Datasets:

krvhrv

/

crawl-books

like 0

the nor- mal distribution. The probability density function for the Cauchy FIGURE L.52 Examples from the RTOG/EORTC late radiation mor- distribution is bidity scoring scheme. 1 é g ù Longitudinal waves PL /C ( x; x0,g) = ê ú p ëê ( )2 x - x 2 0 - g ûú where x0 is the mean value (the centre of the distribution) γ is the full width at half maximum Displacement Direction of wave Hyperlink: http: / /mat hworl d .wol fram. com /C auchy Distr ibu ti on .ht ml FIGURE L.51 For longitudinal waves, the particles oscillate in the same Lossless compression direction as the wave propagation direction. (Courtesy of EMIT project, (General) Image compression can either be lossless or lossy, www .emerald2 .eu) depending on whether the image can be subsequently reproduced Lossy compression 569 L ow contrast detectability Lossy compression is more efficient, but the extent of Image data compression will depend on the tolerance of the observer for L degradation of the image. An example of a lossy file format is JPEG (joint photographic experts group), which allows the user to set a ‘quality factor’ which specifies the amount Transformation of data thrown away. At low-quality factors, JPEG artefacts may appear such as the appearance of ‘blockiness’ within the image. The transformation step for JPEG makes use of Quantisation the discrete cosine transform, to shift the colour data into a more suitable mode for compression and encoding. The next quantisation step (see Figure L.53) is the main source of loss, Encoding Encoding where the colour values are simplified according to the qual- ity factor. Finally the data are encoded using similar methods to the lossless compression, using both run length encoding and Huffman coding. Image file– Image file– Lossy compression is often the compression algorithm of lossless lossy choice for photographs, allowing a useful trade off between qual- ity and size/time. However line drawings or graphs do not lend FIGURE L.53 Steps in compression of images, showing the extra steps themselves to JPEG, where any sharp changes can be distorted. It for lossy files. Lossless files do not undergo the ‘Quantisation’ step which is recommended that images in a master archive are not stored in is the main source of loss. Decompression will simply consist of the steps a lossy format, and it must be noted that frequent retrieval, modi- reversed. fication and archiving will probably degrade the image further each time. Hence lossless file formats should be the first choice exactly (lossless), or whether data are lost in the compression for medical applications, where there is the risk of a misdiagno- (lossy) (Figure L.53). sis from distortions or information lost through the compression Lossless images undergo no quantisation, which is the main process, or where the diagnostic value of the images may not be source of loss for lossy image files. The encoding carried out for known fully. lossless files is based on redundancy reduction schemes, which Abbreviations: DCT = Discrete cosine transform and JPEG = reduce the overall image size but maintain all the information Joint photographic experts group. present. Related Article: Lossless compression Redundancy Reduction Schemes: Firstly the images are treated for interpixel redundancy, taking advantage of any patterns Low contrast present. One example of this is run length encoding (RLE) in which (Diagnostic Radiology) There are several terms specifying the consecutive equal pixel values are replaced with a ‘value-runlength’ subject contrast (mainly used in x-ray radiography). The most couplet. For example, 20 px of value 7, can be instead replaced by often used ones are high contrast and low contrast. the couplet 7–20, which is a reduction of 90% in required storage Low contrast is used to describe images with small differences space. This algorithm works especially well for greyscale images between the optical densities of the adjacent objects. Low contrast and overall achieves compression ratios of around 2:1. detectability is largely influenced by the noise in the image. A The second stage reduces the coding redundancy present. For number of synonyms are used in practice – as soft contrast or example, Huffman coding assigns short codes to the most fre- long-scale contrast. The latter refers to the fact that x-ray films quent values, only using longer codes for those values that occur with big latitude present a large (long) scale of grey levels. For rarely. Although the coding table needs to be saved also, this is more detail see the article on High contrast. usually a small overhead. Related Articles: Subject contrast, High contrast, Film latitude Lossless File Formats: An example of lossless file format is the graphics interchange format (GIF), which allows images to be Low contrast detectability faithfully restored after compression as no information is thrown (Diagnostic Radiology) Low contrast detectability (LCD) is a away. It does have the disadvantage that images with greater than concept that describes the imaging chain’s ability to distinguish 8-bit colour resolution need to be reduced in resolution before they between two objects with slightly different attenuation properties. can be compressed. However this has been addressed in the newer LCD can be used as an image quality parameter to describe the portable network graphics (PNG) file format, which is becoming performance of the imaging system. the scheme of choice for lossless image compression. It hinges on many things including the x-ray spectrum, object/ Lossless file formats should be the first choice for medical anatomy in contention, detector characteristics, noise and the applications, where there is the risk of a mis-diagnosis from dis- observers’ ability to perceive the contrast difference. tortions or information lost through a lossy compression process, TO20/10 and N3 are common test objects that are used in the or even where the diagnostic value is yet unknown. UK to assess the low contrast detectability and threshold. The Abbreviations: GIF = Graphics interchange format, PNG = results can inform contrast detail curves that can provide quanti- Portable network graphics and RLE = Run length encoding. tative evaluations of low contrast and small detail measurements Related Article: Lossy compression of medical images. Further Readings: Alsleem, H. and R. Davidson. 2012. Lossy compression Quality parameters and assessment methods of digital radiog- (General) Image compression can either be lossless or lossy, raphy images. Radiographer 59(2); Hsieh, J. and T. Toth. 2008. depending on whether the image can be subsequently reproduced SU-GG-I-13: Low contrast detectability for X-ray computed exactly, or whether data are lost in the compression. tomography. Int. J. Med. Phys. Res. Pract. Volume 35. Low dose rate (LDR) 570 LSF Low dose rate (LDR) These have a melting point of 69°C and 70°C, respectively, L (Radiotherapy, Brachytherapy) although they are usually worked at temperatures of 90°C and Dose Rates in Brachytherapy: Different dose rates are used 95°C, respectively, since they are easier to pour and mould at in brachytherapy, connected to different treatment techniques. these temperatures. Newer materials such as MCP 96®, melting ICRU, the International Commission on Radiation Units and point 96°C, is an alloy of lead, bismuth and tin with no cadmium Measurements, defined these dose rates in its Report No. 38 ‘Dose content, the latter having a relatively high toxicity. and Volume Specification for Reporting Intracavitary Therapy in Multi-leaf collimators (MLCs) have largely replaced the use Gynaecology’: of blocks from low melting point alloys and lead for shielding of megavoltage photon beams but sometimes blocks are required 1. Low dose rate, LDR when adequate shielding cannot be achieved with MLCs. This a. 0.4–2.0 Gy/h would be the case when the MLC shielding profile is not smooth b. Traditional radium technique; 0.5 Gy/h, 60 Gy with enough for the task because of the leaf width or when a shielded treatment time 5 days island within a field is needed. c. Large amount of clinical data Centres using low melting point alloys for electron treatment d. (NOTE! Ultra low dose rate 0.01–0.3 Gy/h!) normally use a standard thickness of shielding of around 15 mm. This is because for electron energies from 4 to 18 MeV a typical 2. Medium dose rate, MDR thickness of lead shielding would have been 10 mm. Twenty per a. 2–12 Gy/h cent more alloy is needed for lead equivalence so 12 mm would b. More seldom used be required and in practice 15 mm is used, the additional 3 mm is added as a precaution against small air bubbles forming in the 3. High dose rate, HDR alloy as it sets. a. >12 Gy/h = 0.2 Gy/min Output factors must also be measured for these irregular fields b. Treatment times approx. 5–20 min (external beam shaped by cut-outs by comparing the output with standard open therapy belongs here) field. c. Clinical data available Abbreviation: MLCs = Multi-leaf collimators. Related Articles: Block design, Block tray, Custom blocking, 4. Pulsed dose rate, PDR Cerrobend, Block transmission factor, Output factor a. Mimics LDR, using many small ‘HDR pulses’ dur- Further Reading: Podgorsak, E. B. (Technical Ed.), 2005. ing a longer treatment time. Example: 1 pulse per Radiation Oncology Physics: A Handbook for Teachers and hour during 24 h, 0.5 Gy per pulse given in 5 min; Students, International Atomic Energy Agency, Vienna. total dose 12 Gy/day Low-pass filter The radiobiological effects in the tissues irradiated depend on the (General) Image processing of medical images often involves type of applicator used, on the fractionation scheme and on both some kind of low-pass filtering to reduce the noise in the image dose and dose rate distributions. As stated in the ICRU Report that otherwise can be disturbing to the reader of the image. In 38: ‘the clinical experience accumulated with radium techniques the Fourier domain signals are represented by amplitude and fre- cannot be applied to new irradiation conditions without careful quency. In Fourier-transformed images, the frequency has the unit consideration’. This includes consideration of both tumour effects of cm−1. This means that small details are represented by high fre- and effects on normal tissues. quencies and large structures are represented by low-frequencies. Abbreviation: ICRU = International Commission on Radiation A Fourier spectrum describes a signal in terms of amplitudes and Units and Measurements. frequencies. The noise is usually disturbing the image as the type Related Articles: Brachytherapy, Dose rates in brachytherapy, of mottle or ‘salt-and-pepper’, that is local variation in pixel val- see also articles under radiobiology ues that does not have any physiological meaning. This noise will Further Reading: ICRU (International Commission on appear in the high-frequency part of the spectrum. It is therefore Radiation Units & Measurements, Inc.). 1985. Dose and volume the purpose of a low-pass filter to reduce these high-frequency specification for reporting intracavitary therapy in gynecology. amplitudes and let the low-frequency amplitudes pass through the ICRU Report 38, Bethesda, MD. filtering step. A mean-value calculation of a number of samples is also a low-pass filter that reduces the local variations. Mean- Low melting point alloy value filters or weighted mean-value filters are therefore common (Radiotherapy) Frequently part of the radiation field has to be in nuclear medicine work stations. A commonly used 2D filter is shielded to avoid irradiating underlying sensitive structures. This the Butterworth filter that is described as can be done using alloys with a low melting point which can be poured in molten form into a polystyrene mould and then allowed T u v to harden at room temperature. Although in many centres they are ( 1 , ) = 1 ( ( n + , ) 2 D u v / D0 ) largely confined to use with electrons, they are especially useful when complex field shapes are required and blocks can be shaped The filter has two parameters, namely, the order n and the cut-off to account for beam divergence. The alloy can also be melted frequency D0. The order controls the slope of the descent and the down and reused. The use of these materials has replaced lead in cut-off value defines when the descent will start. Note that the cut- many cases which often needed to be milled if complex shapes off frequency is the frequency at which the amplitude has been were required and had a high melting point (327.5°C) making reduced to 0.5. pouring and moulding much more difficult. Popular low melting point products are Cerrobend® and LSF Ostalloy® which are alloys of lead, cadmium, bismuth

and tin. See Line spread function (LSF) L-shell 571 Luminance L-shell LSO (General) The discovery that electrons in atoms occupy different (Nuclear Medicine) Lutetium oxyorthosilicate (Lu2SiO4O:Ce) – a L energy shells was made by Barkla and Moseley from studies of lutetium-based inorganic scintillator activated with cerium, com- x-ray absorption in atoms. These shells were originally labelled monly used in PET detectors. by Barkla with the letters K, L, M, …, Q. At present there are two equivalent nomenclatures used to describe these energy shells: quantum numbers and the x-ray notation of Barkla. Decay Peak Effective Density Attenuation Light Quantum Numbers: Four quantum numbers are required to Constant Wavelength Atomic (g/cm3) Coefficient Yield specify the state of an electron in an atom: (ns) (nm) Number at 511 keV Relative (cm−1) to • The principal quantum number (n) specifies the elec- NaI(Tl) tron energy, it takes values of n = 1, 2, 3, …. • The angular momentum quantum number (ℓ) speci- 40 420 66 7.4 0.87 75% fies the magnitude of the orbital angular momentum; ℓ takes values from 0 to n − 1. It should be noted that the following notation is applied: The relatively high effective atomic number and density of LSO(Ce) mean that unlike traditional NaI(Tl) detectors, it is well suited for the detection of high-energy photons such as ℓ those from positron emitters. It has a greater light yield than 0 s other scintillators of comparable density, e.g. BGO and a 1 p much faster light emission. PET systems with LSO(Ce) scin- tillators are therefore more sensitive than BGO PET systems, 2 d and exhibit reduced dead time effects and better discrimina- 3 f tion between true and random coincidences. The good timing 4 g resolution of LSO(Ce) systems has facilitated the development 5 h of time of flight (TOF) technology in commercially available PET systems. One disadvantage of LSO(Ce) is that it is itself • The magnetic quantum number (m slightly radioactive, with 2.6% of naturally occurring lutetium l) specifies the orien- tation of the orbital angular momentum, it takes values being the long-lived radioactive element,176Lu. This phenom- from −ℓ to +ℓ. enon will increase the number of random events detected in • The spin angular momentum (ms) specifies the spin LSO(Ce) PET systems. angular momentum of the electron, that is spin up (+1/2) Related Articles: LYSO, BSO, NaI(Tl), scintillation detectors, or spin down (−1/2). Time of flight (TOF), Positron emission tomography (PET) Further Readings: Bushberg, J. T., J. A. Siebert, E. M. For example, the electronic structure of carbon, which has six Leidholdt Jr and J. M. Boone. The Essential Physics of Medical electrons, is often written as 1s2 2s2 2p2; this implies that it has Imaging, 2nd edn.; Lewellen, T. K. 2008. Recent developments three occupied orbitals, the two most tightly bound electrons are in PET detector technology. Phys. Med. Biol. 53(17):R287–R317. in the 1s orbital, and there are a further two electrons in both the 2s and 2p orbitals. Lubberts’ effect X-Ray Notation: A second way of labelling electron orbitals is (Diagnostic Radiology) Lubberts’ effect (named after G. to use x-ray notation, a convention commonly employed in x-ray Lubberts, who first published on the effect in 1968) is related spectroscopy. In this notation the principal quantum number (n) is to the use of phosphor materials in x-ray imaging. If x-rays are attributed a letter as shown in the following: absorbed in the upper layer of the phosphorus the light pro- duced at the absorption point travels through the thickness of the phosphorus, thus it spreads and leads to image blur. The Quantum Number Notation X-Ray Notation effect is important for thick phosphorus layers, as in image 1s K1 intensifier, or in indirect flat panel digital detectors using phos- 2s L phorus. In the latter if the light is absorbed in the lower layer 1 2p (ms = +1/2) L of the phosphorus, it reaches immediately the detector and the 2 2p (ms = −1/2) L image is with less blur. The Lubberts effect is related with the 3 MTF and noise in the image and results in decrease of the DQE 3s M1 (detective quantum efficiency) with the increase of the spatial frequency. Hence, there are three L-shells: L1, L2 and L3. Some related terms Related Articles: Detective quantum efficiency, Screen relevant to medical physics include the following: selection Further Readings: Beutel, J., H. Kundel and R. Van Metter, • The L-edge which is the energy of x-rays at which the eds. 2000. Handbook of Medical Imaging, Volume 1 – Physics L-orbital begins to absorb. and Psychophysics, SPIE Press, Washington, DC; Lubberts, G. • To describe an Auger peak three orbitals are needed, 1968. Random noise produced by x-ray fluorescent screens. J. firstly the core hole level, secondly the relaxing elec- Opt. Soc. Am. 58:1475–1483. tron’s initial state and thirdly the emitted electron’s ini- tial state, for example KL1L2. Luminance Related Articles: Atom, Auger electron, Electron (Diagnostics Radiology) See HSL (hue, saturation luminance) LUT 572 LYSO LUT energy photons such as those from positron emitters. The light L (Diagnostics Radiology) See Lookup table yield from LYSO(Ce) is superior to that from other scintillators of similar density, including both BGO and LSO(Ce), meaning LYSO that PET systems with LYSO(Ce) scintillators are more sensitive (Nuclear Medicine) Lutetium-yttrium oxyorthosilicate (Lu and typically exhibit superior energy resolution. The decay con- 2(1- stant is comparable with LSO(Ce) therefore leading to reduced x)Y2xSiO4O:Ce) – a lutetium-based inorganic scintillator activated with cerium, commonly used in PET detectors. dead time effects, better discrimination between true and ran- dom coincidences and time of flight (TOF) capability. As with LSO(Ce), LYSO(Ce) is itself slightly radioactive, with 2.6% of Decay Peak Effective Density Attenuation Light naturally occurring lutetium being the long-lived radioactive ele- Constant Wavelength Atomic (g/cm3) Coefficient Yield ment, 176Lu. This phenomenon will increase the number of ran- (ns) (nm) Number at 511 keV Relative dom events detected in LYSO(Ce) PET systems. (cm−1) to NaI Related Articles: LSO, BSO, scintillation detectors, Time of 41 420 60 7.1 0.86 80% flight (TOF), Positron emission tomography (PET) Further Readings: Bushberg, J. T., J. A. Siebert, E. M. Leidholdt Jr and J. M. Boone. The Essential Physics of Medical Similar to LSO(Ce), LYSO(Ce) has a relatively high atomic num- Imaging, 2nd edn.; Lewellen, T. K. 2008. Recent developments ber and density, making it well suited for the detection of high in PET detector technology. Phys. Med. Biol. 53(17):R287–R317. M M M-mode in the automotive industry. The typical source energies for these (Ultrasound) M-mode, or motion mode, is based on a one systems range between 90kV to 600keV and large area detectors line B-mode measurement, which is continuously repeated (digital detector arrays) are used to acquire images. Shielding (Figure M.1). The penetration depth is shown on the y-axis, and requirements for the implementation of higher energy sources the x-axis shows time (Figure M.2). Stationary reflecting targets (compared to medical CT) may lead to an increase in system are shown as straight lines and moving targets are shown as a weight and hence rotation of the sample instead of the source and repetitive pattern. M-mode is very useful when the degree and detector. This is mainly done with the use of robotic arms. rate of motion are to be measured and are used, for example, for Related Articles: Dragonfly CT, Industrial CT investigation of heart valves. Related Articles: A-mode, B-mode Macroradiography (Diagnostic Radiology) Macroradiography is a radiographic mA selector method producing enlarged x-ray images of the object, by placing (Diagnostic Radiology) The mA selector is part of the fila- it at some distance from the detector/film. The method has advan- ment circuit controls of the high voltage generator (HVG) of an tages in cranial angiography, mammography, etc., but is rarely x-ray equipment. In classical HVG it selects the filament current used these days. Macroradiography requires x-ray tubes with through a variable resistor connected at the primary side of the very small focal spot (below 0.3 mm), as otherwise the enlarged filament transformer. Additionally this circuit includes a set of image will be blurred. If the object is placed mid-way between filament resistors for various combinations (low kV at high mA; the focal spot and the film (i.e. ×2 magnification), the size of the high kV at low mA; low kV at low mA). image blur (geometric unsharpness) will be equal to the size of Contemporary equipment with high frequency generator use a the effective focal spot. In this case, if 0.2 mm blur is accepted system which changes the frequency supplying the filament trans- for this particular image, the focal spot size should be 0.2 mm former, thus varying the filament current (hence the tube current, (i.e. the image of the object is doubled, but the blur is still 0.2 mA). mm). Macroradiography is different (better) than simple optical Direct measurement of the mA is possible only in the second- enlargement, because when magnifying glass is used, all image ary circuit of the high voltage transformer (in the middle point of features (the object and its blur) will be equally enlarged. the secondary winding) – see the block diagram H.25 in the arti- Related Articles: Focal spot, Geometric unsharpness, cle High-voltage generator. During quality control one normally Magnification, Object-film distance, Ultra-fine focus measures the mA variation with kV (mainly the effect of volt- age drop). In this measurement errors above 10% are considered Magic angle unacceptable. Output dose measurement at constant kV can also (Magnetic Resonance) Solid materials feature a rigid arrange- be an indicator for the mA change, but can not supply accurate ment of spins. These constantly experience the local fields of each information for the mA. other leading to inhomogeneous dipole–dipole broadening. From Related Articles: High voltage generator, High frequency gen- a magnetic moment μ at a distance r and polar angle ϕ a second erator, Voltage drop, Filament circuit, Filament resistor spin experiences a Machine learning m Bloc = ± (General) Machine learning is a sub-set of artificial intelligence- 3 (3cos2f -1) r driven methods and algorithms, that provides systems automation via experience-based learning tools, thus reducing the need for as derived from the dipolar Hamiltonian. The ± signs reflect the pre-defined algorithms and human intervention. spin alignment with respect the applied field B0. The solution of Machine learning algorithms classification: the equation 3(cos θ)2 − 1 = 0 → (cos θ)2 = 1/3 or cosq = 1 / 3 that is 55°44′ and called ‘magic angle’. By magic angle spinning • Supervised machine learning algorithms (MAS), the sample at this angle against the B0 field, dipole– • Unsupervised machine learning algorithms dipole couplings could be suppressed in solid state MR spectra. • Semi-supervised machine learning algorithms MAS can also be used to overcome peak broadening in solids • Reinforcement machine learning algorithms due to homonuclear dipolar interaction, chemical shift anisotropy effects and quadrupole field gradient interaction. Dipole–dipole Most common machine learning applications in healthcare include interaction also affects MRI contrast in highly ordered structures but are not limited to diagnostic imaging, radiotherapy planning, like a tendon. At the magic angle, the T2 of the tendon is slightly personalised medicine, electronic health records, research, data increased because collagen, which is responsible for the majority collection, identifying diseases and diagnosis, pharmaceuticals. of tendon composition, has an anisotropic structure. This increase is negligible when TE is long. However, when TE is short as in Macro CT T1 or PD weighted images, the result will be an increased signal (Diagnostic Radiology) Macro-CT systems are usually used for intensity. the inspection of metal parts (>10 cm Al or >1 cm Fe) mostly Related Articles: Chemical shift, Dipole 573 Magnet 574 Magnetic coupling the outer field is removed it is called a ‘hard’ or permanent mag- net and if it disappeared with the outer field it is called ‘soft’. In nature there are some rocks, for example loadstone, that are hard magnets but the usual hard magnet is manufactured. A hard magnet’s natural magnetism decreases when heated M and increases when it is cooled. Related Articles: Electro-magnet, Permanent magnet, Resistive magnet, Superconductive magnets A-mode Magnet, permanent B-mode (Magnetic Resonance) See Permanent magnet Magnet, superconducting M-mode (Magnetic Resonance) See Superconducting magnet Magnetic beam steering (Radiotherapy) Magnetic beam steering is where magnetic fields FIGURE M.1 Principle of A-, B-

and M-mode display. are used to control the direction of charged particles via the Lorentz force. Typically, a series of dipole magnets are used to steer the beam along the vertical and horizontal axis. In particle therapy, this may be utilised along the beam line or to deliver doses by scanning the beam over the target volume. The ability of these magnets to change the field strength is often the restrictive factor for speed in changing energy of the beam. Magnetic coupling (Magnetic Resonance) The term ‘magnetic coupling’ normally refers to what is more correctly termed ‘scalar coupling’ between spins within a molecule. This occurs when two or more nuclei with nonzero spin are connected to each other via chemical bonds, so that the spin state of a given nucleus influences the local magnetic field of the other. The fact that each nucleus can exist in more than one spin state (±½ in the case of a hydrogen nucleus) results in splitting of resonance peaks in the NMR spectrum of the molecule into ‘multiplets’ made up of several components. It is helpful to consider the simple example of a molecule con- FIGURE M.2 M-mode image from the heart region. Observe the small taining two interacting nuclear spins, A and X, each with spin ½. B-mode image with the line indicator. (Courtesy of EMIT project, www .emerald2 .eu) Neglecting scalar coupling, one would expect a spectrum com- posed of two peaks (known as ‘singlets’ in this context). In the presence of through-bond coupling, some of the A nuclei will be Magnet exposed to X nuclei with spin state +½, and others to X nuclei with (Magnetic Resonance) A magnet is a material or object that pro- spin state −½. These two populations of A nuclei will experience duces a magnetic field. Magnets are divided into two groups, the slightly different local magnetic fields, so the spectral peak due permanent magnet and impermanent magnets like electro-mag- to A nuclei will be split into two equal components. Similarly, the nets and superconductive magnets. peak due to X spins will be divided into two due to the presence In all materials there are electrons that rotate in a specific path. of two populations of A spins. The frequency differences between The rotation generates a small magnetic field and if the movement the components in each case are given by a coupling constant, J, of the electrons is coordinated, for example, by an outer magnetic typically 1–15 Hz for coupling between protons and up to 100 Hz field, they will amplify the magnetic field so it can be noticed or so if other nuclei are involved. J is independent of static mag- outside the material. If the generated magnetic field stays on after netic field strength (Figure M.3). A X JA JX (a) (b) FIGURE M.3 Spectrum of two spin ½ nuclei (a) in the absence and (b) in the presence of weak spin coupling. Magnetic dipole 575 Magnetic flux density If a larger number of nuclei are present, a triplet or higher Magnetic field lines order multiplet may arise, depending on the structure of the (Magnetic Resonance) The stray magnetic field outside the bore molecule and the nature of the coupled spins. The impact of of the magnet is known as the stray fringe field. When a magnetic coupling depends on the magnitude of J relative to the chemi- field is generated in a medium the response of the medium is its cal shift, δ, between the nuclei. The behaviour of spins in the magnetic induction B also called flux density. A ferromagnetic strong coupling regime (J ≫ δ) is much more complicated than object aligns itself along the direction of the magnetic flux, which M that of weakly coupled spins (J ≪ δ), and requires quantum- is caused by the presence of a magnetic field in the medium. The mechanical treatment. In the notation used to describe coupled magnetic field lines are a geometrical abstraction, which are used spin systems, strength of coupling may be inferred from the to visualise the direction and strength of a magnetic field. The alphabetical proximity of the letters used to designate the spins direction of the magnetic field lines can be examined by using – for example, AX is a weakly coupled two-spin system and a magnetic dipole. The static magnetic field has no respect for AB a strongly coupled one. The same principles extend to sys- the confines of conventional floors or ceilings. Knowledge of the tems with more than two spins, such as AX3 (e.g. lactate) and extension of the magnetic field lines in the space permits deter- AMNPQ (e.g. glutamate). The strength of coupling may also mination of the area around the MR scanner in which specific depend on static magnetic field strength of the spectrometer safety rules must be followed. The MR site could be conceptually used and weakens at higher fields, improving spectral assign- divided into four zones taking into account the extension and the ment. In addition to SNR gains this points to using higher fields value of the magnetic field isocontours (i.e. lines of equal mag- in MR spectroscopy. netic field strength) Zone I includes all freely accessible areas. Related Articles: Chemical shift, Decoupling Zone II is the interface between the freely accessible Zone I and Further Readings: Homans, S. W. 1992. A Dictionary of the strictly controlled Zones III and IV. It is in Zone II that the Concepts in NMR, 2nd edn., Oxford University Press, Oxford, patient screening is performed as well as the screening of all the UK; Freeman, R. 1988. A Handbook of Nuclear Magnetic persons who have access to Zone III and IV. Zone III is the region Resonance, Longman Scientific & Technical, Harlow, England. in which free access of unscreened persons or ferromagnetic objects or equipment can result in serious injury or death as a Magnetic dipole result of interactions between the individuals or equipment and (Magnetic Resonance) A magnetic dipole is a pair of magnetic the MR static magnetic field. Zone III regions should be physi- poles of equal magnitude but opposite polarity separated by some cally restricted from general public access by suitable physical small distance. Magnetic dipoles can be characterised by the locks. Zone III may project through floors and ceilings of MRI dipole moment equal to the product of the magnetic strength of sites, imposing magnetic field hazards on persons on floors other one of the poles and the distance separating the two poles. The than that of the MR scanner. Zone IV, which is always located direction of the dipole moment corresponds to the direction from within Zone III, is the MR magnet room. The knowledge of the the south to the north pole. Magnetic dipoles are created by cur- value and extension of the isocontours permits identification of rent loops or by quantum-mechanical spin. the area within which the monitoring equipment might not prop- The strength of a dipole magnetic field is given by erly function. / B(r,l) = (M /r3 )( 1 2 1+ 3sin2l) Magnetic flux (Magnetic Resonance) The magnetic flux through a surface area where is the integral of the normal component of the magnetic field r is the distance from the centre times μ over the area: λ is the magnetic latitude (90 − θ) and θ = magnetic colatitude M is the dipole moment ym = òòmH × ds Magnetic dipole moment (Magnetic Resonance) The magnetic dipole moment μ⃗ is the where μ is the permeability of medium (Henry/metre, H/m). most elementary measure of the strength of a magnetic source The SI unit of magnetic flux ψm is the weber (Wb). and it constitutes an intrinsic property of fundamental particles Dimensionally ψm = [ML2/IT2]. The flux is conventionally rep- (quarks and charged leptons). The magnetic dipole moment may resented by imagining lines of induction to be spaced so that the be visualised by imagining the spin of the electric particles as a number through a given area is equal to the flux through that area. spinning gyroscope representing a tiny loop of electric current around the spinning axis. This current loop produces its own Magnetic flux density magnetic field. The connection between the magnetic dipole (Magnetic Resonance) The magnetic flux density (SI unit Tesla) moment of the particle and the spin angular momentum vector is one of the two magnetic vector fields in Maxwell’s equation. J is The magnetic flux through a surface area is the integral of the normal component of the magnetic field times μ over the area: m = gJ where γ is a particle-specific constant called the gyromagnetic ym = ,òòmH × ds ratio. Related Articles: Gyromagnetic ratio, Magnet Further Reading: Haacke, E. M., R. W. Brown., M. R. where μ is the permeability of medium (H/m). The SI unit of mag- Thomson and R. Venkatesan. 1999. Magnetic Resonance netic flux ψm is the weber (Wb). Imaging. Physical Principles and Sequence Design, John Wiley Dividing the magnetic flux by the area A gives the magnetic & Sons, New York. flux density B of flux per unit area. Thus Magnetic moment 576 Magnetic resonance imaging (MRI) y μ B = m = mH A The SI unit of magnetic flux density is Wb/m2 or Tesla, T. + The magnetic flux density B has the same direction as H in M isotropic media with a magnitude μH or B = mH = mrm0H (Wb/m2 or T) FIGURE M.4 Spinning positively charged nucleus produces a magnetic moment μ. where B is the magnetic flux density, Wb/m2 H is the magnetic field, A/m μ One of the consequences of having the magnetic moment is the permeability of medium, H/m μ proportional to the angular momentum is that an atomic magnet 0 is the permeability of vacuum = 4π10−7 H/m μ placed in a magnetic field will precess. r = μ/μ0 is the relative permeability Magnetic polarity Magnetic moment (Magnetic Resonance) Unlike electrical charges, for which the (Magnetic Resonance) The torque exerted on a magnet placed smallest element is the electrical monopole from which the elec- in a magnetic field of unit strength is called magnetic moment. tric field diverges, the elemental source of the magnetic field has Correspondingly the magnetic moment associated to a flat loop two poles of equal magnitude but opposite polarity, referred as a carrying a current and with its plane oriented parallel to the direc- north and a south pole. tion of a magnetic field of unit flux density is equal to the torque Related Articles: Magnet, Magnetic dipole exerted on it. Any charged particle moving in a closed path pro- duces a magnetic field which can be described at large distance as due to a magnetic dipole located at the current loop. The magnetic Magnetic resonance angiography (MRA) moment is a measure of the net magnetic properties of an object (Magnetic Resonance) The term angiography relates to imaging or particle. of vascular structures. Traditional angiography is performed with The torque τ on a magnetic dipole of moment μ in a magnetic x-ray, using a contrast agent. MR angiography or MRA hence induction B is refers to MR examinations that aim to visualise blood vessels and related pathologies. t = m ´ B There are three main MRA techniques: time-of-flight MRA (TOF MRA), phase contrast MRA (PC MRA), and con- This means that the magnetic induction B tries to align the dipole trast enhanced MRA (CE-MRA), respectively. Common to all so that the moment m lies parallel to the induction. If no fric- approaches is the use of 3D imaging sequences providing data tional forces are operating the work done by the turning force will over the vascular tree of interest. 3D data also enable the appli- be conserved. This gives rise to the following expression for the cation of volume-rendering techniques such as maximum inten- potential energy U of the dipole moment μ in the presence of a sity projection (MIP), largely simplifying interpretation of the magnetic induction B: results. TOF MRA relies on the fact that blood flowing into a slice has U = -m × B a higher signal than stationary tissue. Phase contrast MRA uses gradients to encode velocity information in the phase content of A fundamental property of atomic nuclei is that

those with the image. These two techniques do not use any contrast agent. In odd atomic weights and/or odd atomic numbers, for example a CE-MRA, the signal of the blood is further enhanced via a mag- nucleus of the hydrogen atom, which has one proton, possess an netic contrast agent. See related articles for in-depth information angular momentum or spin. Even if the nuclear spin is a property on the different techniques. characterised only by quantum mechanics in the classical vector Related Articles: Contrast enhanced MRA, Maximum inten- model the spin is visualised as a physical rotation similar to the sity projection (MIP), MR angiography MRA, Phase contrast, rotation of a top about its axis. Nuclear magnetism of a nuclear Time of flight (TOF) spin system originates from the microscopic magnetic field asso- ciated with a nuclear spin. A classical argument to justify the Magnetic resonance imaging (MRI) existence of this magnetic field is that a nucleus has electrical (Magnetic Resonance) MRI is the most commonly used abbre- charges and it rotates around its own axis as if it has a nonzero viation for the magnetic resonance imaging technique. Other spin (Figure M.4). abbreviations are, as an example, MRT (magnetic resonance Physically it is represented by a vector quantity called nuclear tomography) or simply MR (magnetic resonance). MRI pro- magnetic dipole moment or magnetic moment. One fundamental vides high quality images of the human body with high spatial relationship of particle physics is that the spin angular momentum resolution without using ionising radiation or invasive tech- and magnetic moment vectors are related to each other by niques. The inherent soft tissue contrast of the method provides differentiation in signal intensity between normal and diseased m = gJ tissue. Recent developments of MR techniques encompass tech- niques for diffusion-weighted imaging (DWI), diffusion tensor where γ is a physical constant known as the gyromagnetic ratio imaging (DTI), perfusion-weighted imaging (PWI), functional and whose value depends on the nucleus, in particular on the ratio MRI (fMRI) and MR spectroscopy (MRS) (Figures M.5 of electric charge over mass (e/m). and M.6). Magnetic resonance imaging (MRI) 577 Magnetic resonance imaging (MRI) At the same year as Hahn discovered spin echoes (leading to Carr and Purcell’s description of the 90°–180°-pulse), the first controlled measurement of the water NMR signal from living systems was made by Shaw and Elsken using a 15 MHz NMR system. The Swedish researcher Eric Odeblad played an impor- tant role in early studies of complex biological systems by deter- M mining relaxation times in animal tissue. Others working in the same area were Bratton and colleagues, and in the 1960s several groups gathered information regarding water T1 and T2 in living systems. Among the conclusions were that the relaxation times were decreased in living systems compared to the free water situ- ation, and early attempts to create discrimination between differ- ent tissue states using NMR were made by Bratton and colleagues in 1965 and by Odeblad in 1968. In 1971, Raymond Damadian found that relaxation rates in malignant rat cells differed from those in normal cells. This finding was rapidly reproduced by others, and may be seen as a starting point for the biomedical use of NMR. After attend- FIGURE M.5 Magnetic resonance angiogram of the head, neck and ing an experiment performed by Hollis, Saryan and Morris, Paul upper thorax. Lauterbur started to investigate the possibilities of creating spa- tial resolution in an NMR experiment, and he successfully cre- ated the first two-dimensional NMR image using continuous wave (CW) NMR in 1973. The basic concept for the realisation of this image was that magnetic field gradients could be used to determine spatial position, since the signal in a well-defined small frequency interval must come from protons having this fre- quency range, and hence from protons at a specific position along the gradient axis. By 1D FT, the signals from different positions can be separated. In order to obtain 2D resolution, Lauterbur used a backprojection technique similar to the one developed for computerised tomography (CT). In 1973, independent of Lauterbur’s experiment, Sir Peter Mansfield used pulsed NMR techniques to obtain spatial resolution in one dimension using one discrete Fourier transform (DFT) of the signal from camphor crystal layers. After the publication of these initial findings, Lauterbur con- tinued his developments and rapidly showed 2D images in living objects, while Mansfield’s group developed a line-scan technique for 2D imaging, including a method for slice selection using a combination of RF excitation and a gradient in one direction and FIGURE M.6 Sagittal T1-weighted magnetic resonance image of the Waldo Hinshaw and colleagues, inspired by Lauterbur’s work, head. developed a ‘sensitive point’ method, in which a point in space was selected using switched gradients. All the early methods thus used the idea of spatial separation by magnetic field gradient The impact of MRI on healthcare and diagnostic imaging has application, but the goal – 2D resolved images – was reached in been tremendous, and the impact of MRI can be compared with different ways. In 1975, Kumar, Welti and Ernst proposed a 2D/3D the introduction of x-rays. The growth of MRI is also reflected in FT method for 2D/3D image reconstruction, and thereby created a the number of installed units at hospitals throughout the world. base for the presently used so-called spin warp technique, which The numbers continue to increase, and several countries in the was described by Edelstein et al. in 1980. In the original proposal, world presently have more than 1 system per 100,000 inhabitants. durations of the encoding gradients were changed between differ- In 2003, Paul Lauterbur and Sir Peter Mansfield were awarded ent views, but in the spin warp technique gradient amplitude was the Nobel prize in physiology or medicine ‘for their discoveries changed instead. As an important part of the current technique, concerning magnetic resonance imaging’. Hoult in 1977 suggested a gradient refocusing method in order The paths of development from the use of the NMR tech- not to lose signal during the slice selective process, and the same nique (see Nuclear magnetic resonance (NMR)) for the study of year Mansfield proposed the use of spin echoes in pulsed imag- biological systems towards today’s MRI technique are indeed ing experiments, as well as a technique for rapid sampling of raw thrilling. signal data, the echo planar technique. Attempts to measure – or at least investigate the possibilities Following the first image of a living object by Lauterbur in for measurements of – NMR signals from living objects were 1974, images of larger objects became possible by building NMR made very early. Bloch inserted his finger into his Stanford spec- units adapted to the new focus of visualising living animals and, trometer and obtained a strong signal while Purcell and Ramsey eventually, humans. Hence, the first image of a human body part in 1948 inserted their heads into the Harvard cyclotron without was published by Mansfield in 1977 using the line scan technique, any observed biological contraindications on themselves. and the first ‘whole-body’ NMR unit, developed by Raymond Magnetic resonance spectroscopy (MRS) 578 Magnetic resonance spectroscopy (MRS) Damadian, an image of the human thorax was obtained the same magnetic resonance in biologic samples. Acta Radiol. 43:469– year. 476; Bratton, C. B., A. L. Hopkins and J. W. Weinberg. 1965. Technique: Medical MRI most frequently relies on the relax- Nuclear magnetic resonance studies of living muscle. Science ation properties of excited hydrogen nuclei in water and lipase. 147:738–739; Odeblad, E. 1968. An NMR-method for determina- When the object to be imaged is placed in a powerful, uniform tion of ovulation. Acta Obstet. Gynecol. Scand. 47(Suppl 8):39– M magnetic field the spins of the atomic nuclei with non-integer spin 47; Damadian, R. 1971. Tumor detection by nuclear magnetic numbers in the tissue all align either parallel or antiparallel to the resonance. Science 171:1151–1153; Hollis, D. P., L. A. Saryan and magnetic field. H. P. Morris. 1972. A nuclear magnetic resonance study of water In order to selectively image different voxels (volume pic- in two Morris hepatomas. Johns Hopkins Med. J. 131:441–444; ture elements) of the subject, orthogonal magnetic gradients are Weisman, I. D., L. H. Bennet, L. R. Maxwell, M. W. Woods and applied. Although it is relatively common to apply gradients in D. Burk. 1972. Recognition of cancer in vivo by nuclear magnetic the principal axes of a patient (so that the patient is imaged in x, resonance. Science 178:1288–1290; Lauterbur, P. C. 1973. Image y and z from head to toe), MRI allows completely flexible orien- formation by induced local interactions: Examples employing tations for images. All spatial encoding is obtained by applying nuclear magnetic resonance. Nature 242:190–191; Mansfield, P. magnetic field gradients which encode position within the phase and P. K. Grannell. 1973. NMR ‘diffraction’ in solids? J. Phys. of the signal. In one dimension, a linear phase with respect to C 6:L422–L426; Lauterbur, P. C., P. D. Klein and S. V. Peterson. position can be obtained by collecting data in the presence of a (eds.). 1973. Proceedings of the First International Conference magnetic field gradient. In three dimensions (3D), a plane can be on Stable Isotopes in Chemistry, Biology and Medicine, pp. 255– defined by ‘slice selection’, in which an RF pulse of defined band- 260, Argonne National Laboratory, Argonne IL; Lauterbur, P. C. width is applied in the presence of a magnetic field gradient in 1974. Magnetic resonance zeugmatography. Pure Appl. Chem. order to reduce spatial encoding to two dimensions (2D). Spatial 40:149–157; Garroway, A. N., P. K. Grannell and P. Mansfield. encoding can then be applied in 2D after slice selection, or in 3D 1974. Image formation in NMR by a selective irradiative process. without slice selection. Spatially encoded phases are recorded in J. Phys. C 7:L457–L462; Hinshaw, W. 1974. Spin mapping: The a 2D or 3D matrix; these data represent the spatial frequencies of application of moving gradients to NMR. Phys. Lett. 48A:87–88; the image object. Images can be created from the matrix using Kumar, A., D. Welti and R. R. Ernst. 1975. NMR Fourier zeug- DFT. Typical medical resolution is about 1 mm3, while research matography. J. Magn. Reson. 18:69–83; Edelstein, W. A., J. M. models can exceed 1 μm3. S. Hutchison, G. Johnson and T. Redpath. 1980. Spin warp NMR The k-Space Formalism: Since x and k are conjugate vari- imaging and applications to human whole-body imaging. Phys. ables (with respect to the Fourier transform) we can use the Med. Biol. 25:751–756; Hoult, D. I. 1977. Zeugmatography: A Nyquist theorem to show that the step in k-space determines the criticism of the concept of a selective pulse in the presence of a field of view of the image (maximum frequency that is correctly field gradient. J. Magn. Reson. 26:165–167; Mansfield, P. 1977. sampled) and the maximum value of k sampled determines the Multi-planar image formation using NMR spin echoes. J. Phys. resolution, that is C 10:L55–L58; Mansfield, P. and A. A. Maudsley. 1977. Medical imaging by NMR. Br. J. Radiol. 50:188–194; Damadian, R., 1 M. Goldsmith and L. Minkhoff. 1977. NMR in cancer: XVI. FOV µ Resolution µ k D max k FONAR image of the live human body. Physiol. Chem. Phys. 9:97–100. (These relationships apply to each axis (x, y and z) independently.) MRI vs. CT: A computed tomography (CT) scanner uses Magnetic resonance spectroscopy (MRS) x-rays, a type of ionising radiation, to acquire its images, mak- (Magnetic Resonance) This term refers to the use of magnetic ing it a good tool for examining dense tissue such as bone. MRI, resonance techniques to investigate chemical composition, as on the other hand, uses non-ionising radio frequency signals to opposed to forming an image. acquire its images and is best suited for non-calcified tissue. When a nucleus is placed in a static magnetic field in order Both CT and MRI scanners can generate multiple two-dimen- to carry out a nuclear magnetic resonance (NMR) experiment, sional cross sections (slices) of tissue and three-dimensional its resonance frequency depends primarily on the identity of the reconstructions. Unlike CT, which uses only x-ray attenuation to nucleus, via the equation ω0 = γB0, where B0 is the static field generate image contrast, MRI has a long list of properties that strength and γ is the gyromagnetic ratio of

the specific nuclear may be used to generate image contrast. By variation of scanning species. However, nuclei are surrounded by electrons, which par- parameters, tissue contrast can be altered and enhanced in various tially shield them from the applied static field. Thus the static ways to detect different features. field experienced by a nucleus will differ from B0 to an extent MRI can generate cross-sectional images in any plane (includ- that depends on the degree of electronic shielding and hence on ing oblique planes). CT is limited to acquiring images in the axial the chemical group in which the nucleus is located. There is a (or near axial) plane. However, the development of multi-detector corresponding variation in resonance frequency, according to the CT scanners with near-isotropic resolution produces data that can equation ω = γBeff = γ(1 – σ)B0, where Beff is the effective field be retrospectively reconstructed in any plane with minimal loss experienced by the nucleus due to the electronic shielding factor of image quality. σ. This phenomenon is known as chemical shift. The frequency Further Readings: Hahn, E. 1950. Spin echos. Phys. Rev. shifts are very small, on the order of millionths of the centre fre- 80:580–594; Carr, H. Y. and E. M. Purcell. 1954. Effects of dif- quency ω0, so that nuclei in all groups can generally be excited fusion on free precession in nuclear magnetic resonance experi- using a narrowband RF pulse. ments. Phys. Rev. 94:630–638; Shaw, T. M. and R. H. Elsken. NMR data collected from a sample containing a given nucleus 1950. Nuclear magnetic resonance absorption in hygroscopic in a range of different chemical groups will therefore produce materials. J. Chem. Phys. 18:1113–1114; Odeblad, E. and G. a signal containing a range of frequency components in pro- Lindström. 1955. Some preliminary observations on the proton portion to the number of nuclei in each type of group. Fourier Magnetic resonance tomography 579 Magnetic susceptibility transformation of this signal will yield a spectrum containing most other hydrogen-containing compounds. Thus proton NMR peaks corresponding to each group. Different chemical com- signals from the body originate overwhelmingly from water, and pounds have different characteristic spectral patterns, which are signals from other compounds are generally not visible. However, used to identify them in a sample containing a mixture of com- lipids are frequently also present in high concentration, resulting pounds. This is the basis of the use of NMR in analytical chem- in so-called chemical shift artefacts. istry, which predates applications in medical imaging by several While MRI is invariably carried out using the hydrogen M decades. nucleus (proton), and specifically hydrogen in water, more gener- NMR spectroscopy came to be applied to biological samples ally MRS may be carried out using any nucleus that is present in and later to live animals and humans, opening the way for medi- sufficiently high concentration in the body, has an NMR-sensitive cal applications. When the technique is applied in a medical con- isotope with sufficient abundance, and has a sufficiently high text, it is generally known as magnetic resonance spectroscopy sensitivity (dependent on the gyromagnetic ratio). In practice this (MRS). Figure M.7 shows a proton NMR spectrum of the brain limits in vivo natural abundance studies to hydrogen and phos- from a healthy human subject. phorus, but carbon and fluorine MRS are feasible using exog- It is important to note that the peaks in an NMR spectrum enous compounds labelled with 13C and 19F respectively. When arise from nuclei in a specific chemical group, but not a specific performing proton MRS in vivo, it is necessary to employ special compound. Thus a given peak may contain contributions from techniques to suppress the overwhelming signal from water and nuclei in the same group in a number of different compounds (e.g. allow other peaks to be observed. This requirement meant that the peak due to methyl protons in choline-containing compounds phosphorus techniques were exploited first, but proton MRS is in proton MRS), while a single compound may give rise to mul- now dominant because of the ease of implementation on a system tiple peaks if the molecule contains the same nucleus in more than designed for (proton) MRI. one chemical group (e.g. the three peaks due to ATP in phospho- In order to carry out MRS in vivo it is necessary to have some rus MRS). means of localising signal acquisition to the desired anatomical The potential clinical utility of MRS arises from the fact that region. The ability to do this was a major breakthrough, since the relative concentrations of different chemical compounds often it allowed acquisition of spectra from specific tissues without change in pathological states. For example, reduction in the level excision from the body. Localisation may be achieved by using of N-acetyl aspartate in proton spectra from the brain is often single voxel selection to focus on a specific volume of tissue, or by indicative of neuronal loss, while reduction in citrate levels in means of chemical shift imaging. spectra from the prostate is a marker for cancer. Using phospho- The resulting MRS spectrum can be quite complex, and rus MRS it is possible to probe cellular energetics and measure sophisticated techniques are needed to identify the component intracellular pH noninvasively. MRS has been variously dubbed peaks correctly and to analyse the spectrum, particularly if quan- the ‘noninvasive biopsy’, the ‘window on metabolism’ and as titative analysis is required. opening up ‘the new neurochemistry’. However, despite over 20 Related Articles: Nuclear magnetic resonance (NMR), years of development few centres perform MRS examinations Chemical shift, Chemical shift artefact, Water suppression, clinically on a routine basis as yet. Single voxel spectroscopy, Chemical shift imaging (CSI), Peak MRS has historical and methodological priority over MRI, assignment, Spectral analysis which may be regarded as a form of proton MRS in which fre- Further Reading: Graaf, R. 2007. In Vivo NMR Spectroscopy: quency differences arise as a result of the imposed gradients (and Principles and Techniques, 2nd edn., John Wiley and Sons, hence are related to spatial position) rather than from chemi- Chichester, UK. cal shift. Chemical shift is a confounding effect in MRI, but its impact is minimised because water is present in tissues in Magnetic resonance tomography a concentration that is orders of magnitude higher than that of (Magnetic Resonance) For magnetic resonance tomography (MRT), see Magnetic resonance imaging (MRI). Related Articles: Magnetic resonance imaging (MRI), Magnetic resonance (MR), MRI See Nuclear magnetic resonance 0.8 Magnetic susceptibility (Magnetic Resonance) Magnetic susceptibility is an inherent 0.6 property of all objects. It is defined as the extent to which the material becomes temporarily magnetised when it enters a mag- netic field such as the main magnetic field of the MRI system. 0.4 Within the body, bone has the lowest magnetic susceptibility and molecules that contain iron, such as haemoglobin, have the highest magnetic susceptibility. At the boundaries between tis- 0.2 sues there is a change in susceptibility which creates a phase change within the magnetic field which creates an artefact in the image. 0.0 There are many methods which can be used to reduce the ppm effect of magnetic susceptibility. Spin echoes are less susceptible 4 3 2 1 to T * 2 due to the 180° refocusing pulse that corrects the suscepti- bility-induced dephasing of the spins. A short TE can be used to FIGURE M.7 Proton NMR spectrum of grey matter in the normal reduce magnetic susceptibility as it allows less time for dephas- human brain. ing and reduces signal loss. A large receiver bandwidth will also Magnetisation preparation 580 Magnetisation transfer contrast (MTC) help as it shortens the minimum TE available. Phase change (Δϕ) Fast IR-prepared gradient echo caused by susceptibility is given by 180° αααα αα α α α α RF Df = g ×Gi × Dr ×TE (M.1) TI M where G γ is the gyromagnetic ratio of hydrogen phase Gi is the internal magnetic field Δr is voxel size ADC TE is the echo time Equation M.1 shows that magnetic susceptibility increases with voxel size and echo time. Hence minimising the voxel size S and echo time will also reduce the magnitude of the magnetic Inversion susceptibility effect. module TI Abbreviation: TE = Echo time. Related Article: Spin echo Magnetisation preparation FIGURE M.8 Structure of a fast IR-prepared gradient echo sequence. A 180 pulse inverts Mz. During TI, Mz undergoes T1 relaxation and after (Magnetic Resonance) The term magnetisation preparation is fre- TI, a train of RF pulses with low flip angles are executed. quently used to identify pulse sequence parts (see Pulse sequence) where the magnetisation is manipulated (prepared) prior to the readout of the signal in order to improve sequence performance Further Reading: Mugler, J. P. 3rd. and J. R. Brookeman. with respect to signal-to-noise, contrast or acquisition time. A 1990. Three-dimensional magnetization-prepared rapid gradient- common type of magnetisation preparation is the application of a echo imaging (3D MP RAGE). Magn. Reson. Med. 15(1):152–157. 180° pulse inverting the longitudinal magnetisation before apply- ing a spin-echo, a fast spin echo or a gradient-echo sequence (see Magnetisation transfer (MT) the following examples). Another type of preparation is applied (Magnetic Resonance) See Magnetisation transfer contrast on already existing transverse magnetisation, which is driven (MTC) back to the longitudinal axis using a 90° pulse, enabling reduced repetition times in so-called driven equilibrium sequences. Magnetisation transfer contrast (MTC) (Magnetic Resonance) Magnetisation transfer (MT) is a contrast EXAMPLE 1 mechanism in tissue that relies on the fast exchange (or cross- relaxation) between bound and free protons. The bound protons By starting a pulse sequence with a 180° inversion pulse, are associated with rotationally restricted macromolecules or thereby inverting the direction of Mz, and introducing a hydration layers. This bound (or restricted) pool is invisible to waiting time denoted TI before readout, objects with dif- direct imaging due to the very short transverse relaxation times ferent T1 will undergo different amounts of longitudinal (T2 about 10 μs). The bound pool has a very broad absorption relaxation during TI. In combination with a 90°–180° pulse line and can be saturated by an RF-pulse several kHz from the train (a spin echo sequence), an inversion recovery experi- resonance frequency of the free pool. The subsequent transfer ment is created (see Inversion recovery) with an increased between the two pools will move longitudinal magnetisation from T1 contrast range relative to a conventional 90°–180° spin the free pool to the pre-saturated bound pool, thus reducing the echo sequence. observable total MR signal (Figure M.9). Fluids including CSF and blood, fat and bone marrow are little affected by MT attenuation. Tissues rich in structural material EXAMPLE 2 Absorption A 180° pulse can also be used in a fast gradient echo sequence. In this case, a fast train of low-flip angle RF pulses are applied after the inversion time, creating T1 con- Off resonance RF Signal trast in the resulting image. Using short repetition times excitation reduction and low flip angles, each k-space line is sampled with rea- sonably constant signal along the longitudinal relaxation Free pool curve (Figure M.8). This method, frequently denoted magnetisation-prepared rapid gradient-echo imaging (MP Bound pool RAGE), enables fast three-dimensional imaging with T1 Frequency contrast (1). ω0 FIGURE M.9 Protons in the free pool will not be directly affected by Related Articles: Inversion recovery, Inversion time (TI), the off resonance RF-pulse. The signal in the free pool will decrease due Pulse sequence to the MT from the bound pool, affected of the RF-pulse. Magnetophosphenes 581 M agnification such as myelinated white matter are strongly affected by MT and Cavity Cathode the MR signal in these tissues is decreased. resonator MT is often used to reduce the background in time-of-flight angiography and can also be used for better visualising multiple sclerosis (MS) lesions in T2-weighted and in gadolinium contrast Probe enhanced T1-weighted images. M Magnetisation transfer effects may also introduce errors in certain applications. For example, in arterial spin labelling (ASL) the RF-pulse for inverting the spins before they enter the area of interest can also excite spins in the interesting area (bound pro- tons). The exchange of protons between the bound and free pool will decrease the signal in the tissue that is not supposed to be affected, creating a false decrease in perfusion values. Related Articles: Absorption, RF-pulse, T Anode

2 Further Reading: McRobbie, D. W., E. A. Moore, M. J. Graves and M. R. Prince. 2003. MRI: From Picture to Proton, FIGURE M.10 Section of magnetron construction. Cambridge University Press, Cambridge, UK. Magnetophosphenes This is picked up by an antenna and passed along an intermediate (Magnetic Resonance) Phosphenes are the flickering visual sen- waveguide to the main accelerating waveguide. sations caused by non-photic stimulation such as pressure on the Magnetrons are cheaper than Klystrons and due to their small eyes and mechanical shocks. They are generated in the retina size, can be mounted within the gantry. This simplifies the con- and not in the optic nerve or the visual cortex. Gradient mag- necting waveguide structure. However their performance is less netic fields induce electrical currents which may excite the optic reliable than Klystrons, and their frequency output fluctuates as nerve and/or retina of the patient undergoing the MR examina- they are sensitive to the earth’s magnetic field. tion at the threshold of perception. This excitation results in the Related Articles: Klystron, Linac visual sensation of flashes of light named magnetophosphenes (magnetic field-induced phosphenes). The same effect could be Magnification obtained also by rapid movements of the head or eyes in a 4 T (Diagnostic Radiology) The magnification of an image is a func- static magnetic field while for lower strength of the magnetic tion of both the focal-to-film distance (FFD) and object-to-film field there is no induction of magnetophosphenes by head or distance (OFD); it is defined as eye movements. Magnetophosphenes have been elicited by cur- rent densities of as low as 17 μA/cm2. It has been suggested that FFD m = (M.2) the threshold for the induction of magnetophosphenes is age FFD - OFD dependent resulting in a lower value in younger subject. Since magnetophosphenes show the lowest threshold response to slow The image magnification is 1 when the object being imaged is time-varying magnetic fields, it has been the practice to use their placed on the film cassette surface, that is the OFD = 0. appearance in the production of guideline to limit exposure to As the OFD is increased the magnification increases, how- MRI. The magnetophosphenes have never been regarded as ever, this also increases geometric unsharpness illustrated by an harmful for the patients. increased penumbra in Figure M.11. To overcome this, the focal Further Reading: Budinger, T. 1981. Nuclear magnetic spot is reduced as the magnification is increased. For example, in resonance (NMR) in vivo studies: Known thresholds for health mammography the FFD is typically 60 cm, if the breast is offset effects. J. Comput. Assist. Tomogr. 5:800–811. Magnetron Focal spot (Radiotherapy) Magnetrons are used in linear accelerators to pro- duce high power microwaves, which are used to accelerate the electrons in the waveguide. An alternative option for the micro- FOD: wave source is a Klystron. Focus-object distance A magnetron is an RF oscillator that extracts energy from electrons in a resonant structure within a magnetic field. It con- FFD: sists of a copper anode block surrounding a central oxide coated Focal-film cathode, as shown in Figure M.10. The anode has symmetrical distance resonant cavities around the cathode, and the whole structure sits within a uniform magnetic field in a vacuum. When a DC pulse OFD: is applied between the cathode and anode, a bunch of electrons Object-film distance are ejected from the cathode and accelerate towards the outer anode. However their path is deflected by the large magnetic field present perpendicular to the plane of electron motion, and hence they form circular orbits, gradually spiralling out to the anode. Increased penumbra The electron distribution is bunched into rotating radial lines, like spokes of a wheel. As each spoke passes an outer cavity, an FIGURE M.11 Magnification and geometric unsharpness of planar electrical field is induced which forms the microwave radiation. radiograph. Main supply circuit, power supply 582 Mammographic phantom by 20 cm the magnification will be 60/40 = 1.5. To reduce the Handbook, Academic Press; World Health Organization. 2011. increased geometric unsharpness produced by the magnification Medical Equipment Maintenance Programme Overview: WHO the focal spot size is reduced which reduces the penumbra. Medical Device Technical Series, World Health Organization. Abbreviations: FFD = Focal-film distance, FOD = Focus- object distance and OFD = Object-film distance. mA-metre M Related Articles: Geometric unsharpness, Focal spot, Focal- (General) An mA-metre (milliampere metre) is an instrument film distance Focus-object distance, Object-film distance, used to measure electric current in an electric or electronic cir- Penumbra cuit. Electric current is measured in amperes (A). mA-metres are often found in x-ray equipment to measure the current in the Main supply circuit, power supply anode circuit of the x-ray tube. (General) A source of power suitable for operating of electrical Related Articles: Operational amplifier, Voltmeter and electronic devices, electrical power supply in most cases. Alternating-current power supply is obtained from mains. Mammographic phantom Direct-current power supply is obtained from batteries, but can (Diagnostic Radiology) Mammographic phantoms are used to also be obtained by rectifying and filtering from alternating-cur- simulate human breasts in quality control tests. Depending on the rent supply or by a power converter. goal of the measurements, different test objects are required. A typical design DC power supply that uses mains consists AEC (automatic exposure control) performance of a mammo- of a transformer, rectifier and a filter. Many supplies also have a graphic equipment can be easily tested using 0.5 cm or 1 cm thick voltage regulator at the output. PMMA plates. Semi-circular or rectangular shapes are available. Related Articles: Line voltage, Voltage regulation The reference phantom is 4.5 cm thick, used to simulate a 5.3 cm thick standard breast (composition: 50% adipose tissue – 50% Mains frequency glandular tissue) but also 2, 3, 4, 5 and 6 cm thick slabs are com- (General) See Line voltage monly adopted to simulate different breast thickness. These cheap objects are suitable to evaluate the repeatability Mains voltage in selecting exposure parameters (Anode, Filter, kV and mAs) by (General) See Mains supply circuit; Power supply AEC and to evaluate AGD (average glandular dose) versus object absorption. Mains voltage drop By adding an aluminium plate (side 10–15 mm, thickness (Diagnostic Radiology) See Voltage drop 0.2 mm) on the top of the PMMA blocks, the SDNR (Signal- difference-to-noise ratio) versus object absorption can also be easily estimated (Figure M.12). Maintenance PMMA slabs are used also to evaluate digital detectors perfor- (General) Maintenance refers to any act or intervention on a mance (response function, noise, uniformity, artefacts, inter-plate medical device, usually performed by a biomedical or clinical variability for Computed Radiography). engineer or technician, aiming to ensure its correct functioning, For image quality assessment, a uniform object is not adequate preventing or correcting any possible fault and allowing its cor- and different types of commercial products exist. rect functioning and operation. According to the WHO, there are Most of the phantoms have a uniform background including three kinds of maintenance that are more commonly performed one or more physical objects to simulate anatomical details, in on a medical device: known locations (Figure M.13). Although they fail to represent the complexity of the human breasts, which can have very different • Inspections: A series of regularly scheduled actions patterns and various lesion types, they are very useful in QC tests, aiming at inspecting the performance and safety of because they offer the great advantage of being reproducible. medical equipment. A common and very easy image quality assessment is a • Preventive maintenance: An intervention that is usu- human-based evaluation. It consists of counting the visible details ally performed along inspections. It includes a series in the phantom image. A rule to assign the scores is defined (typi- of actions (e.g. lubrication, cleaning or replacement of cally: 1 is assigned for well/totally visible detail, 0.5 for not-well/ parts, etc.) intended to extend the life of the medical partially visible detail, 0 if the detail is not visible at all). The sum device and prevent any fault or failure. of scores is taken as overall image quality index and a minimum • Corrective maintenance: An intervention that is usually threshold can be set for the total score and, separately, per each performed after any fault in the medical equipment is type of detail. found or reported. It entails a series of actions intended The human-based methods have inherent limitations, due to to restore the correct functioning, safety and perfor- intra- and inter-observer variability. mance of the medical device. Using suitable software, digital images allow for the calcula- tion of a large number of quantitative image quality indices, used Data coming from maintenance and inspections are usually to describe the so-called technical image quality. These param- reported in management maintenance systems, which can be eters are objective, reproducible and more sensitive than human- paper-based or computerised and help the management of health based methods in assessing image quality differences. technologies within healthcare settings. Technical image quality is not equivalent to the clinical image Related Articles: Standards, Medical equipment manage- quality, which depends also on the wide variability in human ment, Medical device perception and decision criteria. In general, the improvements Further Readings: Almir, Badnjevic and Lejla Gurbeta Pokvić. in technical image quality are likely to predict improvements in 2020. Medical devices maintenance. In Clinical Engineering clinical image quality. Mammography (screen film) 583 Mammography (screen film) M FIGURE M.12 Semi-circular and rectangular plates. (PIXMAM – image courtesy of Leeds Test Objects.) FIGURE M.13 TOR-MAX Leeds Test Object. (Image courtesy of Leeds Test Objects.) QC protocols suggest minimum acceptable level and achiev- Physical contrast in mammography able level for image quality indices. During acceptance/com- missioning test, these phantoms are exposed at the dose level suggested from the manufacturer, to verify the performance of the system. The value of each IQ index defines a baseline (with relative tolerance) to be used for long-term reproducibility tests. The same object is exposed to monitor the performance of the Mass system during the time (long term reproducibility), through ‘con- stancy test’. Hyperlink: www .leedstestobjects .com/ #phantoms -heading Fat Mammography (screen film) (Diagnostic Radiology) Receptor Calcium Introduction: Mammography is radiography of the breasts. It is significantly different from general radiography of other anatomical sites because of characteristics of the breast anatomy FIGURE M.14 Physical characteristics in the breast that must be and pathologic conditions (Figure M.14). imaged with mammography. (Courtesy of Sprawls Foundation, www Equipment Design: The breast is composed of soft tissue. .sprawls .org) Generally there is a background of adipose (fat) tissue which contains slightly higher density normal glandular structures and radiography systems are the ability to form images with high con- pathologic masses if they are present. The very small differences trast sensitivity and low blurring (Figure M.15). in density among the tissues result in low physical contrast and The most common mammography system uses a molybdenum require an imaging procedure with high contrast sensitivity. x-ray tube anode and a molybdenum filter. The x-ray spectrum Small calcifications are important signs of some cancers, produced with this combination gives a high contrast sensitivity especially in the early stages when they are the most treatable. that is generally optimised with radiation dose to the breast. Visualisation of the small calcifications requires an imaging pro- Some systems are designed with an alternative rhodium filter cedure with low blurring. Therefore, two of the characteristics to be used with the molybdenum anode or an alternative rhodium that make a mammography system different from conventional anode track and rhodium filter combination. These are used to Mammography x-ray tube 584 Mammography x-ray tube Anode effect. To make the spectrum as narrow as possible a molybde- Molybdenum Focal spot num anode with a molybdenum k-edge filter is the most common Rhodium 0.1–0.3 mm system. Some units have an additional rhodium filter and some Filter also have an additional anode of rhodium or tungsten. This type of x-ray tube has low output so a 65 cm SSD is usually used. In M addition a film with a high gamma is used to improve the contrast. This film is also low noise so small differences can be visualised. High Resolution: In order to achieve high resolution a small focal spot is used. This is about 0.3 mm. In addition, to avoid Compression cross-over effect, a single emulsion screen film system is used. The

film is mounted in a cassette with the screen behind the film. Grid The overall resolution is about 14 lp/mm with the screen film sys- Film/screen tem having a resolution of 20 lp/mm. Mammography Unit: The unit has a short SID because of Receptor the low output. The x-ray tube is tilted so to improve resolution. The tube is mounted so that the central ray passes through the FIGURE M.15 General design characteristics of a mammography sys- chest wall edge of the breast instead of the centre of the recep- tem. (Courtesy of Sprawls Foundation, www .sprawls .org) tor. This improves visualisation of all of the breast tissue. The unit has a grid of about 5/1. Mammography is done using many projections so the c-arm of the gantry rotates. The unit has a pho- increase breast penetration, especially when imaging the more totimer which is mounted behind the cassette. This is necessary dense breasts. because the photodetector cannot be made sufficiently radiolucent The breast is compressed to produce a more uniform thick- for mounting in front of the receptor. The unit is equipped with a ness, a thinner breast, and to reduce motion during the imaging breast compression system. exposure. Compression has a number of benefits: Both film/screen and digital receptor/display systems are used for mammography. The necessary characteristics are a relatively • It immobilises the breast. This is important since expo- wide exposure dynamic range (film latitude), high contrast trans- sures of greater than 2 s are not unusual. fer and low blurring (Figure M.16). • It spreads the tissues to improve diagnostic accuracy. Technical Factors: The x-ray beam spectrum is controlled by • It reduces breast thickness which reduces Compton the combination of anode material, filter, kV and mA. Selections scatter. are made based on breast size and density to optimise contrast • It makes the breast thickness more uniform. This sensitivity with respect to radiation dose. The selections are made reduces the dynamic range of the x-ray signal and there- either manually by the operator or automatically by some systems. fore allows for higher contrast film. Exposure to the receptor is generally controlled automatically but can be adjusted by the operator for specific breast conditions. The larger of the two focal spots is used for general mammog- The units also have a system for magnification imaging which is raphy and the small spot is used for the magnification technique done without a grid and a 0.15 mm focal spot. which produces the least amount of blurring and the best visibility Because mammography units are so specialised they have a of detail, especially the small calcifications. variety of specialised quality assurance techniques which are nec- High Contrast: To achieve high contrast the imaging is done essary to achieve constancy of image quality. One of the earliest at low kVp. This is usually in the range of 24–32 kVp. This means and most well known was designed by the American College of that most of the contrast is generated using the photoelectric Radiology. Related Article: Digital mammography Anode Mammography x-ray tube Anode M o Molybdenum Focal S L (Diagnostic Radiology) X-ray tubes for mammography Rhodium Filter (Figure M.17) are designed with specific features to provide Filter R h images with two necessary characteristics, high contrast sensitiv- 2 6 MA 1 0 0 ity for visualising soft tissues including cancers and visualisation KV of detail including micro-calcifications. MAS 1 0 0 The contrast sensitivity is determined by the spectrum of the Time 1 . 0 0 x-ray beam that should be optimised to the size and density of a breast in relation to radiation dose. This is achieved using the characteristic radiation produced with anodes of molybdenum or Compression rhodium. Molybdenum is the standard in mammography and rho- Spectrum Exposure dium only in some ‘dual’ anode tubes allowing the operator to Priority C O N T Density + 1 select between the two. Beryllium (Z = 4) is used as the tube window because of its low Automatic attenuation and minimal filtering effect. Molybdenum is the most common filter, but rhodium is also available in some systems. FIGURE M.16 Technique factors that must be adjusted to optimise Tubes for mammography have small focal spots to minimise each mammography procedure. (Courtesy of Sprawls Foundation, www blurring and enhance visibility of anatomical detail, especially .sprawls .org) calcifications. The smallest focal spot is used in the magnification Man sievert 585 Manual afterloading aims at giving a ‘uniform dose’, that is a dose variation (outside the high-dose regions around each source) of about ±10%, to the target volume using radium needles. Sources of varying strength are used, the source strength distribution is non-uniform with more source strength at the periphery of the target volume and sources should be crossed at the ends. Tables are available giv- M ing the mg*h needed to deliver the specified doses for different implant sizes, provided that the implant rules are followed. Related Article: Dosimetry system Further Readings: ICRU (International Commission on Radiation Units & Measurements, Inc.). 1985. Dose and vol- ume specification for reporting intracavitary therapy in gyne- cology. ICRU Report 38, Bethesda, MD; ICRU (International Commission on Radiation Units & Measurements, Inc.). 1997. Dose and volume specification for reporting interstitial therapy. ICRU Report 58, Washington, DC; Gerbaulet, A., R. Pötter, J.-J. Mazeron and E. van Limbergen. (eds.). 2002. The GEC ESTRO FIGURE M.17 Schematic diagram of an X-ray tube for mammography. Handbook of Brachytherapy. Available at the ESTRO web site: (Courtesy of Sprawls Foundation, www.sprawls.org) www .estro .be Manganese technique that provided the highest possible visualisation of small (General) calcifications. Collimation is used to establish the field of view. This can include a ‘half beam’ with a defined edge of the most intense Symbol Mn radiation for the thickest region of a breast near the chest wall Element category Transition metal and reduced intensity because of the heel effect extending beyond Mass number A of stable isotope 30 (100%) the nipple. Atomic number Z 25 Atomic weight 54.9380 kg/kg-atom Man sievert Electronic configuration 1s2 2s2 2p63s2 3p64s2 3d5 (Radiation Protection; General) This is the radiation unit given Melting point 1519 K to collective effective dose. For more information, see the article Boiling point 2334 K on Collective dose. Related Article: Collective dose Density near room temperature 7210 kg/m3 (7.21 g/cm3) Manchester system History: Manganese is found freely in nature. In prehis- (Radiotherapy, Brachytherapy) toric times the element was adopted as a pigment in paint and it Manchester System – Intracavitary Brachytherapy for was also later used by Egyptians and Romans in glass-making. Cervix Cancer: The (intracavitary) Manchester system is based Nowadays manganese has many industrial uses, most notably as a on the use of a combination of appropriately sized ovoids and intra- component in a range of alloys (including stainless steel) and as a uterine tubes loaded in a standard way (originally Ra-sources). cathode material in disposable batteries. All combinations of ovoids and intrauterine tubes are designed Medical Applications: Defibrillation: Manganese dioxide to give the same dose rate to the unique dosimetry point A, pro- button cells have a relatively stable voltage during discharge, such vided that the applicators are positioned correctly in the uterus that they are employed in the medical field for defibrillation of and upper vagina. The Manchester system is a ‘time system based the heart. on the use of standard applicators’. Two points, A and B, are defined in the classical Manchester Manual afterloading system. Point A, the specification point of the Manchester system, (Radiotherapy, Brachytherapy) is defined as a point 2-cm lateral to the uterine canal and 2-cm Source Handling and Loading: The brachytherapy source/s up from the mucous membrane of the lateral fornix of the vagina must be handled and loaded into the applicators for treatment, and in the plane of the uterus. In clinical practice, point A is often many methods have been used over the time. These methods have defined 2-cm up along the axis of the central tube, from the lower been developed primarily to reduce the dose to the personnel, but end, and 2-cm away laterally (two points, left and right). Several also to improve the quality of the treatment itself. definitions have been used in different centres, and the definition Manual afterloading is used when has in principle changed from an anatomy related to an applicator related point. The different definitions of the specification point 1. Applicators, needles, catheters, etc., are inserted make it difficult to compare results from different centres. 2. Correct applicator positions are verified using dummy Point B is a reference point defined on the transverse axis sources through points A, 5 cm from the mid-line (two points, left and 3. It is necessary to improve the accuracy in applicator right). Point B is representative for the dose in the vicinity of the positioning, as there is no risk of dose to staff pelvic wall and for the dose to the obturator lymph nodes. 4. The sources are inserted into the applicators manually Manchester System for Interstitial Brachytherapy: The (interstitial) Manchester system, the Paterson–Parker system, See Source loading in brachytheraphy Manual loading 586 Mass energy absorption coefficient Related Articles: Brachytherapy, Source loading in brachy- mAs-metre therapy, Manual loading, Manual afterloading, Remote afterload- (Diagnostic Radiology) See mAs selector ing, Remote afterloading unit Mass attenuation coefficient Manual loading (Radiation Protection) The amount of attenuation in an absorber M (Radiotherapy, Brachytherapy) is related to the mass or density of the absorbing material. In Source Handling and Loading: The brachytherapy source/s terms of its mass, the attenuation coefficient is given by must be handled and loaded into the applicators for treatment, and many methods have been used over the time. These methods have m/r been developed primarily to reduce the dose to the personnel, but where also improving the quality of the treatment itself. μ is the linear attenuation coefficient Manual loading is used in historical methods, for example ρ is the density of the material when handling radium sources, which were manually introduced and removed; treatment times vary from a number of hours to Related Articles: Attenuation, Linear attenuation coefficient several days. It was also used when speedy insertion techniques were mandatory and when the patient was a radiation source dur- Mass collision stopping power ing the long treatment. (Radiotherapy) The mass collision stopping power is the rate of See Source loading in brachytherapy energy lost by the charged particle resulting from the sum of the Related Articles: Brachytherapy, Source loading in brachy- soft collisions, causing the excitation and ejection of an electron therapy, Afterloading, Manual afterloading, Remote afterloading, which carries a relatively small energy, and the hard collisions Remote afterloading unit that result in the ejection of an electron with relatively large energy transfer (delta ray). Markus chamber For electrons the mass collision stopping power is given by (Radiation Protection) Plane-Parallel chamber was designed in 1975 by Markus to fulfil the requirements of a Bragg–Gray detec- æ dE ö NAZp × r2 2mec 2 tor in electron beams especially at low energy – i.e. to produce a ç - rdx ÷ = 0 è ø Ab2 coll negligible disturbance to electron fluence present in the irradiated medium. Later the design was improved together with chamber é 2 E ù ´ê æ K ö material selection and the Markus chamber is still commercially lnç ÷ + ln (1 + t / 2) + F (t) - dú ëê è I available. ø ûú where mAs selector NA is the Avogadro’s number (Diagnostic Radiology) The mAs selector differs from the mA Z is the atomic number of the medium selector by the fact that it is also linked to the Timer of the HVG r0 is the classical electron radius of an x-ray equipment. This way, the mAs selector selects and me is the electron rest mass measures the quantity of charge (Q = mA × s) during the expo- c is the speed of light in vacuum sure, which directly corresponds to the dose during the exposure. A is the atomic mass of the medium For this purpose the anode current is integrated over the time of β = v/c is the speed of the particle relative to c exposure (using an integrator). The eventual timing error in the EK is the kinetic energy

mAs devices should be related to the fact that 25% change in I is the mean excitation energy of the medium mAs of an exposure is clearly visible (and often related to expo- τ = EK/mec2 sure point in some of the exposure tables used in the radiography δ is the density effect correction practice). Related Articles: High voltage generator, mA selector, Mass defect Filament circuit, Exposure point (Nuclear Medicine) Mass defect refers to the apparent loss of mass when particles form atoms. The mass defect stems from the Mask mode fluoroscopy fact that a bound system has a lower energy than the unbound con- (Diagnostic Radiology) Mask mode fluoroscopy is one of the ini- stituents and therefore has a lower mass than the sum of its parti- tial names of a method used for digital subtraction angiography cles in a free state. When free particles are binding, the excessive (DSA). In this most common method subtraction is performed energy is released as thermal vibrations or photon emissions, that between the two images – the second one (with contrast) and the is the mass is ‘transported’ to another location. first one (without contrast), called mask. The other DSA method is time interval difference subtraction, Mass energy absorption coefficient where the mask is not fixed, but changes in time (i.e. subtraction (Radiation Protection) The energy absorption coefficient, μa, is is performed between the image with contrast and one of the pre- a linear function of the density of the medium. This dependence vious images with less contrast). This method gives information can be avoided using the mass energy absorption coefficient, μa,m, about the changes of contrast in time. defined as follows: The third main DSA method is K-edge subtraction, where the two images are made with different kV – one just below the ma,m = ma /r K-edge of the contrast media, the other – just above this K-edge. In fact this is dual energy subtraction, and as the contrast medium where is most often Iodine, the two energies are around its K-edge. ρ is the density of the material Related Article: Digital subtraction angiography μa,m has dimension L2/M1 and is usually given in cm2/g Mass number 587 Matching layer The radiation intensity attenuation factor associated with Related Articles: Source strength, Contained activity, absorption, exp[−(μa · t)], where t is the absorber thickness, can Equivalent mass of radium, Apparent activity, Reference air be rewritten as kerma rate (RAKR), Air kerma strength exp éë-(ma,m × x)ùû Mass radiative stopping power (Radiotherapy) The mass radiative stopping power is the rate of M where x, defined as the product x = ρ ⋅ t, is the so-called mass energy lost by the charged particle resulting in the production of thickness. bremsstrahlung radiation. The mass energy absorption coefficient can be used to calcu- For electrons the mass radiative stopping power is given by late the absorbed dose in a medium. For instance, if the radiation intensity in a point of the medium, I [(J/(cm2 s)], and the energy absorption coefficient, μa,m [cm2/g], are known, the product (μa,m æ dE ö N Z 2 ç - ÷ = s A éE + 2 K mec ù × B ( , ) I) gives an estimate of absorbed dose at that point. è r 0 r Z EK dx ø A ë û rad The mass energy absorption coefficient does not eliminate the dependence on radiation energy and on the composition of the where medium; hence, in practical conditions an average mass energy NA is the Avogadro’s number absorption coefficient is needed. Z is the atomic number of the medium Related Articles: Energy absorption coefficient, Absorbed me is the electron rest mass dose, Average mass energy absorption coefficient c is the speed of light in vacuum A is the atomic mass of the medium Mass number EK is the kinetic energy (Nuclear Medicine) The mass number is the sum of protons and Br (Z, EK) in the range of energies 0.5–100 MeV varies between neutrons in an atomic nucleus. All atoms of a certain element have 5.33 and 15: the same number of protons but the number of neutrons can dif- fer, that is the same element but different mass numbers. Atoms s = a[e2 (4pe 2 ) 2 = 5 0 0 28 0 / 0mec ] .8 ×1 - × cm2 /atom of the same element but with different mass numbers are referred to as isotopes. where α is the fine structure constant. Mass of radium From this relationship it can be seen that the energy loss by (Radiotherapy, Brachytherapy) Calibration of source strength is bremsstrahlung increases directly with Z, the atomic number of a very important part of a comprehensive brachytherapy quality the medium and more slightly with the electron energy. system. The instruments, ion-chambers and electrometers, used for source strength determinations should have calibrations that Mass stopping power are traceable to national and international standards. (Radiotherapy) The average linear rate of energy loss per unit Specification of source strength for photon emitting sources: of path length x is called stopping power. Dividing the stopping Source strength for a photon emitting source can be given as a power by the density of the medium it results in a new dosimet- quantity describing the radioactivity contained in the source or as ric quantity called mass stopping power. Common units for the a quantity describing the output of the source: mass stopping power are MeV cm2/g. In the mass stopping power the dependence on the density medium is removed except for 1. Specification of contained activity the polarisation effect. The mass stopping power does not differ a. Mass of radium in mg greatly for materials with similar atomic composition. In addi- b. Contained activity in Ci or GBq tion the mass stopping power in a gas is independent on the pres- 2. Specification of output sure because dividing the stopping power by the density exactly a. Equivalent mass of radium; mg Ra eq compensates for the pressure. The mass stopping power could be b. Apparent activity subdivided into a mass collision stopping power and a mass radia- c. Reference exposure rate tive stopping power: d. reference air kerma rate e. Air kerma strength æ 1 dE ö æ 1 dE ö æ 1 dE ö ç - r dx ÷ = ç - r dx ÷ + ç - è ø è ) ø è r dx ÷ col ørad When brachytherapy was introduced as a treatment modality, the only sources available were radium sources. Source strength was Related Articles: Mass collision stopping power, Mass radiative given as the mass of the radium contained in the encapsulated stopping power source. The filtration of the encapsulation was also given; usu- ally 0.5 mm Pt for needles and 1–2 mm Pt for tubes. (The UK’s Matching layer National Physical Laboratory [NPL], a standards/measurement (Ultrasound) The matching layer improves the sensitivity in ultra- institute established at the turn of the last century, acquired its sonic transducers by making the sound transmission more effi- first radium standard in 1913, made by Marie Curie, and specified cient. This is achieved by reducing the reflections in the boundary in terms of mass of Radium.) between transducer and tissue. In modern brachytherapy dosimetry, reference air kerma rate The matching layer has an acoustic impedance value between or air kerma strength is the quantity used to calculate absorbed the impedance of the transducer element and the soft tissue and dose. (Kerma is the kinetic energy released in matter.) should ideally be: Zm = Zst * Zt where Zm is the acoustic imped- See Source strength for a full description of specification of ance for the matching layer, Zt is the impedance of the transducer source strength. and Zst is the impedance of the soft tissue. Matrix array 588 Matrix array λ/4 PZT M Tissue 6 4 5 3 2 M 1 Pixels λ 1 2 (a) Column electrodes 3 V 0 V Row 1 4 V V/2 V/2 Pixel 11 Pixel 12 Pixel 13 6 5 Row 2 V/2 V/2 V/2 Pixel 21 Pixel 22 Pixel 23 FIGURE M.18 Use of matching layer (M) between two media (PZT and Row 3 soft tissue) with differing acoustic impedances. The reflected waves are V/2 V/2 V/2 cancelled by destructive interference between wave 3, 4 and 6. (Courtesy Pixel 31 Pixel 32 Pixel 33 of EMIT project, www .emerald2 .eu) Co C lu o m l C u o (b) n m lu 1 n m 2 n If the thickness of the matching layer is exactly a quarter of 3 a wavelength then 100% of transmission through the transducer/ tissue boundary is possible. What happens in the transducer is FIGURE M.19 (a) Passive matrix of 3 × 3 elements. (b) Circuit dia- that the ultrasound beam reverberates inside the matching layer gram for the passive matrix, the selected pixel 11 is the given voltage and the reflected beams coming back into the transducer disc will V while the non-selected pixels experience cross-talk and are given a cancel out and the beams coming into the tissue are reinforced. voltage of V/2. This is due to the fact that a reflection from a medium with lower acoustic impedance will have a 180° phase shift of the pulse, but a reflection from a medium with higher impedance will have no to the correct voltage all other pixels in the column are also given phase shift (Figure M.18). a small voltage which partially switches them on, this is referred Many transducers have multiple matching layers to provide to as cross-talking. In the example of a 3 × 3 matrix shown in efficient transmission for a spectrum of ultrasound frequencies. Figure M.19b the selected pixel is given a voltage V, whereas the With a well-designed matching layer the demands of the backing surrounding pixels acquire a voltage V/2. (2) Once a pixel row layer will be reduced. has been addressed, the state is not maintained while other rows Related Articles: Backing layer, Reflection coefficient within the matrix are addressed and the pixel state will degrade as the root mean squared of the pixel capacitance. (3) The refresh Matrix array rate of the display is slow. Due to these limitations passive matri- (Diagnostic Radiology) Matrix arrays are used in diagnostic ces are limited in size and have a low contrast ratio. They are not imaging technology when a large number of either detectors or used for medical imaging; instead, active matrices are used. display elements need to be addressed simultaneously. In diag- As opposed to passive matrices, each element within an active nostic radiology matrix arrays are used in both flat panel displays matrix contains a switching element (usually a thin film transis- and x-ray detectors. Matrix arrays are classed as either active or tor, TFT). This switching element allows every element to be passive. individually addressed and for the element state to be maintained Passive matrices are only used in old monochrome liquid crys- when other elements are being addressed. Active matrices can tal display (LCD) devices such as older laptop and mobile phone be produced as a large area matrix (currently in excess of 40 × displays. They are small matrix devices of limited contrast and 40 cm2), which allows it to be used as a fundamental constitu- refresh rate. The matrix is formed by a grid of electrodes with ent in modern digital x-ray detectors and modern LCD displays. each pixel element formed by the overlapping regions of the elec- In medical imaging detectors the active matrix array is used for trodes (Figure M.19a). Row electrodes are referred to as scanning both direct and indirect radiography. Arrays used for both types electrodes and column electrodes as data electrodes. To address of imaging incorporate a two-dimensional array of imaging pix- a specific pixel the row electrode is switched to ground (0 V) els, which consists of a switching element used for data read-out and the column electrode is set to the desired voltage, V (Figure (typically a TFT) and a sensing and storage element. M.19b). The display is addressed row-by-row. Passive arrays suf- The active matrix utilises thin film technology which allows fer several major setbacks: (1) Although the selected pixel is set the deposition of hydrogenated amorphous

silicon (a-Si:H), Row electrodes Matrix array 589 Matrix array Matrix array Multiplexer ADC (Ultrasound) The term matrix array describes an array in which Charge amplifier the elements are arranged both along the main access of the transducer and also in the elevation plane so that better con- trol of the transmitted and/or received ultrasound beam in the Charge collector elevation plane can be achieved. Matrix arrays can be used to electrode M improve slice thickness in a conventional B-mode image or to provide volume acquisition for 3/4 D displays and multiplanar displays. Matrix arrays are sometimes categorised as 1.25-, 1.5- and 2D arrays. In 1.25D arrays, outer elements are switched on to Switch, improve focusing in the far field and switched off for the near field diode or TFT (Figures M.21 and M.22). If focusing is only required in receive mode, then dynamic aperture control can be performed electroni- cally by incorporating outer elements in the received signal from deeper tissue. Gate line 1.25/1.5/2D arrays for slice thickness focussing Date line FIGURE M.20 Active matrix array and peripheral electronics. making it ideal for construction of both TFTs and photodiodes. Large area arrays are formed by plasma deposition of thin layers of the appropriate materials (e.g. amorphous silicon) onto a glass substrate, once deposited they can then be etched to the desired pattern by a process called photolithography (Figure M.20). Figure M.20 shows a typical array used in a medical imag- ing flat panel detector. Within the array each pixel consists of the switching element and an element to detect incoming photons and Inner elements Outer elements Each element fires store them as charge. The image read-out process is controlled by used for near paired for individually-can altering the voltage applied across the switching element. Firstly, field + outer symmetrical produce offset beam to allow each pixel to detect a signal during exposure the voltage for far field dynamic focussing across each switching element is set to an ionisation or ‘off’ state. The signal is then read-out by changing the switching voltage row- FIGURE M.21 1.25D, 1.5D and 2D matrix arrays have different func- by-row to the conducting or ‘on’ state which allows the charge tions. In 1.25D and 1.5D arrays, the additional rows of elements are used stored in each pixel to be drained by the charge collector electrode to improve control of slice thickness in a conventional array. and passed to the multiplexer. The voltage change is controlled by the gate line driver. As the read-out process is controlled by the external circuitry, each row of pixels requires a separate control line driver to alter the switching voltage, and each column its own amplifier. This process is called the active matrix read-out. The active matrix array allows the radiographic image signal to be read-out sequentially, line by line. Fluoroscopic images are acquired in real-time by permitting all other rows that are not being read-out to continue to detect the incoming signal during exposure. Abbreviations: AMA = Active matrix array and TFT = Thin filmed transistor. Related Articles: Thin film technology (TFT), Amorphous silicon, Flat panel detector, Liquid crystal display (LCD), Active matrix array, Active matrix liquid crystal flat panel display Further Readings: Rowlands, J. A. and J. Yorkston. 2000. Flat panel detectors for digital radiography. In: Handbook of Medical Imaging: Volume 1. Physics and Psychophysics, eds., J. Beutel, H. L. Kundel and R. L. Van Metter, SPIE Press, Washington, DC, pp. 223–313; Yang, D. K. and S. T. Wu. 2006. Liquid crys- FIGURE M.22 In conventional arrays (left) focusing in the slice thick- tal display matrices, drive schemes, and bistable display. In: ness is achieved by using an acoustic lens. With a 1.25D array, each ele- Fundamentals of Liquid Crystal Devices, John Wiley and Sons ment is subdivided. Inner elements are used for focusing at near depth Ltd., West Sussex, UK, pp. 274–276. (middle) and the outer elements are used to improve focusing at depth (right). Gate line driver Matrix depth 590 Matrix size Matrix size (Diagnostic Radiology) The matrix size of a digital image is the number of pixels contained in the image. This is often expressed A by giving the number of pixels in each dimension or direction of the image. Matrix size has a direct effect on two important image M characteristics; pixel size and the numerical size or number of bytes to record the image. Pixel size is a factor in determining the spatial resolution or detail of the imaging system. When an image is digitised into a matrix of pixels each pixel becomes an additional source of blur- ring that adds to the other sources within the system, such as focal-spot size and blurring within the receptor. Pixel size is the ratio of the image size, or field of view (FOV), to the matrix size. Increasing the matrix size decreases the pixel size and therefore improves the resolution and visibility of detail. Respectively the use of a larger FOV with the same matrix leads to decreased resolution. Often the matrix of medical imaging systems is square – i.e. with equal number of pixels in X and Y direction. A digital radio- graphic system with a matrix size of 2048 × 2048 pixels will produce a pixel size of approximately 0.2 mm (400/2048) for its FIGURE M.23 With independent control of outer elements (A), improved focusing over a larger depth range can be achieved. maximal image field size (400 × 400 mm). This provides a spatial resolution of 2.5 lp/mm. If the image field of view is smaller (for example 200 × 200 mm), the pixel size will be decreased, and the spatial resolution will be increased – 200/2048 will present pixel size of about 0.1 mm. Similarly, if the system matrix size is doubled to 4096 × 4098, the pixel size will be 0.1 mm (400/4096) and the spatial resolution would be 5 lp/mm. If a CT scanner has a 512 mm FOV and the matrix is 1024 × 1024 pixels, the pixel size will be 0.5 mm (512/1024) and the spa- tial resolution will be 1 lp/mm. However, if the scanner has col- lected a sufficient number of projections, then the raw data can be used for subsequent reconstruction of another smaller image. For example, an image with a FOV of 128 mm can be reconstructed. Matrix In this case the pixel size of the final image will be 0.125 mm array (128/1024), and a spatial resolution of 4 lp/mm. The example above assumes a new reconstruction of the scan from the collected projections. If a displayed image is zoomed to a smaller displayed FOV the spatial resolution will be the same, but the pixels will be optically enlarged on the monitor. It is the matrix (voxels) size at the location of the anatomy being imaged that determines detail and resolution. Another dimension or parameter of a digital image in addition to matrix size (the number of pixels), is the pixel bit depth – the depth of each pixel expressed as the number of bits to record the FIGURE M.24 2D array can be used to provide volume imaging and signal in each pixel. This parameter relates to the contrast char- reconstruction of images in several planes. acteristics of the image. If the pixel has 8 bits it will be able to present 28 = 256 levels of grey (or colours). There is a difference between the bit depth (bits per pixel) In 1.5D arrays, the outer elements are independently con- needed and used to acquire images and for the display of images. nected symmetrically about the central axis (Figures M.21 and A relatively large pixel bit depth (16 bits per pixel) might be used M.23). This permits more versatile control of slice thickness to record both digital radiographs and CT images. This gives a throughout a range of depths by using different pulses and tim- large exposure dynamic range. The images that are then pro- ings of the outer elements. Typical commercial systems use 5–7 cessed and windowed for display do not require as many bits per rows of elements. pixel. In a 2D array (Figures M.21 and M.24), the elements are Contemporary digital images used in medicine have a matrix individually addressed and can be used to image a volume of with depth of 16 bits, where usually 12 bits are used to record the tissue, permitting 3D and 4D imaging and C-plane imaging. image contrast, and the other 4 bits are used for supporting infor- Current commercial systems are reported to use arrays of 50 × mation (for example text or graphs displayed over the image). The 60 elements. 12 bits present 212 = 4096 levels of grey (or colours) is more than enough for the human visual system. 4096 levels of grey is also Matrix depth completely sufficient for various densitometric measurements (Diagnostic Radiology) See Matrix size (measurement the optical density of the pixel, corresponding to Maxillofacial cone beam CT 591 Maxillofacial cone beam CT the radiation absorption of the respective voxel from the anatomi- thick or thin, planar or curved reconstructions in any orientation cal object). (Figure M.25). Historically, the selection of 4096 grey levels has been based Maxillofacial CBCT is replacing dental scan with conven- on the early CT scanners, where the difference between CT num- tional CT, because this technique: bers of air and water has been accepted as 1000, while the most absorbent bones are up to 3 times this absorption difference, thus • Is less expensive M forming a CT number scale from –1000 (air) through 0 (water) • Uses lower radiation dose to +3000 (bone with high absorption), what equals to 4000. The • Guarantees high resolution, optimal visualisation of practice has shown that 4096 levels of grey are also sufficient for mineralised structures and accurate images various densitometric measurements (measurement the optical • Is more rapid and generally during the examination the density of the pixel, corresponding to the radiation absorption of patient is standing or in a natural seated position the respective voxel from the anatomical object). A final example – an image with a numerical size of 2048 × At the same time, also some disadvantages exist, because: 2048 × 16 (also called 4 megapixel matrix with 16 bits depth), will contain approx. 67 Megabits. If the bit depth is reported in bytes • Contrast resolution is limited (1 byte = 8 bits), the raw image file size will be 8 Megabytes (8 • Bone density values are device-depending (which MB). This file size can be reduced by applying various methods implies severe limitations for quantitative evaluation of for image compression. the tissue from CBCT images) • The regions close to dental restorations and implants Maxillofacial cone beam CT (high-density structures) have reduced image quality (Diagnostic Radiology) Maxillofacial Cone beam computed tomography (CBCT) technique is based on a single rotation of the The main clinical applications of Maxillofacial CBCT are sur- ‘source – detector’ assembly around the patient. The source is a gery (dental implantology, impacted and supernumerary tooth, cone-shaped x-ray beam and the image receptor is a digital detec- oral and maxillofacial pathology, maxillofacial traumatology, tor (usually a flat panel detector or sometimes an image intensi- temporomandibular joint disorders, dento-maxillo-facial discrep- fier/charge-coupled device [CCD] detector). During the rotation, ancies and cleft palate), orthodontics, periodontology, endodon- a set of planar images at various projection angles is generated. tics and otolaryngology Afterwards, these projections are reconstructed to volumetric The equipment operates under automatic exposure control, images using algorithms similar to those used in classical com- usually at a fixed kV value. Two or more axial acquisitions are puted tomography. required to verify the correct positioning. Furthermore, the operator can extract a series of 2D cross-sec- The operator can select Field of view (FOV) dimension and tional images (axial, sagittal and coronal views) and, moreover, VOXEL dimensions (75 µm ÷ 200 µm). FIGURE M.25 CBCT images including coronal, sagittal, axial, and 3D views. Maximum dose 592 Maximum exposure time, AORD FOV can be distinguished in For example, the MET of a worker to a narrowband ultraviolet source emitting around 311 nm (like in the figure in this section), Large FOV (for example: 24 cm × 19 cm; 16 cm × 16 cm; can be estimated as the ultraviolet limit a

= 30 Jm–2 divided the 15 cm × 12 cm) source effective irradiance at 311 nm. Medium FOV (for example: 15 cm × 5 cm; 12 cm × 2 cm; If the source has an irradiance of 10 mWcm–2, the effective M 10 cm × 10 cm) irradiance at 311 nm can be estimated as: Small FOV (for example: 8 cm × 8 cm; 8 cm × 5 cm; 5 cm × 5 cm) EEff = 10 ´ S (311 nm) mW cm-2 Large FOV are selected for ‘cranio-facial’ studies; medium and where S (l) is the ultraviolet action spectrum small FOV are used for dento-alveolar investigations. Examples of different FOV selection are illustrated in Figure M.26. EEff = 10 ´ 0.0111 = 0.111 mWcm-2 Maximum dose Hence (Nuclear Medicine) The maximum dose refers to the location MET = 3 mJcm–2/ (0.111 mWcm–2) = 27s with highest registered absorbed dose. The term is typically used in dosimetric calculations, for example internal and external radiotherapy. In non-quantitative measurements, the dose is usu- ally normalised to the location with highest dose. Maximum exposure time, AORD (Non-Ionising Radiation) The maximum exposure time (MET), according to the AORD, is the maximum time in a typical eight- hour working shift that a worker can be exposed to an artificial optical source without any additional protection. The time depends on both the intensity of the source and the spectral regions of emissions. In the case of non-laser sources, the MET is calculated as the ratio between the exposure limit value in the region/s of interest and the effective irradiance for the skin. Whilst in case of ocular exposures, which tend to be non-cumulative, one compares the effective radiance to the limit Related Articles: Action spectra, AORD, Exposure limit val- itself. ues, Irradiance, ICNIRP, Eye, Radiance, Skin FIGURE M.26 FOV in CBCT. Maximum frequency follower 593 Maximum likelihood expectation maximum (MLEM) Further Readings: Coleman, A., F. Fedele, M. Khazova, P. Freeman and R. Sarkany. 2010. A survey of the optical haz- ards associated with hospital light sources with reference to the Control of Artificial Optical Radiation at Work Regulations 2010. J. Radiol. Prot. 30(3):469; ICNIRP A closer look at the thresholds of thermal damage: Workshop report by an ICNIRP task group. M Health Phys. 111(3):300–306; ICNIRP Guidelines on limits of exposure to incoherent visible and infrared radiation. Health Phys. 105(1):74–91, 2013; ICNIRP Guidelines on limits of expo- sure to laser radiation of wavelengths between 180 nm and 1,000 µm. Health Phys. 105(3):271–295, 2013; ICNIRP Guidelines on limits of exposure to ultraviolet radiation of wavelengths between 180 nm and 400 nm (incoherent optical radiation). Health Phys. 87(2):171–186, 2004; ICNIRP Revision of the guidelines on limits of exposure to laser radiation of wavelengths between 400 nm and 1.4 µm. Health Phys. 79 (4):431–440, 2000. Maximum frequency follower FIGURE M.27 Sonogram from carotid artery. (Courtesy of Tim (Ultrasound) The method for obtaining the maximum and mini- Hartshorne, Leicester Royal Infirmary, Leicester, UK.) mum frequency followers (frequency envelopes) from the sono- gram has been described by Gibbons et al. (1981) and improved by Evans et al. (1989a,b) and consists of comparing the spectrum The sonogram on Figure M.27 is showing maximum velocity of the Doppler shift signal with a threshold and marking the first envelope and mean velocity envelope, as well as fiducial mark- and the last points where the spectrum is greater than the thresh- ings for start of systole, peak velocity and end of diastole (cross old. The main advantage of using the maximum frequency fol- markers). lower rather than the mean frequency follower is that the former Related Articles: Doppler ultrasound, Sonogram is less affected by the thump wall filter, which removes power Further Readings: Evans, D. H., W. N. McDicken, R. at low frequencies both from wall movements (which is the rea- Skidmore and J. P. Woodcock. 1989a. Doppler Ultrasound son why it is used) and from slow moving blood (an undesir- – Physics, Instrumentation and Clinical Applications, John able but unavoidable artefact). Furthermore, a comparison of Wiley & Sons, Chichester, UK, pp. 49–50; Evans, D. H., F. S. the behaviour of the mean and maximum frequency allows a Schlindwein and M. I. Levene. 1989b. An automatic system for qualitative estimation of the flow profile: with parabolic flow the capturing and processing ultrasonic Doppler signals and blood mean frequency should be close to half the maximum while for pressure signals. Clin. Phys. Physiol. Meas. 10(3):241–251; plug flow both followers will be approximately the same (Evans Gibbons, D. T., D. H. Evans, W. W. Barrie and P. S. Cosgriff. et al. 1989a,b). 1981. Real-time calculation of ultrasonic pulsatility index. Med. Since the Doppler shift produced by a moving target is pro- Biol. Eng. Comput. 19:28–34. portional to the velocity of the target, and since there are a large number of targets (scatterers) moving with a range of velocities Maximum likelihood expectation maximum (MLEM) in a blood vessel, the Doppler shift signal is a wide-band signal (Nuclear Medicine) The most commonly used iterative recon- containing information about the velocity distribution of blood struction methods in commercially available systems are the within the sample volume. maximum-likelihood expectation maximum method (commonly The best way to obtain information from the Doppler shift sig- denoted the MLEM) and its accelerated version called ordered- nal is to perform real-time spectrum analysis, enabling the assess- subsets expectation-maximum (OSEM). In practice, the equation ment of the evolution in time of the blood velocity distribution that describes the process is given by and the usual way of displaying the resulting information is the old Doppler sonogram (link). I new Ii P i = h j ji å h å From each of the individual spectra that make up the sono- ji j å h I old jk k gram one can obtain the mean frequency (proportional to the j k mean velocity), which is the ‘normalised first moment’ where I is the image to be created N -1 å P is the measured projection pi ´ i f = i=0 N -1 (M.3) hji is the probability (sometimes called the transfer matrix) that å p the pixel i i=0 i will contribute to the projection bin j where Pi is the intensity of each individual frequency bin In its simplest form that assumes no photon attenuation, no i is the counter of frequency bins in the Fourier transform of scatter contribution and collimator blur, hij takes the value of the Doppler shift signal Δf(t) unity along the ray-of-view for the current projection angle. This then reduces the formula to Joining the points obtained from Equation M.3 over time pro- duces the intensity weighted mean frequency follower and, pro- new Iold I i P = j vided that the angle of insonation is known, the mean velocity i å h å å I old ji j k follower. j k Maximum (minimum) intensity projection 594 Maximum permissible concentrations The summation term under I old i is needed because the backpro- jection step is a summation step and therefore a normalisation with the number of projection angles is essential to keep the num- ber of counts in the reconstructed images the same as have been acquired. M Related Articles: OSEM, Reconstruction Maximum (minimum) intensity projection (Diagnostic Radiology) Maximum intensity projection (MIP) and minimum intensity projection (mIP) are volume rendering tech- niques for image display. A two- or three-dimensional model of CT data set (or part of it) can be created by displaying the maxi- mum or minimum CT numbers encountered along the viewing angle. These techniques are used, for example, for CT angiogra- phy (MIP) or tracheobronchial system (mIP) for optimal visuali- sation of contrast differences between different tissues. Related Article: Volume rendering Maximum (minimum) intensity projection (MIP) (Magnetic Resonance) The MIP is a very useful image rendering technique for three-dimensional MR angiographic data. A pro- jection line is defined through the slab at a certain angle, and the projection value is given by the maximum pixel intensity along the projection line. In this way, bright structures form the projec- FIGURE M.29 Maximum intensity projection (MIP) of the MR angi- tion image, suppressing background with lower signal intensity. ography based on the image shown in Figure M.23. This property makes the MIP well suited for both TOF MRA and contrast-enhanced MRA (Figures M.28 and M.29), as these tech- Maximum permissible concentrations niques visualise blood vessels as bright. (Radiation Protection) Maximum permissible concentration is Clinically, a set of radial projection lines are often defined, the maximum quantity per unit volume of a radioactive material and scrolling through the resulting images gives the impression in air, water and foodstuffs that is not considered an undue risk of viewing the vascular tree from different angles. to human health. An analogue to the maximum intensity projection is the -mini- In order to evaluate the maximum permissible concentration mum intensity projection, often abbreviated MinIP. in case of occupational exposure to radionuclides, ingestion and Related Article: Magnetic resonance angiography (MRA) inhalation dose coefficients shall be taken into account. This means the committed effective dose per unit intake via ingestion corresponding to different gut transfer factors (i.e. the proportion of the intake transferred to the body fluid in the gut) for various chemical forms; and the committed effective dose per unit intake via inhalation for the default lung absorption types (fast, moderate and slow). The aforementioned quantities are calculated taking into account the latest models for the respiratory tract, the bioki- netic models for systemic activities and the appropriate transfer factors for the components of the intake cleared from the lungs to the gastrointestinal tract. The inhalation and ingestion dose coefficients and all the nec- essary parameters for the calculation of the occupational expo- sure in relation to all known radioisotopes in use are provided by International Commission on Radiological Protection (ICRP) and International Atomic Energy Agency (IAEA) publications. The various concentrations present will determine the number of working hours in the given conditions. In case of public exposure, the same methodology is applied, taking into account the different dose limits for the public and the total permanence in the environment. Different age groups, namely less than 1, 1–2, 2–7, 7–12, 12–17 and more than 17 years, are considered. Increased transfer factors are used for infants from 0 to 3 months. The maximum permissible concentration allowed for each radionuclide depends on the earlier-explained parameters and effects on the human body. The total amount of radionuclides in FIGURE M.28 One slice from a 3D CE-MRA dataset showing the renal the working place and in the environment are established tak- arteries. ing into account the fact that the dose limits for occupationally Maximum permissible dose (MPD) 595 Maxwell gradients exposed workers and for public, respectively, shall be respected. exposure limit value (ELV) is equivalent to MPE and is used in As a result the competent authorities (at local and national levels) the Control of Artificial Optical Radiation at Work regulations. are responsible for issuing authorisations, monitoring the working MPE values are laser specific and are dependent upon the conditions and checking the environment in order that the maxi- wavelength, pulse duration or exposure time and the tissue that mum permissible concentrations of radioisotopes in food, air and is exposed. In the retinal hazard region (380–1400 nm) the size water are compatible with the maximum permissible amounts in of the retinal image must also be considered when calculating M the body. the MPE. Methods for calculation of MPE levels are described Further Readings: IAEA (International Atomic Energy in PD IEC/TR 60825-14:2004, Safety of laser products: A user’s Agency). 1996. International basic safety standards for protection guide. against radiation and for the safety of radiation sources. Safety MPE values are used for risk assessment means and in estab- Series No. 115, International Atomic Energy Agency, Vienna, lishing the boundaries required for laser-controlled areas. Austria; ICRP publications 66 (1994) and 68 (1994). Related Articles: Exposure limit value (ELV), Retinal hazard region, Controlled area (Laser) Maximum permissible dose (MPD) Further Readings: Medicines and Healthcare Products (Radiation Protection) Up to the 1950s, all dose limits set for Regulatory Agency, Lasers, intense light source systems and workers occupationally exposed to ionising radiation was based LEDs – guidance for safe use in medical, surgical, dental and aes- on avoiding deterministic effects. In other

words, workers were thetic practices, Crown copyright, September 2015; Health and not allowed to reach a threshold dose. No consideration was Safety Executive, Control of Artificial Optical Radiation at Work taken in setting dose limits for the risk of causing cancer or Regulations 2010, S.I no. 1140; www .l egisl ation .gov. uk /uk si /20 10 other stochastic effects in the population of exposed workers. /11 40 /pd fs /uk si _20 10114 0 _en. pdf; PD IEC/TR 60825-14:2004, From 1934, the limit set was called the maximum permissible Safety of laser products: A user’s guide. dose. Evidence emerging from the survivors of the atomic bombs at Maximum target absorbed dose Hiroshima and Nagasaki, together with the follow-up of patients (Radiotherapy) The International Commission on Radiation undergoing radiotherapy for non-malignant disease such as anky- Units and Measurements (ICRU) recommends a common method losing spondylitis, and subsequently developing cancer, con- of dose specification which could be generally adopted to permit vinced the international community that a dose limit merely set to a comparison between the treatment practices. Normally a non- avoid deterministic effects was not good enough. uniform dose distribution is obtained in the target volume and In 1956, ICRU proposed a revised complete system of radia- therefore for practical reasons it is useful to report specific doses tion quantities to describe radiation exposure and resultant dose such as the maximum target absorbed dose. to a human, together with a set of radiation units. These were The maximum target absorbed dose is the highest absorbed exposure, unit the Roentgen, radiation absorbed dose to mea- dose in the target area that can be regarded as ‘clinically mean- sure the deposition of energy in tissue, unit the rad, and the ingful’. The latter term implies that at least a minimum area is Roentgen equivalent man to measure the equivalent whole body irradiated to the dose level designated as ‘maximum’. The mini- effect of a partial or series of partial body exposures, unit the mum area recommended for this purpose is 2 cm2, unless the rem. whole target area is less than 4 cm2, in which case a minimum The revision of the system of units by ICRU then prompted area of 1 cm2 should be taken to define the maximum target ICRP a year later to issue new advice on dose limitation, revis- absorbed dose. ing the concept of the maximum permissible dose (MPD) to take The ICRU indicates the 2 cm2 value, considering this as the full account of both the avoidance of deterministic effects, and a smallest area for which the absorbed dose can be calculated with consideration of the limit of acceptability of exposing the working confidence. population to a stochastic risk of cancer. Related Articles: Mean target absorbed dose, Minimum tar- In defining this new limiting dose, they also determined that get absorbed dose, Modal target absorbed dose, Median target the annual MPD should be reduced to 5 rem (equivalent to 50 absorbed dose, Hot spots mSv in the modern unit, that is 2.5 times the current dose limit Further Reading: ICRU (International Commission on for radiation workers). They also introduced the concept of cumu- Radiation Units and Measurements). 1978. Dose specification lative dose into the dose limits by specifying the annual MPD in for reporting external beam therapy with photons and electrons. terms of age at exposure with the formula ICRU Report 29, Washington, DC. MPD = 5(N -18) rem Maximum velocity (Ultrasound) See Maximum frequency follower where N is the age of the worker at exposure. Maxwell gradients This formula indicated that radiation workers could not be less (Magnetic Resonance) In an idealised application of a gradient than 18 years of age. Above that age their annual dose had to be field in MRI, a linear magnetic field strength variation is intro- an average of 5 rem over their working lifetime. duced along one direction in the bore of the scanner. However, Related Articles: Roentgen, Absorbed dose, Deterministic Maxwell’s equations demonstrate that it is not possible to produce effects, Stochastic effects an isolated linear variation in just one direction. Additional gra- dients in the orthogonal directions called ‘Maxwell’ or ‘concomi- Maximum permissible exposure (MPE) tant’ gradients will also exist. (Non-Ionising Radiation) The term maximum permissible expo- With the application of a gradient of amplitude G in the sure is specific to lasers. The MPE is the highest level of laser z-direction, the z component of the field H at point a distance z exposure to the eye or skin that is considered safe. The term from the isocentre is given by Mayneord F factor 596 Mean dose per cumulated activity 1. Hz = H0 + Gz Control where room 2. H = (Hx, Hy, Hz) The divergence of the vector H is M ¶H ¶H 3. Ñ × H = G + x + x ¶y ¶y Rat hole From Maxwell’s equations, 4. ∇ · H⃗ = 0 so Maze ¶H 5. x ¶H + x = G ¶y ¶y The rate of change of field with position (i.e. the gradient) along a plane orthogonal to the applied gradient is seen to be non- zero. These gradients are the Maxwell or concomitant gradients. Maxwell gradients are a source of small but unwanted spatial vari- ations in the resonance frequency of spins in an imaged volume. Further Readings: Gao, J. H., A. W. Adamson and J. C. Gore. FIGURE M.30 Schematic design of a radiotherapy room with maze. 1992. Effect of selective excitation and phase uniformity of con- The dotted lines show the extent of the primary beam, note the thicker comitant field gradient components. Phys. Med. Biol. 37(8):1705– walls in this area. The maze should afford at least two reflections for suf- 1715; Norris, D. G. and J. Hutchinson. 1990. Concomitant ficient protection against 10 MV photons. magnetic field gradients and their effects on imaging at low mag- netic field strength. Magn. Reson. Imag. 8:33–37. entrance, which removes the need for a large heavy door. Some Mayneord F factor kind of interlocked barrier is still recommended to prevent free (Radiotherapy) Mayneord F factor is an equation derived based access to the room, for example a waist high gate. This design of on inverse square law to correct for the change in percentage room is considered more patient friendly and enables faster access depth dose values as a result of changes in focus to skin (FSD) to the patient and room in an emergency. However the maze must distance in radiotherapy treatment set-up. be big enough for beds and equipment to be physically wheeled For instance, the per cent depth dose of a radiotherapy beam round. at depth d in a patient is %DD1, if the focus to skin distance of the The provision of a maze may be difficult where an existing set-up changes from FSD1 to FSD2, the per cent depth dose value room is upgraded to a higher energy unit, and a protective door (%DD2) of the new set-up can be determined using the following may be the solution. Alternatively, if there is a pre-existing maze equation: that needs to be improved for a higher energy unit, then there are a number of measures that can be incorporated: 2 F = {(FSD2 + dm ) / (FSD1 + dm )} • Extra baffles and lintels can be used to effectively increase the length of the maze. ´{( 2 FSD1 + d ) / (FSD2 + d )} • A 5% Boron or Lithium loaded polyethene lining can be used to reduce the energy of fast neutrons. where F is the Mayneord F factor. • A light interlocked door can be placed at the inner maze The equation only takes into account the changes due to entrance, but used only for the higher energy treatments. inverse square law effect. It does not account for any changes in scattering dose. For this reason, F factor works well for small MDCT (Multidetector CT) fields. The factor has been used in manual treatment planning for (Diagnostic Radiology) See Multidetector CT (MDCT) accommodating FSD changes of up to about 40% with acceptable error. MDR (medium dose rate) (Radiotherapy, Brachytherapy) See Medium dose rate (MDR) Maze (Radiotherapy) Radiotherapy suites are designed to protect staff Mean absorbed dose to air and the general public from primary and scattered radiation. The (Radiation Protection) This is purely a qualitative term used to entrance to the room needs special consideration, and will either define the mean absorbed dose to air in conditions when the radia- use a heavily shielded mechanical door, or a maze. Figure M.30 tion field is variable. shows a typical radiotherapy room layout including an entrance Related Article: Absorbed dose maze. A correctly designed maze will prevent primary and reduce scattered radiation reaching the room entrance, including Mean dose per cumulated activity neutrons. In general, if the maze offers at least two reflections, (Nuclear Medicine) The mean dose per cumulated activity S is then this will provide sufficient protection for 10 MV photons. In used to calculate the radiation dose received by a target organ rk most cases a maze eliminates the need for more shielding at the from activity located within a source organ rh. The procedure is Mean electron energy 597 Mean free path often complicated and time consuming, especially when there is æ a large number of source organs and multiple emission types to d ö Ed = E0 çç1 - ÷÷ consider. The quantity S has the unit ‘Gy/Bq s’. S is calculated for è Rp ø a number of source–target pairs and for several radionuclides. S where is determined by the amount of energy emitted by the cumulated E0 is the average incident electrons energy at the phantom activity in the source organ (Δ) and the anatomic relationship surface M between the two organs and the target organ composition (Φ). The d is the depth of measurement mean dose per unit cumulated activity is Rp is the practical range in the same units S (rk ¬ rh ) 1 = åfi (rk ¬ rh )D Further Reading: Harder, D. 1966. Spectra of primary and i mk i secondary electrons in material irradiated by fast electrons. IAEA (M.4) S ( Technical Report 58, Vienna, Austria. rk ¬ rh ) = åFi (rk ¬ rh )Di i Mean energy imparted (Radiation Protection) Absorbed dose is a macroscopic quantity Given S and the cumulated activity à the total radiation dose D to defined as the quotient of dE/dm, energy imparted per unit vol- a target organ from a specific source organ is ume of absorber. D (r r A k ¬ h ) = ´ S (rk ¬ rh ) (M.5) At a microscopic level energy absorption is a stochastic pro- cess. The stochastic quantity is z, the specific energy imparted, Related Articles: Cumulated activity, Equilibrium absorbed where dose constant, Absorbed fraction, MIRD formalism z = e/m Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, where ε is the energy imparted in an absorber of mass m, in a Philadelphia, PA, p. 414. specific event. The formal definition of absorbed dose is Mean electron energy D = de /dm (Radiotherapy) The energy spectrum of an electron beam depends on the spread in energy of the intrinsic linac beam and the modi- where d ε– is the mean energy imparted in joules (J). fication by the energy loss and scatter in any material through Related Articles: Absorbed dose, Specific energy imparted which the beam passes. In Figure M.31, the typical distribution of Further Reading: ICRU (International Commission on the electron beam energy is shown before leaving the accelerating Radiation Units and Measurements). 1998. Fundamental structure of the linac [graph (a)]; at the phantom surface [graph quantities and units for ionizing radiation. ICRU Report 60, (0)]; and at a depth z in the phantom [graph (z)]. Due to the com- Bethesda, MD. plexity of the spectrum there is no single parameter which can characterise the electron beam; therefore several parameters are defined such as the maximum energy Emax, the modal energy E Mean free path p and the mean energy E–. (Radiation Protection) As a beam of photons passes through an The mean electron energy as a

function of depth in the phan- absorber, it is attenuated, due to collisions with the molecules tom is needed in some dosimetric protocols for the choice of cor- within the material. The number of photons in the beam decreases rection parameters for absorbed dose measurements with an ion according to the relationship chamber measurement. The Harder’s relation is employed to esti- n( x) = n(0)e-x / l mate the mean electron energy: (M.6) φE/φE(Ep) (φE)z/φE(Ep,z) (φE)0/φE(Ep,0) (φE)a/φE(Ep,a) 1.0 (z) (0) (a) r 0.5 z r0 ra E(MeV) 0 Ez Ep,z Emax,z E0 Ep,0 Emax,0 E0 Ep,a Emax,a FIGURE M.31 Electron beam parameters at three different positions: (z) at a depth z in the phantom (0) at the phantom surface and (a) before the beam leaves the accelerating structure of the linac. Mean glandular dose 598 Mean transit time (MTT) where Related Article: Air kerma l is the mean free path Further Reading: IPEM (Institute of Physics and Engineering n(0) is the number of photons per second in the beam passing in Medicine). 2005. The commissioning and routine testing of x = 0 mammographic x-ray systems. IPEM Report 89, York, UK. n(x) is the number of photons that have travelled an additional M distance, x, without being scattered out of the beam Mean life (Nuclear Medicine) See Average lifetime of atoms Mean free path may also be calculated using the following equation: Mean target absorbed dose (Radiotherapy) The International Commission on Radiation 1 l » M 7 Units and Measurements (ICRU) recommends a common method np 2 ( . ) r of dose specification, which could be generally adopted to permit where a comparison between the treatment practices. Normally a non- l is the mean free path uniform dose distribution is obtained in the target volume and n is the number of molecules therefore for practical reasons it is useful to report specific doses r is the collision radius of the molecule such as the mean target absorbed dose. For its calculation, it is necessary to calculate the dose in a If the absorber is a gas at atmospheric pressure, the mean free large number of discrete points (lattice points), uniformly distrib- path is equal to several hundred atomic diameters. uted in the target area. The mean target absorbed dose is then Related Articles: Absorber(s), Mean free path for photons in calculated as the mean of the absorbed dose values at these points. attenuation Mathematically, Further Reading: Shankland, R. S. 1955. Atomic and Nuclear Physics, The Macmillan Company, New York. 1 åDi, j N AT Mean glandular dose (Radiation Protection) The dose to which patients are exposed where during mammography procedures is normally defined in terms N is the number of the points in the matrix of mean glandular dose (MGD) measured in mGy, rather than in Di,j is the dose at lattice point i,j located inside the target area terms of effective dose. Breast tissue is relatively highly radiosen- (AT) sitive, and patients may be healthy and subject to exposure as part of a screening program. Therefore, doses during mammography Related Articles: Maximum target absorbed dose, Minimum must be optimised so that they are as low as reasonably practi- target absorbed dose, Median target absorbed dose, Modal target cable to minimise the risk of cancer in the exposed population, absorbed dose, Hot spot whilst ensuring that the resultant images allow early diagnosis Further Reading: ICRU (International Commission on of possible breast cancers such that overall prognosis for those Radiation Units and Measurements). 1978. Dose specification patients is greatly improved. for reporting external beam therapy with photons and electrons. Calculation of MGD is described in IPEM Report 89 and uses ICRU Report 29, Washington, DC. a standard breast model based on the work of Hammerstein et al. (1979) and Dance et al. (1990, 2000a). A 4.5-cm thick Perspex Mean transit time (MTT) phantom of semi-circular cross-section in the horizontal plane of (Magnetic Resonance) The mean transit time (MTT) is the aver- diameter 16 cm is used, and the incident air kerma measured. age time it takes for a non-diffusible tracer (representing the The measured air-kerma is related to the MGD for a standard blood) to pass through a microvascular or capillary system from breast using conversion factors, applied as shown in Equation the arterial to the venous side. In normal brain tissue MTT is on M.7. The entrance air kerma for a 4.5-cm thick Perspex phantom the order of 4–6 s. The function h(t) describes the distribution is equivalent to that for a 5.3-cm thick breast with a glandularity of times required by the different tracer molecules when pass- of 29% in the central region. The model also has 0.5 cm thick ing through the system (following an instantaneous tracer input). adipose layers at the top and bottom and has been found to be MTT can be calculated using the central volume theorem: For a typical for breasts of this compressed thickness for patients aged system with blood flow F, the blood volume V is the sum of all 50–64 years: paths taken by the tracer within the microvascular system. If the product h(t)dt is the tracer fraction that leaves the system between D = K45 × g53 × c53 × s (M.8) time t and t + dt, and this tracer fraction has passed through a volume Ft, the total blood volume V is given by the summation where of all Ft volumes weighted by the tracer fraction h(t)dt at time t: D is the mean glandular dose K45 is the entrance kerma for 4.5-cm thickness of Perspex ¥ ¥ g53 is the g-factor for the 5.3-cm thick standard breast V = ò(Ft )h(t )dt = Fòth(t )dt c53 is the conversion factor to allow for the glandularity of the 0 0 5.3-cm thick standard breast s is the spectral correction factor The MTT is given by The g and c factors are dependant upon the half value layer of ¥ the spectra used and can be estimated using the tables shown in MTT = ò ( ) V th t dt = IPEM Report 89. F 0 Measuring chamber 599 Mechanical index (MI) It may also be useful to remember that the tissue residue function Further Reading: Knoll, G. F. 2000. Radiation Detection and R(t), that is the fraction of tracer that remains in the tissue at a Measurement, 3rd edn., John Wiley & Sons, Inc., New York, pp. given time t following an instantaneous bolus, is given by 140–144. t Mebrofenin R (t ) = 1 - òh(t )dt (Nuclear Medicine) Mebrofenin is a diagnostic radiopharma- M 0 ceutical used to assess liver hepatocytic function. Mebrofenin is which implies that dR(t)/dt = −h(t). By using an integration by attached to Technetium-99m for HIDA scintigraphy. Technetium- parts of the function 1 · R(t) the following relationship is obtained: 99m-HIDA radiopharmaceuticals are extracted from the blood by the hepatocytes in the liver. Imaging of HIDA compounds is, ¥ ¥ therefore, used to show the distribution of functioning hepato- ò1 × R (t )dt = lim étR (t )ù + t × h(t )dt cytes and to assess the formation and secretion of bile and subse- t®¥ ë û ò quent passage to the gallbladder and the small intestines. 0 0 Indications for Mebrofenin imaging vary but it is often used Since the volume must be finite, it is evident that [tR(t)] → 0 when for assessment of biliary pain, congenital anomalies of the bili- t → ∞ [1]. Hence, MTT can be calculated as ary system, evaluation of liver transplantation or imaging of post- operative patients to exclude, for example, bile leakage. ¥ Related Articles: Positron emission tomography, Perfusion MTT = òR (t )dt imaging, Radionuclide generators 0 Further Reading: Sharp, Gemmell and Murray. 2005. Related Articles: Cerebral blood flow, Cerebral blood volume, Practical Nuclear Medicine, 3rd edn., Springer. Perfusion imaging Further Reading: Meier, P. and K. L. Zierler. 1956. On the Mechanical index (MI) theory of the indicator-dilution method for measurement of blood (Ultrasound) The mechanical index is an indicator of the rela- flow and volume. J. Appl. Physiol. 6:731–744. tive risk of mechanical bioeffects resulting from diagnostic ultrasound. Measuring chamber Bioeffects from ultrasound can be broadly divided between (Radiation Protection) The measuring chamber is an air-filled thermal and mechanical effects. Possible mechanical effects of ionisation chamber (Figure M.32) used for the measurement of ultrasound include streaming of fluid and cavitation. Cavitation gamma radiation exposure. The ionisation charge is propor- is known to occur at outputs much higher than diagnostic ultra- tional to the exposure and the ionisation current to the exposure sound, for example in lithotripsy. Although there is no evidence rate. of adverse mechanical effects occurring in clinical practice using It is based on the Bragg–Gray principle: diagnostic ultrasound, there is concern that there should be limits on ultrasound output to ensure safe practice. Dm = W ´ Sm ´ P (Gy) The output display standards (ODS) were put forward by the American Institute of Ultrasound in Medicine (AIUM) and the where National Electrical Manufacturers Association (NEMA) in 1992 D as a guide for users to monitor the output level, and by association, m is absorbed dose in a given material W is equal to the average energy loss per ion pair in the gas relative risk of ultrasound scanners. S The risk of adverse mechanical bioeffects increases with m is a relative mass stopping power of the material to that of the gas increased rarefactional pressure (also known as negative pressure) P is a number of ion pairs formed in the gas per unit mass and decreases as ultrasound frequency increases. The mechanical index is a non-dimensional index defined as A gas-filled ionisation chamber can be used for the indirect measurement of the absorbed dose in an arbitrary medium. P 1 2 r.3 ( f / awf ) MI = Related Articles: Extrapolation ionisation chamber, Ionisation CMI chamber where Pr.3 is the peak derated rarefaction pressure using 0.3 dB/cm MHz (derated for energy absorption by tissue) Collimator fawf is the ultrasound working frequency CMI = 1 MPa MHz −1/2 where CMI is a correction factor for the X-ray model tissue target being considered – * The upper limits set by the Food and Drug Administration + (FDA) in their track-3 guidance are MI = 0.23 for ophthalmic applications MI = 1.9 for all other applications R MI and TI (thermal index) are generally displayed on the ultra- sound screen as an aid to the user. MI has been found to be useful Signal in setting levels of output for contrast agent use. Low MI contrast agents can be used with low outputs to enable continuous scan- FIGURE M.32 Scheme of an air-filled ionisation chamber. ning to highlight flow in small arteries. High MIs can be used to Mechanical interlock 600 Medical device disrupt the contrast agent bubbles enabling techniques to measure volume re-filling using interval scanning. Further Readings: Abbott, J. G. 1999. Rationale and deriva- tion of MI and TI – A review. Ultrasound Med. Biol. 25:431–441; Ter Haar, G. and F. A. Duck. 2000. The Safe Use of Ultrasound in M Medical Diagnosis, BMUS/BIR Publications, London, UK. Mechanical interlock (Radiation Protection) See Interlock Mechanical isocentre (Radiotherapy) The mechanical isocentre is the point about which the various moveable parts (gantry, collimator, treatment couch) of the treatment machine rotate. While the isocentre is ideally a point, in reality it is typically a small sphere of radius less than 1 mm within which the various axes of rotation intersect. FIGURE M.33 Diagrammatic representation of a single element trans- To verify the mechanical isocentre, a front pointer must first ducer which is rotated to provide a swept area which is used to produce be inserted into the head of the linear accelerator. The collimator a sector scan. is then rotated and the walkout of the front pointer observed com- pared to a piece of graph paper attached to the treatment couch. Medial Typically it should be possible to achieve a radius less than 0.5 (General; Clinical) Directional anatomical terms describe the mm. The walkout of the front pointer should also be checked relationship of structures relative to other structures or locations as the couch is rotated, again typically on the order of 0.5 mm. in the body.

‘Medial’ means towards the mid-line, middle, away Finally, the position of the tip of the front pointer with respect to from the side (for example, the middle toe is located at the medial a fixed point should then be checked at the four main orthogonal side of the foot). gantry angles, and the deviation should be less than 1 mm. See Anatomical relationships Related Articles: Isocentre, Radiation isocentre Median target absorbed dose Mechanical locking system (Radiotherapy) The International Commission on Radiation (General) A mechanical locking system is any system which lim- Units and Measurements (ICRU) recommends a common method its or prevents access to some physical space such as a room or the of dose specification which could be generally adopted to permit inside of an equipment cabinet. Such a lock may also be fitted to a comparison between the treatment practices. Normally a non- an electrical power switch to prevent unauthorised use. uniform dose distribution is obtained in the target volume and This may be anything from a simple pin driven through a cog therefore for practical reasons it is useful to report specific doses locking it onto its shaft, to a complex electro-mechanical safety such as the median target absorbed dose. interlock system to prevent physical access to a radiotherapy treat- The median target absorbed dose is the central value among ment bunker during irradiation. the set values of the absorbed dose at all the lattice point in the A practical lock needs at least one stable state – the locked target area when arranged according to magnitude. position requiring significant mechanical force, tool, or code to Related Articles: Mean target absorbed dose, Maximum tar- change it to the unlocked position. Such a lock may be activated get absorbed dose, Minimum target absorbed dose, Modal target by chance or by default, and where this might pose a problem a absorbed dose, Hot spots ‘deadlock’ having two stable states is preferable. Further Reading: ICRU (International Commission on Radiation Units and Measurements). 1978. Dose specification Mechanical transducer for reporting external beam therapy with photons and electrons. (Ultrasound) The term mechanical transducer is used to ICRU Report 29, Washington, DC. describe a transducer where movement of the transducer ele- ments is used to form the image by ‘sweeping’ through a plane Medical device or volume. Most conventional commercial systems use array (General) A medical device is any kind of device, system or soft- transducers where the elements are fixed and beams are steered ware, which is intended by its manufacturer to be used, alone or electronically. Mechanical transducers are still used for the in combination with others, for the diagnosis, monitoring, pre- following: vention, treatment and rehabilitation of any disease or condition affecting human health, or intended to restore or support a physi- • 2D B-mode imaging. The transducer is mechanically ological process, or replace an organ, or intended to alleviate the swept around an axis and a sector image provided from end of life, or for birth control. Therefore, medical devices include a single element (Figure M.33). mathematical models or algorithms intended by its designers to • 3D imaging. A curvilinear array is mechanically swept be used for one of the intents mentioned above. It is worth making in the elevation plane providing a series of images from this remark, as still today many scholars associate the concept of a swept volume. medical device to the concept of electric medical equipment. • High frequency (>20 MHz) imaging. Difficulties in Another common belief is that only objects touching the constructing small arrays have led researchers to use patients are defined as medical devices. In fact, devices that do single element transducers moved in a line or plane to not directly come in contact with the human body, but only with provide a plane or volume of ultrasound data. its derivatives (e.g. blood), for the purposes of analysis (e.g. a Medical equipment management 601 Medical equipment management centrifuge) or for the purpose of alteration before re-implantation procurement, donations, inventory, installation and maintenance or administration (e.g. a hemodialyser) are also considered medi- of medical equipment, training for safe use and finally decom- cal devices. Medical devices are defined and regulated by many missioning). The proper use and management of such equipment, national and international bodies across the world. Those regula- in fact, improves and fosters its safety, efficiency, effectiveness tions are often inspired by the USA Food and Drug Administration and appropriateness. HTM begins by screening of the needs of (FDA) and by the European Commission, which has recently the country and ends with the decommissioning of the medi- M published the newest Medical Device Regulation 2017/745 (to be cal device. HTM is usually performed by clinical engineers and enforced in May 2020) and the In-vitro Diagnostic Regulation entails different tasks. These include: 2017/746 (to be enforced in May 2023) across the world. Medical devices can be further divided into: • Following good procurement practices: This entails an adequate process of selecting equipment for healthcare • Invasive: Medical devices that are meant to penetrate facilities, including needs screening, cost-effectiveness any natural body orifice (e.g. mouth) or any artificially analysis and technical evaluation. created body orifice (e.g. stoma). • Following appropriate solicitation and provision of • Non-invasive: Medical devices that are meant to work donations. on intact skin (e.g. thermometer). • Installation of medical devices: This involves following • Active: Medical devices that are using power sources installation manuals of new equipment in healthcare (e.g. battery) external to the human body and excluding facilities. gravity. • Medical equipment management systems: This involves • Non-active: Medical devices that do not rely on exter- developing a complete management (documentation) of nal power sources (except for gravity and the energy the medical equipment of the hospital. Records usually directly generated by the human body). include serial number, inventory number, manufacturer, • Implantable: Medical devices that are meant to be device group and type, operating status, service history, implanted and remain in the human body for a certain warranty, class of the device, the availability of user, period of time (transient, short-term, long-term). technical, maintenance and installation manuals. • Non-implantable: Medical devices that are not intended • Training of the personnel: Training for proper use of to be implanted and remain in the human body after a medical devices, as per instructions of the manufac- surgical procedure. turer, reduces incidents, fosters patient safety and may increase the performance and productivity of the Medical devices are classified according to their risks into four department. classes, based on their invasiveness and duration of use – i.e. • Maintenance and servicing of equipment: The mainte- Transient (less than 60 minutes), Short (60 minutes to 30 days), nance programme is paramount to the correct, efficient Long (more than 30 days). and safe use of medical devices. There are three kinds The EU classes are hereby presented in order of increasing of maintenance: Inspection, preventive and corrective risk, followed by an example: maintenance (new equipment also features predictive maintenance). • Class I – thermometer • Quality assurance: This entails planning and perform- • Class IIa (Class II) – syringe ing quality control and safety tests to favour the correct • Class IIb (Class II) – condom and safe operationalisation of the medical device as per • Class III (Class III) – pacemaker the instructions of the manufacturer. • Accident investigation: An analysis carried out to deter- Accordingly, in-vitro diagnostics can be divided into: mine the causes of an accident (i.e. root cause analysis). • Risk management: A series of provisions to prevent, • Class A (Class I) – microbiological culture media analyse and solve any possible risk related to the use of • Class B (Class II) – pregnancy self-tests healthcare technologies. • Class C (Class III) – genetic Tests • Equipment disposal: An evidence-based decision about • Class D – ABO typing the disposal of a medical device. This may be due to its obsolescence or to the continuous need for costly repairs or for serious safety reasons. Abbreviation: FDA = Food and Drug Administration. • Inspection or conformity assessment of medical device: Related Articles: Specification of medical device; Standards; A series of tests to evaluate the correct functioning of Life cycle of equipment; Medical equipment management; the medical devices as per the manufacturer’s instruc- Maintenance tions, their safety and performance for pre- and post- Further Reading: European Commission, Medical Device market surveillance purposes. Regulations 2017/745, In-Vitro Diagnostics Regulations 2017/746, FDA Regulation of Medical Devices. Definition of Medical Device by IFMBE CED. Abbreviation: HTM = Health Technology Management. Hyperlink: https://ced .ifmbe .org/ Related Articles: Medical device, Standards, Life cycle of equipment, Medical equipment management, Maintenance, Clinical engineering Medical equipment management Further Readings: Badnjević, A., M. Cifrek, R. Magjarević (General) Also referred to as health technology management and Z. Džemić (eds.). 2018. Inspection of Medical Devices, Series (HTM), it defines the processes of dealing with any medi- in Biomedical Engineering, Springer, Singapore; Medenou, D. cal equipment (e.g. planning, needs assessment, selection, et al. 2019. Medical devices in Sub-Saharan Africa: Optimal Medical exposure 602 Medical Internal Radiation Dose (MIRD) formalism assistance via a computerized maintenance management sys- system where a change in luminance can be identified (termed the tem (CMMS) in Benin. Health Technol. 9(39):219–232; WHO Just Noticeable Difference (JND) index). This ensures consistent Health technology management resources; Medical Equipment and accurate interpretation of images regardless of image display Management – Wikipedia; Roberto, M., E. Iadanza and F. Dori. device. 2015. Clinical Engineering: From Devices to Systems, Academic The quality assurance of medical displays is vital to ensure M Press; Willson, K., K. Ison and S. Tabakov. 2014. Medical that they are uniform, void of artefacts and, if required, follow Equipment Management, Taylor & Francis. the DICOM GSDF. This is completed with ‘test patterns’, images that, if displayed correctly, will produce, for example, a certain Medical exposure luminance or uniformity. The test objects are assessed qualita- (Radiation Protection) A medical exposure is the deliberate direct tively by eye, or quantitatively using a light meter to ensure the exposure of a human to ionising radiation, from either a radioac- display meets the required standard. tive substance or electrically produced radiation equipment, for Related Articles: DICOM (Digital imaging and communi- the purposes of diagnosing or treating a clinical condition. It cations in medicine), PACS (Picture archiving and communica- may also include those exposures of individuals carried out for tion systems), HSL (Hue, Saturation, Luminance), Bit, Bit depth, research on new or improved medical procedures or medicinal Image display, Voxel, Pixel, Grey levels, Display Quality Control, products. Grayscale Standard Display Function, Human Visual Response All other exposures, including the deliberate exposure of Function individuals for non-medical purposes such as occupational or Further Readings: AAPM Task Group 270, Display Quality immigration screening, for medico-legal proceedings, etc., are Assurance, American Association of Physicists in Medicine considered not to be ‘medical exposures’ for the purposes of (2019); Kagadis, G. C. 2013. Medical imaging displays and their regulation. use in image interpretation. RadioGraphics 33:275–290; National Related Articles: Occupational exposure, Public exposure Electrical Manufacturers Association. 2011. Digital Imaging and Communications in Medicine (DICOM) Part 14: Grayscale Medical image display Standard Display Function, NEMA; The Royal College of (General) The accuracy of a medical diagnosis is reliant on the Radiologists. 2019. Picture Archiving and Communication correct transfer, storage and display of data acquired on imag- Systems (PACS) and Guidelines on Diagnostic Display Devices, ing modalities. The international standard of DICOM (Digital 3rd edn. Imaging and Communications in Medicine) ensures standardisa- tion of such processes. Medical Internal Radiation Dose (MIRD) formalism Medical imaging devices will transform a signal (e.g. the num- (Radiotherapy) Medical Internal Radiation Dosimetry (MIRD) ber of x-ray photons hitting a pixel element within a detector) to task group of the Society of Nuclear Medicine has developed an electrical signal. These electrical signals will be processed mathematical formalism for internal dose calculations in nuclear (with images reconstructed if required) to form, depending on medicine (Bolch et al., 2009; Loeavinger and Berman, 1976). modality, a 2D or 3D image of pixel values (or voxel values in 3D This formalism computes the internal absorbed dose to a speci- imaging). These pixel values can then be displayed on an image fied target organ from a radiopharmaceutical that accumulates display device. within a set

of source organs (please note that a target organ can For the majority of medical images, pixel values will be dis- also be a source organ), see Figure M.34. The radiopharmaceuti- played as shades of grey from white to black. The number of cal is first taken up by the source organ and then its concentration shades of grey between white and black (i.e. the number of differ- (and therefore activity) is gradually diminished due to its phar- ent pixel values) that can be displayed in an image is termed its bit macokinetics and physical radioactive decay. In order to calculate depth. An image with a bit depth of 12 will be able to display 212 = the absorbed dose to the target organ, the activity of the source 4096 levels of grey. The bit depth of an image may be greater than organs must be calculated/estimated first, as a function of time. the range of grey levels that can be distinguished by the human This enables the energy, carried away by particles and/or photons eye. An image may hence be ‘windowed’ to display a small range emitted during the decay process, to be determined. Finally, the of pixel values across the full grey level display range of an image fraction of this energy absorbed in the target organ is calculated. display device. Simplified MIRD formalism: Images may be displayed on an imaging modality itself but will Time-Integrated Activity: As mentioned above, elimination of also be transferred via a Picture Archiving and Communication a radiopharmaceutical is due to biological and physical processes. System (PACS) to be reported on a specifically designed moni- As such the effective half-life, T1/2eff , of the radiopharmaceutical tor. The characteristics of the monitor will depend on its required in an organ is given by function. Characteristics include resolution, quoted as the number of pixels in a display in units of megapixels (1 megapixel = 1 × 1 1 1 = + (M.1) 106 pixels), and luminance, quoted in candelas per meter squared T1/2 eff T1/2 Physical T1/2 Biological (cd/m2). A human observer will most commonly study medical images. where T1/2 Physical is the decay half-life of the particular radioiso- This being the case, differences in pixel intensity should be tope used, and T1/2 Biological is the half-life related to biological pro- directly proportional to brightness differences as perceived by cesses/pharmacokinetics. As such, the probability of the isotope the human eye. This is naturally not the case; the human eye is decay in an organ is given by its effective decay constant as: more proficient at identifying differences in luminance in brighter ln images than in darker images. The relationship between image leff = (2) T (M.2) pixel value and displayed luminance hence must be calibrated to eff match this behaviour. The mechanism by which this completed Let’s assume that Ao amount of isotope activity is administered to is the DICOM Grayscale Standard Display Function (GSDF), the body. If there is no significant characteristic uptake time for which relates displayed luminance to points in the human visual the radiopharmaceutical following the administration, then the Medical Internal Radiation Dose (MIRD) formalism 603 M edical Internal Radiation Dose (MIRD) formalism M FIGURE M.34 Schematic example of source and target organs. Source organs are organs that accumulate significant amounts of a radiopharmaceuti- cal. Target organ is an organ intended to receive radiation dose for imaging or therapy purposes. Adapted from Loevinger et al. initial activity in a source organ, rs, is Ai (t = 0). In general, Ai = fS i, where i = 1, …, k, then the mean energy emitted per radioactive Ao, where fS (<1) is the fraction of the total administered activity transition, Δ, is calculated as: taken up by the source organ. k The activity of the radiopharmaceutical in the source organ, rs, k as a function of time, can be then expressed as: D = å EiYi = D ) i å i (M.7 i A(rs ,t ) = Ai (0)exp(-leff t ) (M.3) where Δi is the radiation energy emitted by the ith emission (in J or Gy . kg). The time integrated (or cumulated) activity, A , (total number of Absorbed Fraction: Only a portion of the energy Δ, emitted radioactive disintegrations) originating within the source organ per disintegration by a source, will be deposited in the target up to time t = TD from time of administration t = 0, is: organ of mass, M. This is known as the absorbed fraction, ϕ. The absorbed fraction depends on the energy and type of the emit- TD ®¥ ted radiation, shape, size, mass and composition of source and A (rs ,TD ® ¥) = ò A(t )dt target organs, anatomic relationship (i.e. distances) between the 0 source and target organs. When the absorbed energy is divided A by the mass, M, of the target organ, this is defined as the specific = i @ 1.443Teff Ai (0) (M.4) absorbed fraction: leff f = 1.443Teff fS Ao éëBq.sù F = (M.8) û M The integration period often used is infinity. Ideally, it should be Mean Absorbed Dose: Finally, the total energy emitted by the source matched to the biological endpoint studied in combination with the organ and absorbed in the target organ of mass, M, is given by: time period in which the relevant absorbed dose is delivered (TD). An associated quantity, time integrated activity coeffi- A ” f D = = A ” ¦ éGyù ( .9) M ë û M cient, a (S ), in the source organ, rs, is defined as: where Φ is the specific absorbed fraction. A a (rs ,Td ) = é A ësùû (M.5) For multiple radioactive decay channels equation 9 changes to: o k Total Radiation Energy Emitted: Once the time integrated activ- A å EiY k ifi ity is calculated, the total amount of radiated energy as a result of D = i = A åEiYi (fi / M) the decay processes can be estimated. This depends on the num- M i (M.10) ber of disintegrations, number of primary and secondary particles k (including photons) emitted and energy carried away by these = A åDiFi éëGyù particles. A useful term to use is the mean energy emitted per û i radioactive transition (or per unit of time integrated activity) in the source, Δ, defined as: where Φi is the specific absorbed fraction for the ith decay channel. k S-Factor: Looking at the equation 10 above, the factor å DiFi D = EY éëGy.kg/Bq.sùû (M.6) i has in fact the meaning of the energy absorbed in the target organ, where E is the particle energy in Joules, (remember Gy is J/kg), rt, per decay or per unit time-integrated activity, A in a source and Y is the number of particles per nuclear decay in (Bq.s)–1. organ, rs. Correspondingly, in the MIRD formalism, the S-factor Appropriate energy conversion factors must be applied if particle or the S-value is defined as: energy is used in keV/MeV units. k The factor Δ must be calculated for each decay channel of an S (rt ¬ rs ) = åDiFi éëGy/Bq.sùû or éëmGy/MBq.sùû (M.11) isotope under consideration. If there are multiple decay channels, i Medical lasers 604 Medical lasers As such, the MIRD dose relationship for a single source and tar- General Time-Dependent Formulation of MIRD: The get can be expressed as: time-dependent formulation of MIRD corrects the restrictions of the simplified formalism. It then describes the mean absorbed D (rt ) = A (rs )S (rt ¬ rs ) (M.12) dose D(rT ,TD ) to target tissue, rT, over a defined dose integration period, TD, following administration of the radiopharmaceutical M or in case of several sources: to the subject, using the following equation TD D (rt ) = åA (rs )S (rt ¬ rs ) (M.13) D (rT ,TD ) = åòA(rS ,t)S (rT ¬ rS ,t )dt (M.14) S rS 0 where D is the mean absorbed dose to the target organ, summing where A(rS,t) is the time-dependent activity of the radiopharma- contributions from all source organs. ceutical in the source tissue, rS,. S(rT ¬ rS ),t) is the radionuclide- The pharmacokinetic aspects (including uptake and elimi- specific term representing the mean absorbed dose rate to target nation of the radiopharmaceutical) are included in the time- organ rT, at time t, following administration, per unit activity pres- integrated activity, A . While the effects of physical decay of the ent in source tissue rS. radionuclide, distance between source and target, plus the organ Further Readings: Bolch, W. E., K. F. Eckerman, G. Sgouros size and its configuration on the target dose are contained in the and S. R. Thomas. 2009. MIRD pamphlet no. 21: A generalised S-factor. schema for radiopharmaceutical dosimetry—standardisation When the target organ is also the source organ, the absorbed of nomenclature. J. Nucl. Med. 50(3):477–484. doi: 10.2967/ dose is often referred to as the self-absorbed dose. This would be jnumed.108.056036; Loevinger, R. and M. Berman. 1976. the largest contribution to the target dose. A Revised Schema for Calculating the Absorbed Dose from The cross-absorbed dose is referred to as the absorbed dose, Biologically Distributed Radionuclides. MIRD Pamphlet No. 1, where the source and target organs are different. Revised ed, Society of Nuclear Medicine, New York; Loevinger, S-factors are often obtained from Monte Carlo simulations R., T. F. Budinger and E. E. Watson. 1983. MIRD Primer for using simplified anatomical models or phantoms, using analyti- Absorbed Dose Calculations, Society of Nuclear Medicine, Inc., cal, voxel-based and surface graphics approaches. New York, ISBN 0-932004-25-3. Please note that this simplified MIRD formalism assumes an instantaneous uptake of the radiopharmaceutical, a mono-expo- Medical lasers nential time-dependent removal of radionuclide from the source (Non-Ionising Radiation) The use of lasers is widespread in the as well as constant mass of the target and source organs during medical field. The table below shows some common types of the irradiation. medical laser and their application. Laser Laser Type Wavelength (nm) Characteristics CO2 Gas 10 600 Wavelength highly absorbed by intracellular water applied in removal of tissue and sealing of blood vessels. Wavelength does not penetrate deeply into tissue (c. 100–300 µm) therefore damage to deeper tissue is minimal. Delivered by a mirror system in an articulated arm as the wavelength is readily absorbed by most standard optical fibres. Argon 488; 514 The blue/green wavelengths of the Argon laser are well absorbed in haemoglobin and melanin. These wavelengths pass through water and can therefore be used in aqueous environments such as the bladder. Excimer: 193 Extremely precise photoablation of tissue with no damage to surrounding Argon Fluoride (ArF) 248 tissue. Krypton Fluoride (KrF) 308 Xenon Chloride (XeCl) 353 Xenon Fluoride (XeF) Nd:YAG Solid state 1064 (532) Nd:YAG lasers are frequency doubled to achieve a KTP:YAG beam. The (Neodymium doped ytrium 1064 nm wavelength is transmitted through water and can be used with a aluminium garnet) wide range of delivery devices. Photothermal tissue ablation is not as successful as for other lasers (CO2, KTP:YAG) and the beam energy is scattered and well absorbed by proteins in tissue so greater damage to deeper tissue is caused. KTP:YAG (Potassium- 532 Absorbed well in haemoglobin and melanin and transmitted through water. titanyl phosphate) More efficient than the Nd:YAG for photothermal coagulation and and frequency doubled vaporisation of tissue. Can be used with optical fibres. Nd:YAG Er:glass 1540 Non-ablative skin resurfacing. (Erbium) (Continued) Medical physics 605 Medical physics Ho:YAG (Holmium) 2100 Strongly absorbed by water, used for photothermal cutting and ablation of bone and cartilage. Output is pulsed and can be Q-switched for use in lithotripsy. Er:YAG 2940 Strongly absorbed by water, more suited to superficial applications and less (Erbium) common than other YAG lasers. Alexandrite 755 Well absorbed by deoxygenated haemoglobin, longer visible wavelength and M long pulse width allows deeper penetration and treatment of larger vessels. Ruby 694.3 The first laser to be demonstrated. Had wide applications in several medical areas but has now been mostly replaced by more efficient lasing media with similar wavelengths. Diode lasers Semi- 400–450 Generally compact and therefore easily transportable. Wavelengths can be conductor 600–900 transported via conventional optical fibres. 1100–1600 Pulsed dye Dye 300–1800 Visible wavelengths longer than 532 nm allow deeper penetration. 1100–1600 Further Reading: Nouri,

Keyvan. 2011. Lasers in implanted radioactive sources given the state of technology; (9) Dermatology and Medicine, Springer-Verlag, London, UK, ISBN planning, directing, conducting and participating in supporting 978-0-85729-280-3. programs and remedial procedures to ensure effective and safe use of ionising and non-ionising radiation and radio nuclides in Medical physics human beings by physician specialist; (10) formulating radiation (General) Medical physics is a profession which deals with the protection guides and procedures specific to hospital environment application of the concepts and methods of physics for the pre- and other professional groups and organisations and conducting vention, diagnosis and treatment of human diseases with a spe- specialised measurements and producing protocols to minimise cific goal of improving human health and well-being. A specific radiation exposure of patients, staff and the general public; (11) definition of the International Organization for Medical Physics participating in and contributing to the development and imple- (IOMP) gives the following description for the professional activi- mentation of national and international standards, laws and regu- ties of a medical physicist: lations relating to patient safety, particularly to radiation and Medical physicists apply knowledge and methodology of sci- radioactive materials; (12) teaching principles of medical phys- ence of physics to all aspects of medicine, to conduct research, ics to physicians, residents, graduate students, medical students, develop or improve theories and address problems related to diag- technologists, and other health-care professionals by means of nosis, treatment and rehabilitation of human disease. They are lecturers, problem solving and laboratory sessions; (13) prepar- directly involved with patients and people with disabilities. Tasks ing, publishing and presenting scientific papers and reports and include (1) conducting research into human disorders, illnesses (14) supervising and managing radiation workers and other health and disabilities and investigating biophysical techniques asso- professional workers. ciated with any branch of medicine; (2) conducting specialised Medical physicists work mainly in hospitals, but also in uni- examinations of patients and the disabled, improving patient care versities, research institutions, regulatory bodies, industries, etc. and clinical services and developing innovative imaging and non- For this reason, it is not possible to establish the exact global num- imaging diagnostic procedures for specific medical applications; ber of medical physicists. Most medical physicists are members (3) developing novel instrumentation and physiological measure- of the International Organization for Medical Physics (IOMP), ment techniques, mathematical analysis and applications of com- which currently (2020) has about 30,000 members in 87 coun- puters in medicine in response to clinical need for patients and tries. However, this workforce has uneven global distribution, aiding to everyday living for the disabled; (4) ensuring the quality, varying from about 25 specialists per million population in the safety testing and correct maintenance and operation of treatment USA, to less than one per million in Africa (see article IOMP for machines, x-ray equipment, radiation treatment planning comput- approximate numbers of medical physicists in various regions in ers; medical uses of ultrasound, MRI and infrared and the cor- the world). Due to the efforts of IUPESM, IOMP and IFMBE rect delivery of prescribed radiation doses to patients in radiation the occupation medical physicist is listed under number 2111 in therapy; (5) ensuring the accuracy of treatment unit parameters the International Standard Classification of Occupations (ISCO- and settings used for a patient’s treatment, including correct trans- 08), published by the International Labour Organisation (ILO). fer of parameters between the simulator, treatment plan and the Current surveys predict almost tripling of this workforce globally treatment unit, and periodic review of each patient’s chart; (6) by 2035. The IOMP Journal Medical Physics International pub- calculating dose distributions and machine settings, design and lishes regular information about the status of medical physics in fabrication of treatment aids and treatment-beam modifiers for various countries of the world. individual patient treatments; (7) in vivo measurement to verify Related Articles: IOMP, IUPESM the dose delivered to a patient; participation at patient discussion Further Readings: Atun, R., D. A. Jaffray, M. B. Barton, conferences; (8) advising and consulting with physicians on the F. Bray, M. Baumann, B. Vikram, T. P. Hanna, F. M. Knaul, Y. physical and radiobiological aspects of patients’ treatments, and Lievens, T. Y. M. Lui, M. Milosevic, B. O’Sullivan, D. L. Rodin, the development of treatment plans in such applications as use E. Rosenblatt, J. Van Dyk, M. L. Yap, E. Zubizarreta and M. of ionising radiation in diagnosis, therapy, treatment planning Gospodarowicz. 2015. Expanding global access to radiotherapy. with externally delivered radiation as well as use of internally Lancet Oncol. 16:1153–1186; Medical Physics World. 2008. Medical physics expert (MPE) 606 Medical Radiation Protection Education 24(2):21; Tabakov, S. 2016. Global number of medical physicists b. Treatment times approx. 5–20 min (comparable to and its growth 1965–2015. J. Med. Phys. Int. 4(2):20167, 78–82. external beam therapy) Available free from: www .m pijou rnal. org /p df /20 16 -02 /MPI- c. Clinical data available 2016- 02 -p0 78 .pd f; Tabakov, S., P. Sprawls, A. Krisanachinda 4. Pulsed dose rate, PDR and C. Lewis (eds.). 2011. Medical Physics and Engineering a. Mimics LDR, using many small ‘HDR pulses’ dur- M Education and Training – Part I, ISBN 92-95003-44-6, ICTP, ing a longer treatment time example: 1 pulse per Trieste, Italy. Available free from: www .e meral d2 .eu /mep/ e -boo hour during 24 hours, 0.5 Gy per pulse given in 5 k11 /E TC _BO OK _20 11 _eb ook _s .pdf; Tsapaki, V., S. Tabakov min; total dose 12 Gy per day and M. Rehani. 2018. Medical physics workforce: A global per- spective. Phys. Med. 55:33–39. The radiobiological effects in the tissues irradiated depend on the type of applicator used, on the fractionation scheme and on both Medical physics expert (MPE) dose and dose rate distributions. As stated in the ICRU Report (Radiation Protection) The medical physics expert is defined in 38, ‘the clinical experience accumulated with radium techniques the European Directives as: ‘an expert in radiation physics and cannot be applied to new irradiation conditions without careful technology whose training and competence to act is recognised consideration’. This includes consideration of both tumour effects by competent authorities and who, as appropriate, acts or give and effects on normal tissues. advise on patient dosimetry, on the development and use of com- Abbreviation: ICRU = International Commission on Radiation plex techniques and equipment on optimisation, on quality assur- Units and Measurements. ance, including quality control, and on other matters relating to Related Articles: Brachytherapy, Dose rates in brachytherapy, radiation protection within the scope of the Directive’. see also articles under radiobiology The Directive further states, ‘In radiotherapeutic practices, a Further Reading: ICRU (International Commission on medical physics expert shall be closely involved. In standardised Radiation Units and Measurements, Inc.). 1985. Dose and volume therapeutical nuclear medicine practices and in diagnostic nuclear specification for reporting intracavitary therapy in gynecology. medicine practices, a medical physics expert shall be available. ICRU Report 38, Washington, DC. For other radiological practices, a medical physics expert shall be involved, as appropriate, for consultation on optimisation includ- Medium frequency portable x-ray machine ing patient dosimetry and quality assurance including quality (Diagnostic Radiology) See High frequency generator control, and also to give advice on matters relating to radiation protection concerning medical exposure, as required’. Individual national legislation and guidance implementing the Medical Radiation Protection Education Directive provides more detailed information on the education, and Training project (MEDRAPET) training and experience required to be appointed as an MPE and (General) The overall aim of this EU-funded project was to the duties to be undertaken. improve the implementation and harmonisation of the Medical The fields of competence of an MPE are normally radiotherapy Exposure Directive 2013/59/EURATOM related to radiation physics, diagnostic radiological physics, nuclear medicine physics protection education and training of medical and healthcare and medical health physics; an MPE is competent and authorised professionals in the EU member states. It provided learning to practice independently in one or more of these fields. In addi- outcomes in terms of detailed knowledge-skill-competence tion to these fields of competence in some cases the MPE is also inventories for all medical and healthcare professions (includ- involved in other clinical and biomedical activities. ing specialisations) involved in the use of ionising radiation in Further Reading: European Union 1997 Council Directive medicine. The document is an essential reference for all medi- 97/43/Euratom of 30 June 1997 on Health Protection of Individuals cal physicists involved in the teaching of medical and healthcare against the Dangers of Ionizing Radiation in Relation to Medical professionals. Exposure Official journal NO. L 180, 09/07/1997 P. 0022–0027. The project was coordinated by the European Society of Radiology (John Damilakis john .damilakis @med .uo c .gr). The other project partners were: Medium dose rate (MDR) (Radiotherapy, Brachytherapy) Dose Rates in Brachytherapy: Different dose rates are • European Federation of Organisations for Medical used in brachytherapy treatment techniques. The International Physics (EFOMP) Commission on Radiation Units and Measurements, ICRU, • European Federation of Radiographer Societies (EFRS) defined these dose rates in its Report No. 38 ‘Dose and Volume • European Society for Therapeutic Radiology and Specification for Reporting Intracavitary Therapy in Gynecology’: Oncology (ESTRO) • European Association of Nuclear Medicine (EANM) 1. Low dose rate, LDR • Cardiovascular and Interventional Radiological Society a. 0.4–2.0 G/h of Europe (CIRSE) b. Traditional technique; 0.5 Gy/h, 60 Gy with treat- ment time 5 days Related Article: European Training and Education for c. Large amount of clinical data Medical Physics Experts project (EUTEMPE) d. (NOTE: Ultra low dose rate 0.01–0.3 Gy/h) Further Reading: The MEDRAPET guidance document was 2. Medium dose rate, MDR published by the European Commission as Radiation Protection a. 2–12 Gy/h Series 175. Its full title is Guidelines on Radiation Protection b. Seldom used Education and Training of Medical Professionals in the European 3. High dose rate, HDR Union. It can be found here: https :/ /ec .euro pa .eu /ener gy /si tes /e ner a. > 12 Gy/h = 0.2 Gy/min /fi les/ docum ents/ 175 .p df Metal-oxide–semiconductor field-effect 607 Metal-oxide–semiconductor field-effect MEFOMP MEP workshop topics include, but not limited to: interna- (General) The Middle East Federation of Organizations for tional collaboration, new technologies, workforce, capacity build- Medical Physics (MEFOMP) was founded in 2009 as a Regional ing through education and training. MEP workshop speakers Organization of IOMP. As of 2019, the Federation consists of 12 are among the most prominent leaders of IUPESM, IOMP and national member organisations, representing about 1000 physi- IFMBE, regional experts in the field of biomedical engineer- cists and engineers working in the field of medical physics. ing and medical physics and representatives of the partnering M Since its inauguration, the main objective of MEFOMP has organizations. been to harmonise and promote the best practice of medical phys- Hyperlink: www .iupesm .org /mep ics in the Middle East region. MEFOMP includes the medical physics societies from the fol- lowing countries: Bahrain, Iraq, Jordan, Kuwait, Lebanon, Oman, Metal artefact Palestine, Qatar, Saudi Arabia, Syria, Yemen and UAE. (Diagnostic Radiology) Metal objects in the patient, such as dental Hyperlink: www .mefomp .com fillings, prostheses, surgical clips and electrodes, can cause severe streaking artefacts in the image (especially in computed tomog- raphy). Sometimes these are called ‘star artefacts’ – Figure M.35. Mega electron volt The artefact can be caused by a combination of low signal, beam (General) See Electron volt hardening, partial volume, undersampling at sharp interfaces and limited dynamic range of the detectors and display. Meisberger polynomial Metal artefacts can be eliminated by removing of metal (Radiotherapy, Brachytherapy) objects, where possible, or avoidance of them, for example, by Point Source Calculations: Dose distributions around brachy- gantry angulation. Otherwise metal artefact reduction software therapy sources are dominated by the inverse square law behav- may be available to reduce their appearance. This software identi- iour. Different slowly varying functions have been used to describe fies projection data which has ‘over-ranged’ due to high attenua- the deviation of the dose distribution from 1/r2, and the Meisberger tion of the metal object. This data is then replaced by interpolating polynomial is one of them. This polynomial was published in data from both sides of the metal object (Figure M.36). Images 1968, and coefficients were given for the brachytherapy sources used at that time; gold, iridium, caesium, radium and cobalt. The Meisberger polynomial has been

used for a long time in treatment planning systems to characterise the radial behaviour of cylindrical brachytherapy sources. Related Articles: Source models, Point source calculation Further Reading: Venselaar, J. and J. Pérez-Calatayud. (eds.). 2004. A Practical Guide to Quality Control of Brachytherapy Equipment, ESTRO Booklet No. 8, Brussels, Belgium. Melanoma (Non-Ionising Radiation) Melanoma is the deadliest form of skin cancer, which generates from melanocytes (the skin cells responsible for pigment). Melanoma is common in areas of the skin which are most exposed to the sun (such as head or back of legs). However, it can appear also in areas of the body which are normally covered. A very promising treatment for melanoma is immunotherapy. Related Articles: AORD, Basal cell carcinoma, UV light haz- ard, UV dosimetry Further Reading: Blumenberg, Miroslav. 2018. Human FIGURE M.35 Example of streaking artefacts on a CT scan (neck) due Skin Cancers: Pathways, Mechanisms, Targets and Treatments, to presence of metal in patient (dental implant). London, UK. MEP Workshop (Medical Physics and Engineering Workshop) (General) The Medical Engineering and Physics (MEP) work- shop is organized by the International Union for Physical and Engineering Sciences in Medicine (IUPESM). The workshop is focused on experience sharing, networking and collaboration between medical physicists and biomedical engineers globally. It is hosted as part of leading medical physics and biomedical engi- neering events, including the IUPESM World Congress. Attenuation Interpolated The MEP Workshop is focused on common professional strat- profile profile egies, bringing experts on-site different regions and setting up the scene for further collaboration. The workshop is largely sup- FIGURE M.36 Illustration of software method for reduction of metal ported by IUPESM partner organizations. artefacts. Metal-oxide–semiconductor field-effect (MOSFET) 608 Metallic implant M Original image With metal artefact reduction FIGURE M.37 Effectiveness of metal artefact reduction software (CT scan – spine). showing the effectiveness of metal reduction software are shown whose central part is made of metal (steel, which is insulated and in Figure M.37. grounded). The end parts of this envelope are still made of glass Some manufacturers use an extended CT number scale to and vacuum inside is similar. reduce metal artefacts caused by over-ranging of attenuation In more recent designs the entire envelope is made of metal values. (using black ceramic insulation inside and additional ceramic Related Articles: Artefact, Beam hardening, Cone beam arte- insulators at the high voltage connections). A typical ceramic is fact, Helical artefact, Image artefact, Motion artefact, Partial vol- beryllium oxide due to its very high thermal conductivity. Often ume effect (artefact), Ring artifact the metal x-ray tubes are also called ceramic x-ray tubes. The metal tube envelopes have also better heat dissipation Metal-oxide–semiconductor field- abilities and are more robust. They can be used with anode dicks effect (MOSFET) transistor with mass on the order of 2000 g (i.e. with double heat storage (Radiotherapy) A metal-oxide–semiconductor field-effect tran- ability) what make them especially useful for angiography and CT sistor (MOSFET, MOS-FET or MOS FET) is a field-effect tran- scanners. sistor (FET with an insulated gate) where the voltage determines The metal tubes use a composite (double) exit window – made the conductivity of the device. MOSFETs are particularly useful of beryllium and aluminium (the x-rays first pass through Be, and in amplifiers due to their input impedance being nearly infinite, after this through Al). These tubes have smaller inherent filtra- which allows the amplifier to capture almost all the incoming tion as firstly Be absorbs less radiation than the glass and sec- signal. The main advantage is that they require almost no input ondly they can be placed closer to the housing, what diminishes current to control the load current, when compared with bipolar the oil absorption. The inner side of the metal envelop is specially transistors. Two basic forms of MOSFET are available: ‘depletion’ coated to absorb the infrared radiation from the heated anode. and ‘enhancement’. In the depletion type, the transistor requires At present more and more x-ray tubes use this type of envelope the Gate-Source voltage (VGS) to switch the device ‘OFF’. The (Figures M.38 and M.39). depletion mode MOSFET is equivalent to a ‘Normally Closed’ Related Articles: Glass envelope, Filament heating, X-ray switch. In the enhancement type, the transistor requires a Gate- tube, Anode, Liquid metal bearing Source voltage (VGS) to switch the device ‘ON’. The enhancement mode MOSFET is equivalent to a ‘Normally Open’ switch. Metallic implant Advantages of MOSFETs include: they have very small active (Magnetic Resonance) Magnetic field–related translational volumes; they have instantaneous readouts (for on-line dosim- attraction and torque present hazards to individuals with cer- etry); they offer permanent dose storage (can be read multiple tain implants or devices. The presence of an intracranial aneu- times), and are waterproof. Disadvantages of MOSFETs include: rysm clip made from ferromagnetic materials is contraindicated they have a finite lifetime (~100 Gy); they have an energy depen- for MR procedures because magnetically induced forces may dence and a temperature dependence. With adequate characteri- displace these clips causing serious injury to the patient or his sation, MOSFETs can be useful in making measurements within death. The risk of the clip displacement can be a minimal risk high dose gradient fields. Examples of their clinical application in other parts of the body because after some time fibrosis and include: in-vivo dosimetry, small field output factor measurement encasement of the clip help to keep it in a stable position. The (e.g. for radiosurgery) and measurement of build-up curves for pulsed radio frequency (RF) magnetic fields, which are used high energy photon beams. to obtain the MR signals from tissue induce electrical currents in conductive metal implants and determine their heating. The Metal x-ray tube magnitude of the increased heating of tissues due to the presence (Diagnostic Radiology) In order to diminish the metalisation of the metallic implant depends on the dimensions, orientation, of the glass, some manufacturers produce x-ray tube envelopes shape and location of the metallic implant in the patient. This Metastable nucleus 609 Microdosimetry Metastasis (Nuclear Medicine) Metastasis refers to cancer cells that have detached themselves from the primary tumour and colonised in another organ or tissue. The new cancer colony is referred to as a metastasis. A detached cancer cell (or cell cluster) can spread through M the body via the blood stream and the lymphatic system. If cells from a primary lung cancer spread to and metastase in the liver, they are then categorised as metastatic lung cells rather than liver cancer. MFO (Multifield optimisation) (Radiotherapy) See Multifield optimisation (MFO) Microbubbles (Ultrasound) Contrast agents for ultrasound are made up of micro-bubbles that dramatically change the acoustic properties of the medium and thereby provide a stronger signal from blood. FIGURE M.38 Mammographic metal tube (the lower part is still made More details and detection strategies are provided under Contrast of glass). agents. Micro CT (Diagnostic Radiology) See Small animal CT Microdosimetry (Radiation Protection) The deleterious effects of ionising radiation occur as stochastic events at a sub-cellular level with dimensions on the order of microns. Conventional quantities for measuring radiation dose and radiation quality are absorbed dose (D) and linear energy transfer (LET). These quantities are mea- sured over macroscopic distances. Microdosimetry, as the name suggests, is concerned with the deposition of energy at microscopic levels. It is important in developing an understanding of how and why biological effects may differ following irradiation by the same absorbed dose and radiation quality. Microdosimetric events are stochastic in nature and there are two principal microdosimetric quantities, lineal FIGURE M.39 Fully metal x-ray tube. (Image by PHILIPS.) energy (y) and specific energy (z). Lineal energy is the quotient of the energy imparted, ε, in a single event and the mean chord length, l, of the volume in which increases heating of surrounding tissues primarily concentrated the deposition occurred; thus, in a small volume and can burst. Metallic implants produce also susceptibility artefacts in magnetic resonance imaging. The artefacts are related with the metal characteristics and the ori- e y = entation of the implant in the magnetic field. The artefact can l therefore be minimised by optimally positioning patient in the magnet. lineal energy is measured in units of keV/μ and differs from LET, Related Article: Implant a non-stochastic quantity, in that it has no energy cut-off value. Specific energy (z) is the quotient of energy imparted, ε, and Metastable nucleus the mass of the volume in which the event occurred; thus, (Nuclear Medicine) A metastable nucleus is an excited nucleus ‘trapped’ in an excited state. The metastable nucleus will even- e z = tually de-excite down to its ground state. In the periodic table m nomenclature an m is placed after the atomic number to mark that the nucleus is metastable. A common metastable nucleus The mean specific energy, z–, which is the expectation value of a used in nuclear medicine is 99mTc; the metastable technetium specific energy between z and z + dz is a non-stochastic quantity nucleus has a half-life of approximately 6 h, meaning that half and equivalent to the absorbed dose, D. the excited states will have de-excited after 6 h. A 140 keV pho- Lineal energy and specific energy differ in that the former is ton suitable for emission imaging is generated in the de-excita- the deposition from a single event whereas the latter may result tion process. from more than one event. Measurements of microdosimetric quantities are made with Metastable state proportional chambers. These are gas-filled devices with compo- (Nuclear Medicine) See Metastable nucleus nents made from tissue equivalent materials. They are designed to Micro-MLC 610 Midpoint dose be Bragg–Gray cavities imitating tissue dimensions on the order Further Readings: Graham, D. T., and P. Cloke. 2003. of microns. Principles of Radiological Physics, Elsevier Science Limited, Related Articles: Absorbed dose, Linear energy transfer Edinburgh, UK, www .bipm .fr /en /si /si _brochure Micro-MLC Micro-PET M (Radiotherapy) For small fields such as those used for brain (Nuclear Medicine) A pre-clinical detector system for imaging the tumours or boost fields in the head and neck, better resolution of radionuclide distribution in small animals. The spatial resolution the field margins may be required than for larger PTVs. Several in a micro-PET system is primarily determined by the positron miniature multileaf collimators have been developed to be used physics, namely the positron path length. A conventional PET for these cases with 1.5–6 mm leaf widths projected at the linac scanner has better resolution than a conventional SPECT but for isocentre. the micro systems the relationships is reversed. The obvious medi- Micro MLCs (sometimes referred as mini MLCs) have been cal benefit with a small animal PET imaging system is that radio- typically configured as self-contained accessories that can be pharmaceuticals can be tested on animals and evaluated before attached to the collimator of a linear accelerator for specific clinical trials. treatment techniques and removed for conventional use of the Related Article: Micro-SPECT machine. Fibre-optic transmission lines are used to commu- nicate with a PC-based digital control system. The secondary Microprocessor jaws of the accelerator are set to a fixed field size during the use (General) The microprocessor is now one of the main compo- of the micro MLC so that the leaves of the micro MLC need nents of every computerised system. While it is usually a single be only long enough to cover a reduced maximum field size chip –central processing unit (CPU) – usually 16-bit, 32-bit or (Figure M.40). 64-bit design, special microprocessors are used for digital sig- Another non-conventional MLC system is the MIMiC device nal processors (DSP) and graphics processing units (GPU). These provided by NOMOS Corporation. It is designed to collimate microprocessors are main components of any imaging system. the x-ray field to a fan-beam that is dynamically modulated by Hyperlink: http: / /en. wikip edia. org /w iki /M icrop roces sor short-stroke leaves as the gantry of the accelerator is rotated. The modulated fan beam irradiates a transverse plane of the patient that Micro-SPECT is 2-cm thick. Each leaf pair has only two positions: completely (Nuclear Medicine) A pre-clinical detector system for imag- open or completely closed. This system is sometimes referred to ing the radionuclide distribution in small animals. The camera as bimodal MLC. parameters are related in a similar way as for a large scale SPECT Abbreviation: MLC = Multileaf collimator.

system. System resolution, for example is primarily determined Related Articles: Multileaf collimator, Stereotactic radiosurgery by the bore diameter of the SPECT system which gives a superior Further Readings: AAPM (American Association Physicist resolution compared to a conventional SPECT system. There is in Medicine). 2001. Basic applications of multileaf collimators. also the obvious medical benefit of trying new radiopharmaceuti- Report of Task Group No. 50, Radiation Therapy Committee, cals on animals instead of patients. A micro-SPECT can be used AAPM Report No. 72, Madison, WI; www .b rainl ab .co m /dow together with a micro-PET, CT or MR scanner in a multimodality nload /pdf/ Brain LABIG Broch ureEn glish .pdf. system. Related Article: Micro-PET Micron (General) Micron (symbol μ) is a unit of length corresponding to Micro-switch micrometre (symbol μm) in the International System of Units (SI): (General) A micro-switch is an electro-mechanical switch that needs very little force for closing or separating the electrical 1m = 1mm = 10-6 m contacts. They are often used in control and regulation circuits where a movement of the operator or a mechanical tool controls Related Article: Système Internationale (SI) a process. It is often used as a stop switch in medical electro- mechanical devices and medical imaging equipment that has movable parts. Microwaves (General) Microwaves are electromagnetic waves (i.e. energy waves propagated by paired, transversely oscillating electric and magnetic fields) and are part of the electromagnetic spectrum. The microwave part of the spectrum is characterised by wave- lengths on the order of approximately 1 mm − 1 m, frequencies of between roughly 0.3 and 300 GHz. Microwaves are in the non- ionising part of the electromagnetic spectrum. Related Article: Electromagnetic energy spectrum Midpoint dose (Radiotherapy) The midpoint dose is the dose delivered at the FIGURE M.40 m3® high-resolution multileaf collimator produced by midline of a patient treated by two photon beams directed along BrainLab (www .b rainl ab .co m /dow nload /pdf/ Brain LABIG TBroc hureE the same axis from opposite sides of the treatment volume. If the nglis h .pdf ). parallel opposed beams are equally weighted and normalised to Minification gain 611 MIRD formalism the midpoint value, the dose distribution can be made uniform for example 5% accuracy. The MDA parameter is often used in within the irradiated volume depending on the patient thickness order to set a detection limit for a detector system to assure that and the beam energy and flatness. The dose near the patient sur- the MDA is measurable while monitoring a suspected radioactive face increases compared to the midpoint dose when the patient contamination area (Bq/m2). MDA is also important in order to thickness increases or the beam energy decreases. In Figure M.41 determine the ability of nuclear medicine equipments to measure the depth dose curves for parallel opposed beam normalised to lower levels of activities, mainly liquid scintillation counters or M the midpoint dose value are reported for different beam energies. NaI(Tl)-well-counters. The parameter may also be used to deter- The so-called tissue lateral effect consists in the increase of the mine detection limits for noninvasive equipments, such as scin- surface dose relative to the midpoint dose. tillation cameras and PET scanners, keeping the administered The ratio between the maximum peripheral dose and the mid- activity low enough without impairment of the detector signal and point dose is reported in Figure M.42 for different patient thick- diagnostic accuracy. ness and beam energies. Different algorithms have been developed for the determina- tion of MDA for different equipments and situations which can be Minification gain found elsewhere. (Diagnostic Radiology) See Total brightness gain Related Articles: Quality control, Activity, Optimisation Further Reading: Knoll, G. F. 2000. Radiation Detection and Minimum detectable activity (MDA) Measurement, 3rd edn., John Wiley & Sons, Inc., New York. (Nuclear Medicine) The minimum detectable activity (MDA) (also called minimum detectable amount of activity), of a radionuclide Minimum target absorbed dose may be defined as the smallest activity in a sample that can be (Radiotherapy) The International Commission on Radiation detected with better than a certain percentage counting accuracy, Units and Measurements (ICRU) recommends a common method of dose specification which could be generally adopted to permit a comparison between the treatment practices. Normally a non- Midline uniform dose distribution is obtained in the target volume and 130 therefore for practical reasons it is useful to report specific doses 60Co 120 4 MV as the minimum target absorbed dose. 10 MV The minimum target absorbed dose is the lowest absorbed 110 dose in the target area. No area limit is recommended when 100 reporting minimum target absorbed dose. 90 25 MV Related Articles: Mean target absorbed dose, Maximum tar- get absorbed dose, Modal target absorbed dose, Median target 80 absorbed dose, Hot spots Further Reading: ICRU (International Commission on 70 Radiation Units and Measurements). 1978. Dose specification 60 for reporting external beam therapy with photons and electrons. ICRU Report 29, Washington, DC. 50 40 MIP (maximum intensity projection) 0 5 10 15 20 25 (Magnetic Resonance) See Maximum (minimum) intensity -pro- Depth (cm) jection (MIP) MIRD Committee FIGURE M.41 Depth dose curves for parallel opposite beams for dif- (Nuclear Medicine) The Medical Internal Radiation Dose (MIRD) ferent beam energies. Committee is a committee sanctioned by the Society of Nuclear Medicine (SNM) that provides fundamental quantities used for risk assessment, radiation dosimetry/protection and radiation ther- apy. The MIRD Committee develops methods, models, assump- 1.4 tions and mathematical models for calculation of the internal radiation dose. 1.3 MIRD formalism 60Co (Nuclear Medicine) The MIRD method is used to calculate the total absorbed dose to patients from nuclear medicine examina- 4 MV 1.2 tions. The method is named after the Medical Internal Radiation Dose Committee of the Society of Nuclear Medicine. 10 MV With the mean absorbed dose per cumulated activity S, it 1.1 is possible to calculate the radiation dose to target organs from 25 MV radioactivity accumulated in one or several source organs. The source and the target organ can be the same organ; in fact often 1.0 the major part of the absorbed dose originates from radioactivity 10 15 20 25 30 within the target organ itself, often referred to as self dose. Other Patient thickness (cm) organs are considered to be source organs if they contain a con- centration of radioactivity higher than the average concentration FIGURE M.42 Ratio of the maximum peripheral dose to the midpoint dose. of radioactivity in the body. Percentage depth dose Maximum peripheral dose ratio midpoint dose Mirror image artefact 612 Misadministration The radiation dose to a target organ is determined in three steps: Further Readings: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, 1. Cumulated activity, Ã: The first step is to determine the Philadelphia, PA, pp. 406–417; Loevinger, R. F., T. Budinger and E. kinetics of the tracer, that is basically the time spent by E. Watson. 1988. MIRD Primer for Absorbed Dose Calculations, the radioactivity in a specific organ and also the amount The Society of Nuclear Medicine Inc., New York; Snyder, W., M. M of activity present in each source organ. This is usu- Ford and G. Warner. 1978. Estimates of specific absorbed fraction ally done with a gamma camera, but can involve other for photon sources uniformly distributed in various organs of a methods such as blood, urine and faeces samples. The heterogeneous phantom. MIRD Pamphlet No. 5 (revised), Society activity concentration in an organ will show a temporal of Nuclear Medicine, New York; Stabin, M. G., R. B. Sparks and variance, that is from initial uptake to total clearance, E. Crowe. 2005. OLINDA/EXM: The second-generation personal and the behaviour is modelled by a time-activity func- computer software for internal dose assessment in nuclear medi- tion which differs between organs. The cumulated activ- cine. J. Nucl. Med. 46(6):1023–1027. ity is the integration of the time-activity function, i.e. the total numbers of disintegrations in the source organ. 2. Equilibrium absorbed dose constant, ‘Δ Mirror image artefact ’: Secondly the (Ultrasound) A mirror image artefact (sometimes called a ghost total amount of emitted energy from the radioactivity in signal) occurs due to specular reflection at a large smooth inter- the source organ is calculated. Decisive parameters are face (‘large’ referring to an interface surface greater than the the energy of the emitted radiation and radiation type. beam width). A mirror artefact is most pronounced with highly 3. Absorbed fraction, ‘ϕ’: The fraction of the emitted reflective interface surfaces, which are boundaries with a high energy source organ absorbed in the target organ is amplitude reflection coefficient. determined. The dose contribution depends on the ana- The artefact occurs when the pulse on its return path (after it tomic relationship between the target and source organ has been reflected back from a large highly reflective interface) (size, shape and distance between the two), radionuclide is interrupted by another object which causes the pulse signal to characteristics (energy, particle or non-particle radia- be reflected (or scattered) back in the outgoing direction where tion, etc.) and the material composition of the tissue. it is again reflected off the large interface before being returned To calculate the radiation dose, ‘Δ’ and ‘ϕ’ must be calculated uninterrupted back to the transducer. for each emission, i. The radiation dose procedure is somewhat The ultrasound system measures time delays between trans- tedious, especially when dealing with different emission types mitted and received signals. The axial depth position on the and many source target pairs. To simplify the calculation proce- screen image is the product of an assumed propagation speed (speed of sound in tissue, commonly estimated at 1540ms–1 dure, the mean dose per cumulated activity S is introduced. S is ) calculated for each radionuclide and for individual source–target and time (total transmit-receive duration). The additional pair. The total radiation dose received by a target organ from a reflection(s) increase the pulse transmit-receive time and there- specific source organ is fore inaccurately places the second (mirrored) target posterior to the large interface surface. This can be seen in the figure below. D (rk ¬ rh ) = A 1 ´ åfi (rk ¬ rh )Di m k i (M.9) = A ´ S (rk ¬ rh ) where D(rk ← rh) is the radiation dose received in target organ rk from source organ rh: mk is the target organ mass. More informa- tion about the different steps in the radiation dose calculation can be found in their respective articles (see Further Readings). To create patient-specific S-factors, a set of 3D CT slices are required. From these, the segmentation of source and target organs is performed. This is then run through a Monte Carlo code to generate a table of S-factors. This process is so far too tedious to use in the clinic, so a set of S-factors derived from standard phantoms are almost always used. In this step, assumptions are made about the human anatomy, namely that all patients share the same internal structure, when in fact the structure varies from patient to patient, especially patients with pathological abnormal- ities. S-factors can be found from MIRD, OLINDA, ICRP. The models (or phantoms) used to model the human body were published by the MIRD Committee of the Society of Nuclear Medicine. Several phantoms were developed; 1-, 5-, 10-, 15-year- old, and adult individuals. The development of new phantoms has gone a long way and S factors for more human looking phantoms can be found in the literature. Misadministration Related Articles: Cumulated activity, Equilibrium absorbed (Radiotherapy) Misadministrations in radiotherapy (either in dose constant, Absorbed fraction, Mean dose per cumulated external beam radiotherapy or brachytherapy) can take many dif- activity, Limitations to the MIRD formalism ferent forms. Possible scenarios include: Misregistration 613 Mobile target volume • Irradiation from a source other than the one ordered in nuclear magnetic resonance spectroscopy. Phys. Med. Biol. • Irradiation of the wrong person 51:R579–R636. • Radiation administered by a route other than the route prescribed by the authorised physician MLC QA • Radiation administered to a part of the body other than QA considerations for MLCs include the following: the part of the body specified by the authorised physician M • The activity of a therapeutic radiopharmaceutical dif- • Leaf position accuracy fering from the prescribed activity

by more than 10% • Collimator leakage (tongue and groove leakage) • Any error (e.g. equipment malfunction, error in com- • Leaf speed control putation, calibration, time of exposure or treatment • Interlocks geometry) that results in a discrepancy of at least 10% between total prescribed and total delivered dose Abbreviations: IMRT = Intensity-modulated radiotherapy, • Any error that results in a discrepancy of at least 50% MLC = Multileaf collimator, MSF = Multiple static field and PTV between the prescribed and delivered dose per fraction = Planning target volume. Related Article: Intensity-modulated radiotherapy (IMRT) Note: The above scenarios are examples only and it is impor- tant to check local regulations since they may differ from the details above. MLCs and IMRT MLCs are used in the delivery of intensity-modulated radiother- apy (IMRT) enabling the fluence of the beam to be modulated Misregistration over time and position. There are two methods, the multiple static (Nuclear Medicine) Image registration refers to the process of field (MSF) technique, and the dynamic MLC technique. The accurately aligning one image with another. This is particularly MSF technique uses the MLCs to create a sequence of 2D-shaped useful in nuclear medicine and PET as these images suffer from fields, only moving the MLC when the beam is off. The dynamic a lack of anatomical detail. When image registration is per- MLC method uses continuously moving leaves whilst the radia- formed and the images are not accurately aligned this is known as tion is on. Varying the velocity of each leaf pair can create any misregistration. required 2D profile. Related Articles: Image fusion, Image registration, Registration Further Reading: Hill, D. L. G., P. G. Batchelor, M. Holden and D. J. Hawkes. 2001. Medical image registration. Phys. Med. Mobile shield Biol. 46:R1–R45. (Radiation Protection) A mobile shield is a protective shield used for the protection of the worker during the use of ionising Mixed radionuclides radiation. (Nuclear Medicine) The term mixed radionuclides refer to the use Fluoroscopy and Interventional Applications: Additional of multiple radioisotopes in nuclear medicine. Different imaging protective mobile shields should be available in fluoroscopy and radionuclides (i.e. gamma emitters) are labelled to two different in interventional radiology rooms, which include tracers in order to study two separate biological processes that spatially overlap. The two tracers can be spatially separated using 1. Ceiling suspended protective screens an energy separation. Another application involves the use of one 2. Protective lead curtains mounted on patient table imaging radionuclide and one radionuclide with a therapeutic 3. Protective lead curtains for the operator if the x-ray tube purpose labelled to the same tracer. The imaging radionuclide is placed in an over the couch geometry and if the radi- allows the user to perform in vivo evaluation of the targeting to ologist must stand near the patient tissue ratio. Mobile X-Ray Equipment: Mobile shield, with generally 2.5 Mixing time (T mm Pb equivalent, is used to protect the operator when mobile M) (Magnetic Resonance) Mixing time is the term used to refer x-ray equipment is used to perform x-ray investigation, in the to the interval between the second and third 90° pulses in the ward, at the patient’s bed. The mobile shield will have a glass STEAM pulse sequence. During this interval, magnetisation is window with the same amount of protection (lead equivalent) in stored along the z-axis and undergoes T1-relaxation, introducing order to allow the operator to see the patient while performing the a weighting exp(−TM/T1) into the final signal. This weighting is investigation. rarely significant, and there is a more significant reduction in T2- Mammography: Usually light, transparent shields are incor- weighting due to storage of the magnetisation in the z direction, porated with the equipment in order to allow the operator the see yielding an effective reduction in TE. the patient and remain in the room during the exposure. Movement, including diffusion, during TM is a known problem Further Readings: IAEA (International Atomic Energy with STEAM. Physiological motion during this interval can result Agency). 1996. International basic safety standards for protection in further loss of signal from the VOI, over and above the inherent against radiation and for the safety of radiation sources. Safety 50%. A variety of approaches exist to compensate for this. Series No. 115, International Atomic Energy Agency, Vienna, The development of multiple quantum coherences during T Austria; Sutton, D. G. and J. R. William (ed.). 2000. Radiation M results in complex behaviour of J-coupled species during the shielding for x-ray. BIR report, Chilton, England. STEAM sequence. Related Articles: Magnetic coupling, STEAM Mobile target volume Further Readings: Frahm, J., K.-D. Merboldt and W. Hänicke. (Radiotherapy) The mobile target volume has been suggested as 1987. Localized proton spectroscopy using stimulated echoes. J. an extension to the ICRU 50 formalism (gross tumour volume, Magn. Reson. 72:502–508; Keevil, S. F. 2006. Spatial localization planning target volume, clinical target volume, etc.). The mobile Mobile unit 614 Modulation transfer function target volume would include an additional ‘safety’ margin to prescribed dose to the tumour volume in more complex treat- allow for subject motion such as respiration. ments that also conform more closely to the shape of the tumour. Related Articles: Gross tumour volume, Planning target vol- To deliver this complex conformal therapy, machines do not ume, Clinical target volume just adjust to the shape of the tumour using multi-leaf collimators, Further Readings: ICRU (International Commission on but also vary the output (conventionally recorded as the monitor M Radiation Units and Measurements). 1993. Prescribing, report- units – MU, equivalent to cGy – per minute). This type of con- ing and recording photon beam therapy. ICRU Report 50, formal therapy is known as intensity-modulated radiotherapy Washington, DC; ICRU (International Commission on Radiation (IMRT). When the machine is capable of delivering this treat- Units and Measurements). 1999. Prescribing, recording and ment all whilst the head is continuously moving round the patient, reporting photon beam therapy (Supplement to ICRU Report 50), it is referred to a volumetrically modulated arc therapy (VMAT). ICRU Report 62, Washington, DC; Urie et al. 1991. The role of During IMRT or VMAT, delivering a prescribed dose to uncertainty analysis in treatment planning. Int. J. Rad. Oncol. the tumour volume takes longer in terms of beam on time than Biol. Phys. 21:91–107. it would using the old conventional point and shoot geometry. However, the radiation leakage rate from the machine head will Mobile unit remain the same, potentially having an impact on the shielding (Diagnostic Radiology) Mobile unit is the medical jargon used design for the treatment room. for any type of small low-power x-ray equipment (radiographic Therefore, in calculating the required shielding in terms of or fluoroscopic) on wheels. Mobile units are powered mainly by thickness of concrete, we must introduce a factor to allow for high frequency generators and use low-power x-ray tubes. Some this modulation – the ratio between the monitor units it takes to radiographic mobile units use capacity discharge generators or deliver an IMRT/VMAT treatment, and the monitor units it would mono-block generators. Mobile units are convenient to move into take to deliver the same prescribed dose using conventional beam patient rooms or the operational theatre, as well to be used as field therapy – i.e: units in military hospitals. All mobile units are powered either by a battery or directly from the mains. MU (IMRT) Modulation Factor = MU ( Conv) Modal target absorbed dose (Radiotherapy) The International Commission on Radiation Evidence shows that this factor can vary between 2–10, depend- Units and Measurements (ICRU) recommends a common method ing on the treatment. This implies that the contribution of head of dose specification which could be generally adopted to per- leakage radiation to the overall radiation dose outside the treat- mit a comparison between treatment practices. Normally a non- ment room will be 2–10 times that it would be for standard con- uniform dose distribution is obtained in the target volume and ventional radiotherapy. As a rule of thumb for shielding design therefore for practical reasons it is useful to report specific doses calculations a factor of five is normally used for IMRT, and 2.5 such as the modal target absorbed dose. for VMAT. The modal target absorbed dose is the absorbed dose that Related Articles: Intensity-modulated radiotherapy (IMRT), occurs most frequently at calculation grid points in the target area. Workload factor (W), Use factor (U), Orientation factor, The modal target absorbed dose depends on the dose cal- Occupancy factor (T) culation method and exceptionally more than one modal target absorbed dose can be found in a particular patient. Modulation transfer function Related Articles: Mean target absorbed dose, Maximum tar- (Diagnostic Radiology) Modulation transfer function (MTF) rep- get absorbed dose, Minimum target absorbed dose, Median target resents the response of an imaging system to an input signal of absorbed dose, Hot spots continuously varying spatial frequency. Modulation of a signal Further Reading: ICRU (International Commission on is defined as Radiation Units and Measurements). 1978. Dose specification for reporting external beam therapy with photons and electrons. I(max) - I(min) ICRU Report 29, Washington, DC. Modulation(M) = I(max) + I(min) Modalities where (General) A modality refers to a family of imaging systems based I(max) is the maximum intensity of the signal (e.g. white in on the same physical phenomena, for example SPECT and PET. greyscale image) Modality also refers to an individual imaging system in dual or I(min) is the minimum intensity (e.g. black in greyscale image) multi-modality system, for example the SPECT in a SPECT-CT system. This can also be interpreted as the total contrast of the image. The important radiological issue is what a radiologist will be able Mode conversion to distinguish in the images (see the following hyperlink for a (Ultrasound) When a longitudinal wave crosses the boundary (at discussion of issues), but the MTF is also reported for standard an angle) into a solid material, some of the energy can cause par- digital cameras. MTF is defined as the ratio between the modula- ticle movement in the transverse direction. This is called mode tion of the resulting image (M(i)) and the modulation of the input conversion. object (M(o)) as function of spatial frequency (ν): Related Articles: Transversal wave, Longitudinal wave M(i) MTF(n) = Modulation factor M(o) (Radiation Protection) Modern radiotherapy machines have In the determination of the performances of medical imaging sys- developed significantly over the past 20 years to deliver the tems, MTF gives a complete description of the spatial resolution Modulation transfer function (MTF) 615 Modulation wheel in the spatial frequency domain. Two main procedures can be fol- Hyperlink: http: / /www .imag ingec onomi cs .co m /iss ues /a rticl lowed for the evaluation of MTF. es /MI _2 001 -05 _1 0 .asp Bar Test Object: These are built to represent an object with groups of black and white line pairs, giving an input image with Modulation transfer function (MTF) increasing spatial frequency. Each group of line pair constitutes (Magnetic Resonance) Modulation transfer function is a means a specific spatial frequency, expressed as line pair per length unit of quantifying the spatial resolution of an MRI system. MTF spe- M (e.g. lp/mm), and the measurement of the modulation of each cifically looks at perceived contrast within the image. To measure group represents a set of experimental points for the determina- MTF, a test object with a high contrast, angled block is scanned tion of the MTF curve. MTF is expressed as per cent variation in (Figure M.44). An edge response function can then be obtained the image modulation normalised to the modulation at low fre- from the image. By differentiating the edge response function, a quency where there is no significant loss of the signal. line spread function is produced. The full width half maximum Fourier Transformation: Spatial resolution is related to the measure is taken from the line spread function to give a good capability of an imaging system to represent the sharpness of a indication of the spatial resolution. high contrast edge. Starting from the experimental measure- ment of the edge response function (ERF) of a high contrast edge Modulation wheel image is possible to evaluate, by derivation, the line spread func- (Radiotherapy) In passive scattering, a modulation wheel or ‘pro- tion (LSF) which represents the response of the system to a

high peller’ is used to spread the beam energy and adjust the range or contrast linear signal (Dirac delta function). The Fourier transfor- mation of the LSF represents the MTF of the image system in the spatial frequency domain. An image has two essential components: edges where intensi- ties (densities, grey values) change, and contrast, the amount by which they change. A sharp edge is a line along which change is steep. Imperfect imaging systems will affect such lines. They may spread the region on which the change happens, or they may reduce size of the change, and both effects amount to the same up when the image is normalised. This duality explains how con- trast (or modulation if normalised) and edge sharpness relate. To describe sharp edges in the frequency domain, we need high spa- tial frequencies. The MTF, also called spatial frequency response, describes how the system affects different spatial frequencies dif- ferently (Figure M.43). Abbreviation: MTF = Modulation transfer function. Related Articles: Point-spread function, Optical transfer func- tion, Contrast inversion Further Reading: Barrett, H. H. and K. K. Myers. 2004. Foundations of Image Science, Wiley Series in Pure and Applied Optics, Wiley, New York. FIGURE M.44 Eurospin test object showing MTF object. Spatial freq in MTF Spatial freq out 1 1 0.9 1 0.5 0.5 0 0.8 0 –0.5 –0.5 –1 –1 0.7 0.6 0.5 1 0.4 0.5 0 1 –0.5 0.3 0.5 –1 0 –0.5 0.2 –1 1 1 0.5 0.1 0.5 0 0 –0.5 –0.5 –1 0 –1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 FIGURE M.43 MTF diagram of an imaging system showing the input spatial frequencies (left), and how these are modulated by the system- – output spatial frequencies (right). Note the degradation of the amplitude of the output for higher frequencies. Modulator 616 Molecular imaging M FIGURE M.45 Moiré pattern artefact. M.45a (left) – The Moiré pattern begins to appear with the distortion of the test object patterns in the third column (from left to right) and is very prominent with the increasing of spatial frequency after this column. M.45b (right) – visualisation of test object from 1a at the same TV monitor – Moiré artefact disappear due to the angle between the two periodical patterns (test object and horizontal TV raster). depth of penetration of a mono-energetic proton beam (Khan and object over TV monitor – in this case the interference is between Gibbons, 2014). The wheel is usually made of several plastic seg- the periodic structure of the test object and the monitor raster ments of increasing thicknesses. As the wheel rotates, the range (Figure M.45a). The Moiré pattern artefact also depends on the and energy of the pristine Bragg peak is reduced by the increas- angle between the two interfering periodic structures – e.g. the ing thickness and width of the segments, creating the required artefact in Figure M.45 does not appear when the Huttner test spread-out Bragg peak (SOBP). object is placed at about 45 degrees to the TV monitor raster Related Articles: Bragg peak spreading, Beam modulation, (hence the requirement for such positioning of the test object dur- Range modulation, Spread-out Bragg peak (SOBP) ing Quality Control procedures) (Figure M.45b). Further Reading: Khan, F. M. and J. P. Gibbons. 2014. Khan’s Related Articles: Zebra-stripe artefact the Physics of Radiation Therapy, 5th edn., Wolters Kluwer Further Reading: Kim, Y., D. Oh and A. Hwang. 2017. Health. Small-scale noise-like moiré pattern caused by detector sensi- tivity inhomogeneity in computed tomography. Opt. Express Modulator 25(22):27127–27145. (Radiotherapy) The modulator in a linac generates high voltage (approximately 100 kV) and current in short pulses (the order of Molar mass microseconds) to the radiofrequency power source for the produc- (Nuclear Medicine) Molar mass (abbreviated M) is the mass of 1 tion of the microwave radiation (either klystron or magnetron) and mol of a chemical element or compound. The molar mass unit is to the electron gun. g/mol. The molar mass of a molecule is the sum of the isotopes Typically the modulator circuitry is kept in a separate cabinet individual molar masses. which can be located in the treatment room, or in a control room nearby, and with the high voltages involved interlocks on the door Mole of the modulator cabinet are used. (Nuclear Medicine) For a small sample it is inconvenient to spec- More detail on the circuitry used and information on the PFN ify the large number of molecules. Instead the mole is used which and PRF are given in Further Reading. corresponds to the number of atoms in 12 g of the isotope 12C. Abbreviations: PFN = Pulse forming network and PRF = This number is 6.02 × 1023 atoms and the number or constant is Pulse repetition frequency. called Avogadro’s number. Related Articles: Linear accelerator, Klystron, Magnetron Related Article: Avogadro’s number Further Reading: Greene, D and P. C. Williams. 1997. Linear Further Reading: Benson, H. 1996. University Physics, Accelerators for Radiation Therapy, 2nd edn., IOP Publishing, revised edition, Wiley, New York, p. 369. Philadelphia, PA. Molecular excitation Moiré patterns (Nuclear Medicine) This is a process in which a molecule is (Diagnostic Radiology) Moiré patterns exist in images produced excited from a low energy state to high energy state. The energy with various imaging modalities. In principle these are artefacts, state of a molecule is modelled by the molecular Hamiltonian resulting by the interference of at least two overlaying periodic where each solution or eigenvalue represents an energy state patterns (which are often with different spatial frequencies). The that is the sum of electronic, rotational, nuclear and translational term Moiré comes from the name of a type of textile with a rip- components. pled appearance. In medical imaging, Moiré pattern artefact can be related to Molecular imaging monitor, anti-scatter grid, aliasing, etc. A typical Moiré pattern (Nuclear Medicine) Visualisation, characterisation and measure- artefact is observed in fluoroscopy, when visualising Huttner test ment of biological processes at molecular and cellular scale in Molecular mass 617 Molybdenum humans and other living systems. Molecular imaging includes chemical element (e.g. oxygen gas O2) or different elements (e.g. two- or three-dimensional imaging as well as quantification over water H2O). time. While the bulk of clinical molecular imaging uses radio- Molecules cannot be defined for ionic crystals, covalent crys- tracers in nuclear medicine and PET for two- or three-dimen- tal networks or metallic substances, although these normally sional imaging, the term molecular imaging can also be used to consist of repeating units. Glasses contain atoms held together by characterise areas such as MR spectroscopy, optical fluorescence chemical bonds but they are amorphous meaning that a molecule, M imaging and contrast ultrasound imaging. or even a regular repeating unit, cannot be defined. Conventional imaging studies that use x-rays (radiography, flu- Related Articles: Atom, Molecular mass oroscopy, CT) produce anatomical images of the body. Molecular imaging and nuclear medicine allow the visualisation of the func- Molière scattering theory tional processes of the body at the cellular and molecular level, (Radiation Protection) The Molière scattering theory attempts thus producing a physiological image. X-ray imaging uses the to describe the small-angular distribution of electrons that have x-ray source, which is outside the patient, to cast a ‘shadow’ of the undergone elastic collisions with the atomic nuclei of an absorb- patient’s anatomy on to the imaging detector (‘transmission imag- ing medium – that is, the electrons do not lose energy during these ing’). Nuclear imaging introduces ‘tracers’ inside the patient’s scattering interactions, but do change direction. body and forms images by acquiring the emissions with external This theory, together with the Klein–Nishina differential detectors (‘emission imaging’). cross-section, Bethe–Bloch equation and others, attempts to Related Articles: Magnetic Resonance Spectroscopy, describe interactions between ionising radiation and matter at Radionuclide Imaging, Ultrasound Contrast Agents an atomic level, and forms the mathematical basis for radiation Further Readings: Bushberg, Seibert, Leidholdt and Boone. dosimetry based on Monte Carlo statistical modelling. 2012. The Essential Physics of Medical Imaging, 3rd edn., Related Articles: Elastic scatter, Klein–Nishina differential Lippincott Williams and Wilkins, Philadelphia, PA; Cherry, cross-section, Bethe–Bloch equation Sorenson and Phelps. 2012. Physics in Nuclear Medicine, 4th edn., Elsevier; Fact Sheet: What is Nuclear Medicine and Molecular Imaging? Molybdenum Hyperlink: www .s nmmi. org /A boutS NMMI/ Conte nt .as px ?It (General) emNum ber =5 648. Molecular mass Symbol Mo (Nuclear Medicine) The molecular mass is the weight of one Element category Transition metal molecule expressed as a multiple of 1/12 of the total weight of a Mass number A of stable isotopes 92 (14.84%); 94 (9.25%); 95 12C atom. Molecular mass is sometimes referred to as molecular (15.92%); 96 (16.86%); 97 weight. The molecular mass of a single molecule is not always (9.55%); 98 (24.13%); and 100 identical to the average molecular mass in a sample. The molecular (9.63%) mass differs because of the natural occurrence of isotopes. Atomic number Z 42 Atomic weight 95.94 kg/kg-atom Molecular targeting Electronic configuration 1s2 2s2 2p6 3s2 3p6 3d10 4s2 4p6 4d5 (Nuclear Medicine) Molecular targeting may be defined as the 5s1 specific concentration of a diagnostic tracer or therapeutic agent Melting point 2896 K because of their interaction with a molecular species. The targeted Boiling point 4912 K molecular species could be a mutated locus of DNA or its protein Density near room temperature 10,280 kg/m3 or RNA product; a gene product of normal sequence and struc- ture, aberrantly expressed in a given tissue; or a transcriptionally normal gene product whose structure or function has been modi- History: During the two world wars, molybdenum alloys were fied by abnormal RNA splicing or post-translational processing. commonly employed in the armour plating of tanks. In modern Molecular targeting agents differ from conventional physiologic times molybdenum compounds are used in pigments, catalysts tracers, such as radioiodide or myocardial and cerebral perfusion and electrodes. Molybdenum is also used as target material in the indicators, as physiologic tracers are directed toward processes production of some x-ray tubes. characteristic of a normal tissue or cell type and are dependent on Isotopes of Molybdenum: Molybdenum has 35 known iso- the cooperation of multiple gene products or on non-specific bulk topes, 7 of them are stable. The isotope of interest in medical processes, such as diffusion, membrane permeability or electro- physics is 99Mo produced either as a product of uranium-235 fis- static interactions. sion or as neutron activation product. Whereby, stable molybde- num-98 is bombarded with thermal neutrons in a nuclear reactor. Molecular weight (Nuclear Medicine) See Molecular mass Isotope of molybdenum 99Mo Molecule Half-life 65.94 hours (General) A molecule is an electrically neutral group of two or Mode of decay β−, γ more atoms connected by covalent bonds. Atoms attached by Maximum decay energy, Emax β−: 1.214 MeV, γ: 0.74 non-covalent bonds, such as hydrogen bonds or ionic bonds, are not strictly regarded as molecules. However, the term may some- times be used to describe certain charged organic molecules Medical Applications: Technetium generator: The daughter or biomolecules. A molecule may contain atoms of the same product of 99Mo is 99mTc, a radionuclide widely used in diagnostic Molybdenum breakthrough 618 M onochromator nuclear medicine studies (6.01 h half-life). To ensure that tech- The absorbed dose rate at a reference point in the phantom is also netium radiopharmaceuticals are readily available, many clinical proportional to Np. nuclear medicine departments run molybdenum-based techne- tium generators on site (in hospital radiopharmacies). Monitoring Related Articles: Technetium-99m, Technetium generator, (General) Monitoring refers to the observation, detection and M Target of x-ray tube periodic analysis of the trend of variable quantities such as physi- cal, chemical and biological parameters, actions and processes, Molybdenum breakthrough also through the use of detection and measurement tools, in (Nuclear Medicine) Molybdenum breakthrough occurs when the order to perform systematic and continuous surveillance on sub- radioactive parent 99Mo is partially eluted with the daughter 99mTc jects – such as human patients in hospitals, animals in veterinary from a radionuclide generator. Samples used for clinical imag- clinics or on biomedical equipment in healthcare facilities. It ing must not contain any 99Mo because it will contribute to an also includes recording evidence, such as periodic maintenance unnecessary radiation dose to the patient. One can estimate the results, that must be compared with the related parameters in 99Mo contamination by using specific detector arrangement with

order to evaluate the status of the device. a thick photon absorber (e.g. 30 mm of lead) between the crystal Systematic control and monitoring with periodic maintenance and the sample. The 140 keV photons emitted from 99mTc will be allow evaluation and correction of errors and anomalies found, in attenuated by the shield. But for the 740–780 keV photons from order to make improvement actions to guarantee quality processes 99Mo the absorber is relatively transparent. and adequate performance of equipment in healthcare facilities on an ongoing basis, but also to carry out prospective analysis Monitor chamber regarding desirable advantages or potential adverse events. (Radiotherapy) A monitor chamber is included in the head of a linear accelerator to control continuously the radiation beam. Monitor unit The beam monitor consists of a set of flat transmission ionisa- (Radiotherapy) Monitor units (MUs) are used in radiotherapy tion chambers which are usually mounted close to the accelerator to measure out the dose delivered by a linear accelerator. They window between the flattening filter and the collimators so that are the equivalent of ‘beam-on’ time for a Cobalt-60 machine. the beam passes through them. The monitor chamber must cause Commonly, the linear accelerator is calibrated to deliver 1cGy/ minimal perturbation of the radiation beam and therefore the elec- MU at reference conditions for photon beams and at dmax for a 10 trodes are formed by deposition of carbon or metal onto a thin cm × 10 cm field at SSD = 100 cm for electron beams. substrate of mica or a plastic foil such as Melinex or Kapton. The MUs are measured by the pair of transmission monitor ioni- substrate has a typical thickness of 0.1 mm. The reduced thickness sation chambers mounted in the treatment head of the linear of the electrodes permits a reduction of the radiation scattering, accelerator. The dose monitor chambers are arranged either as the bremsstrahlung contamination of the beams, the broadening a redundant dose monitoring combination or as a primary/sec- of the electron energy spectrum and the production of low energy ondary dose monitoring combination. When they have measured secondary electrons. The collection efficiency of the monitor the exact amount of radiation to pass through, the machine is chamber must be about 99% at the high dose rate of the linear switched off. If one chamber fails, then the second one will inter- accelerator and this requires a polarising voltage of about 300 V. rupt the beam. The monitor chamber consists of multiple parallel electrodes Abbreviation: MU = Monitor unit. which form a double ionisation chamber system for two indepen- dent dose monitoring channels. The chamber plates are mounted Mono-block generator so that one is rotated 90° on the beam axis from the other. This (Diagnostic Radiology) Although called mono-block HVG, this is arrangement allows the on-line testing of the beam symmetry and not a new design of HVG, but is a construction of the x-ray tube flatness in the beam’s radial and transverse planes, respectively. housing, which incorporates not only the x-ray tube, but also the Beam flatness and symmetry are controlled by feedback circuits high voltage transformer, the rectifiers, etc. All components are that run from the ion chambers to the bending magnet’s beam immersed in isolation oil in the mono-block box. Most often these steering coils. units are used for dental radiography, having a stationary anode The response of the monitor chamber depends on the mass tube and two pulse rectifier. This design can also be used with of gas between its electrodes and therefore it is necessary to cor- low-power rotating anode tube in some mobile x-ray equipment. rect the system response for any variation of temperature, humid- Mono-block generator is also called single block generator. ity and pressure in the chamber volume in case of an unsealed Related Articles: High voltage generator, Voltage waveform monitoring chamber filled with air. Sealed chambers have been also used. The response of the monitor chamber is influenced Monochromator by the backscatter radiation from the jaws and is dependent on (Non-Ionising Radiation) A monochromator is a specialist photo- the radiation energy. The prescribed dose needs to be delivered biological source which can emit narrowband non-coherent light, reproducibly for each patient treatment. To achieve this routinely, with a half maximum bandwidth of few nm. one set of electrodes of the monitor chambers is used to monitor The core elements of the monochromator are: the beam output, so that the dose can be measured and delivered reproducibly. • a broadband source emitting in the 200–1000nm range The units of the signal recorded by this set of electrodes are • a monochromator chamber, like those used in referred to as monitor units (MU). One MU of dose has been spectroradiometers. delivered when the monitor chambers have detected a preset dose. The ionisation current Ic is proportional to Nc, number of Light monochromators are Risk Group 3 optical sources, as even radiation quanta per unit of time passing through the chamber a short unprotected exposure to some of the emissions can be haz- and to Np, number of radiation quanta incident on the phantom. ardous (Figure M.46). Monoclonal antibodies (mAb) 619 Monte Carlo method M FIGURE M.46 Schematic of a light monochromator (single chamber). Related Articles: Photobiological lamp safety, thickness, (namely the thickness of a specific material which Spectroradiometer, Solar simulator reduces the intensity of the radiation entering the material to half) Further Reading: Mackenzie, L. A. and W. Frain‐Bell. 1973. will remain the same. Therefore the 1st HVL is equal to the 2nd The construction and development of a grating monochromator HVL. The homogeneity factor (HF) is the ratio between the 1st and its application to the study of the reaction of the skin to light. and the 2nd HVL and is a measure of the polychromaticity of the Br. J. Dermatol. 89(3):251–264. beam. In case of mono-energetic beams, HF is equal to one. Mono-energetic beams of ionising radiation are very difficult Monoclonal antibodies (mAb) to obtain. They would represent a great advantage for medical (Nuclear Medicine) Antibodies are the body’s natural defence therapy as well as diagnosis with ionising radiation. The devel- against foreign substances known as antigens (e.g. disease- opment of synchrotrons and other advanced equipment as laser causing bacteria and viruses). Each antibody has an affinity for plasma accelerators make it possible to select monochromatic a specific antigen, and some antibodies, once activated, will con- beams; but this kind of equipment is still far from the routine tinue to provide resistance (this is the basis for the production of medical practice. vaccines). In the past antibodies have been produced by injecting a lab- Monte Carlo calculations oratory animal with an antigen and when antibodies have been (Nuclear Medicine) Monte Carlo is a numerical simulation formed, collecting these from the blood serum. One problem with method used to solve problems that involve stochastic events this method is that it produces a very small quantity of usable with defined probabilities. Physical processes, such as photon antibody. Monoclonal antibody technology on the other hand pro- and electron interactions, can be modelled provided that the vides a technique to produce large quantities of pure antibodies. processes can be accurately estimated by probability density This is achieved by forming a hybrid of cells that produce antibod- functions. With all the physical preconditions modelled, simu- ies naturally together with cells that can grow continually in cell- lations of a very large number of histories (e.g. photon or elec- culture. The result of this cell fusion is called a hybridoma and it tron tracks) are required to provide good statistical properties will continually produce antibodies. These antibodies are called in the result. A uniformly distributed series of random numbers monoclonal antibodies (abbreviated as mAb) because they come is essential in Monte Carlo simulation since these numbers are from only one type of cell (the hybridoma cell). Antibodies pro- used to properly select between the different processes that can duced in the conventional manner are called polyclonal because occur. they come from substances that contain many kinds of cells. In a Monte Carlo simulation it is possible to change different The technique for producing monoclonal antibodies was first parameters in order to investigate their influence on the system. A described by Kohler and Milstein (1975). They found that when typical effect that is very difficult or even impossible to determine myeloma cells were fused with antibody-producing mammalian experimentally is the component in an SPECT image caused by spleen cells, the resulting hybridomas produced large amounts of photons scattered in the object. With Monte Carlo simulation it monoclonal antibodies. is relatively straightforward since all details about a history are In nuclear medicine, antibodies against tumour specific anti- known and from this it is easy to ‘tag’ a photon as being scattered. gens are used for tumour targeting. They are then labelled with a Related Article: Acquisition time radionuclide and can be used for diagnostics (radioimmunodiag- Further Readings: Ljungberg, M., S. E. Strand and M. A. nostics) or therapy (radioimmunotherapy). King. 1998. Monte Carlo Calculations in Nuclear Medicine, Abbreviation: mAb = Monoclonal antibody. Institute of Physics Publishing, Bristol, UK, pp. 1–11; Zaidi, H. Related Articles: Radioimmunotherapy, Tumour targeting and G. Sgouros. 2003. Therapeutic Applications of Monte Carlo Further Reading: Kohler, G. and C. Milstein. 1975. Calculations in Nuclear Medicine, Institute of Physics Publishing, Continuous cultures of fused cells secreting antibody of pre- Bristol, UK, pp. 1–23. defined specificity. Nature 256(5517):495–497. Monte Carlo method (Radiotherapy) The Monte Carlo method involves the use of ran- Mono-energetic beam dom number based statistical sampling to model a series of physi- (Radiation Protection) Mono-energetic beams have a single cal processes. It is particularly useful for modelling a series of value of linear energy transfer (LET). The half value layer (HVL) physically independent events. Monte Carlo photon transport simulation 620 Moodle A particularly common application of Monte Carlo in radi- 1 ology and radiotherapy is the modelling of ionising radiation transport. The concept is that the ionising radiation is described by modelling many independent samples and the statistics of 0.75 R the samples is analysed to determine the cumulative behaviour M of many particles of the radiation. For each particle, the energy, 0.5 direction, etc. are randomly sampled at the start of the particle’s ‘history’ (where history is used to describe the chain of events θ that happens to each individual particle that is modelled). The 0.25 interaction of the radiation with matter is then randomly sampled according to the probability of interaction and the nature of that interaction. Any secondary radiation that is created from the 0 interactions is then followed similarly, in order to describe the –180 –90 0 90 180 entire radiation effect. This methodology may be used to model Angle the dose delivered to the patient by a treatment (such as exter- nal beam radiotherapy or brachytherapy), to determine the dose FIGURE M.48 Cumulative probability distribution from Figure M.39, required to form an x-ray image, or to calculate the dose response with sampled random number R and the corresponding angle, θ. of a detector such as an ionisation chamber. A series of off-the-shelf packages are often used for Monte Carlo radiation transport, these include EGS/BEAM (where EGS Hyperlinks: The EGS/BEAM package (http: / /www .irs. inms. stands for electron gamma shower), GEANT, PENELOPE and nrc .c a /pap ers /C CRI99 /n ode 21 .ht ml); The GEANT package others. Often the packages require a user file to specify the geom- (http://wwwasd .web .cern .ch /wwwasd /geant/); The PENELOPE etry and physical requirements of the problem to be solved. package (http://www .nea .fr /abs /html /nea -1525 .html) Statistical Sampling: A variety of approaches are available for statistically sampling events to determine a particle’s history. Monte Carlo photon transport simulation An example is the cumulative probability distribution, CPD. In (Nuclear Medicine) The Monte Carlo method is useful for solving this approach, if the probability is p(x) as a function of the inde- very complex problems of a stochastic nature. In nuclear medi- pendent variable, x, then the summed probability between −∞ cine, radiation is used for imaging and for the treatment of cancer. and x: P(x) = ò x p(y)dy is calculated. A random number, R, In imaging, it is mainly the photon transport that is used to create

-¥ between 0 and 1 is generated and the value of x for which P(x) = images of radionuclide distributions. The Monte Carlo simulation R is used to find x by inversion. of a photon transport includes simulation of photo-absorption, To illustrate this, consider the following. Figure M.47 shows a Compton scattering and coherent scattering. By sampling from hypothetical Gaussian angular distribution for scattering of a parti- probability distribution functions, each individual photon can be cle. Figure M.48 shows the cumulative probability distribution. We followed in a phantom towards a simulated camera. It is neces- may sample this distribution in 1 by generating a random number sary to estimate for each interaction, the reduction in energy as between 0 and 1 (R in Figure M.48). We then read off the value of well as the scattering angles. The simulation needs to include a the angle for which the CPD is equal to R. This gives a value of θ. large number of histories in order to maintain good statistics in Abbreviations: CPD = Cumulative probability distribution the results. and EGS = Electron gamma shower. Examples of Monte Carlo codes for photon transport simu- Related Articles: Radiotherapy treatment planning, X-rays, lation in nuclear medicine applications are the SIMIND code, Electrons SIMSET code and the GATE/GEANT4 code. More information Further Reading: Rogers, D. W. O., B. A. Faddegon, G. X. about these codes can be found on the Internet. Ding, C. M. Ma, J. We and T. R. Mackie. 1995. Beam – a Monte- Carlo code to simulate radiotherapy treatment units. Med. Phys. Moodle 22:503–524. (General) Moodle is one of the major virtual learning environ- ment (VLE) platforms used nowadays in higher education, and very popular in the field of medical physics. 0.015 Moodle’s development began in 1999 and it exists in its pres- ent form (as at 2020) since 2001. It was developed in Perth, Australia by Martin Dougiamas. Moodle is one of the three most 0.01 popular VLE (together with WebCT and Blackboard). These plat- forms command about 80% of the total VLE market. Moodle has over 50% of market share in academia in Europe, Latin America and Oceania and taking into account the big US higher education 0.005 market, across the globe Moodle and Blackboard share the lead- ing position. Moodle’s typical features, as one of the leading VLE plat- 0 forms in higher education, are: –180 –90 0 90 180 Angle • The possibility to manage courses, users and roles: The Moodle platform is typically used to create pro- FIGURE M.47 Hypothetical probability distribution for scattering fessional structured course content. There are basically angle. three major roles – Teacher, Student and Manager – and Probability Cumulative probability MOS-FET detector 621 Motion artefact several subsidiary ones and they are subject to hierar- At the phantom surface two energy parameters are of particu- chical relations (i.e. control of access to content for the lar interest; the most probable energy Ep,0 that corresponds to the different roles, tracking student progress; managing and peak of the distribution and the mean energy E0 . These param- monitoring student attendance and participation). eters can be easily determined by practical measurements and are • Online assessment: utilised to specify the electron beam quality. The Moodle platform is suitable for both summa- The most probable energy of the electron beam at the surface M tive and formative assessment. It allows the creation of of the phantom Ep,0 is related to the practical range in water Rp, various types such as offline tasks (i.e. in the form of expressed in centimetres, by essays) as well as different multiple question types such as: single/multi-line answer; multiple choice answer; Ep,0 = 0.22 + 1.98RP + 0.0025R2 P (MeV) drag-and-drop order; essay; true or false/yes or no; fill in the gaps; agreement scale, etc. The energy losses of the electron beam when it traverses layers of • User feedback: matter results in a broader spectrum. Experimental data indicate Moodle is a very convenient medium for students’ that there is approximately a linear relationship between most exchange of feedback both with teachers and their peers probable energy Ep,0 on the phantom surface and at the depth z through discussion groups, forums, etc. and is given by The prerequisites for introducing a VLE system such as Moodle for the purposes of e-teaching medical physics æ z ö and other image-rich subjects are listed in the Further Ep,z = Ep,0 çç1 - ÷÷ Reading section. è Rp ø where Related Articles: e-learning Ep,0 is the most probable energy at the surface Further Reading: Tabakova, V. 2020. e-Learning in Medical Rp is the practical range of the electron beam Physics and Engineering: Building Educational Modules with z is the depth in the phantom Moodle, CRC Press. Hyperlink: www .moodle .org Related Articles: Mean electron energy, Electron practical range MOS-FET detector (Radiation Protection) See Metal oxide semiconductor field- Motion artefact effect (MOSFET) transistor (Diagnostic Radiology) Motion artefacts can be the result of vol- untary (swallowing, breathing), or involuntary (cardiac motion, Most probable energy peristalsis) patient movement. These artefacts are often seen (Radiotherapy) In a linac the electron beam in the accelerating in CT scanning. Motion during the course of a scan results in component reaches a specific final energy that depends on the streaks in the scan plane images. This is due to inconsistencies design of the accelerating structure. The intrinsic electron beams in the acquired projection data. Movement in between rotations are almost mono-energetic because their energy distributions are results in misregistration artefacts in multiplanar reconstructions very narrow. After leaving the accelerating structure the spec- (MPRs) and 3D reconstructions. trum degrades and the clinical electron beams are characterised During the course of a CT scan, patients are usually requested by a number of parameters such as the maximum energy Emax, the to hold their breath in order to eliminate breathing artefacts. They most probable energy Ep and the mean energy E . In Figure M.49, are given positioning aids, and in some cases may be immobil- the typical distribution of the electron beam energy is shown ised. Young children or uncooperative patients may need to be before leaving the accelerating structure of the linac [graph (a)]; scanned under a general anaesthetic. at the phantom surface [graph (0)]; and at a depth z in the phantom Motion artefacts can be corrected with the use of software [graph (z)]. which reduces the weighting of views at the beginning and end of φE/φE(Ep) (φE)z/φE(Ep,z) (φE)0/φE(Ep,0) (φE)a/φE(Ep,a) 1.0 (z) (0) (a) r 0.5 z r0 ra E (MeV) 0 Ez Ep,z Emax,z E0 Ep,0 Emax,0 Ea Ep,a Emax,a FIGURE M.49 Electron spectra and their parameters. Motion artefacts 622 Moving grid M (a) (b) FIGURE M.50 Patient motion artefact without and with correction: (a) Patient motion artefact – a streak through the abdomen and (b) software cor- rection of this artefact. a rotation (Figure M.50). However, modern, multislice CT scan- results can be greatly affected by patient motion, such as breath- ners have much reduced examination times and therefore motion ing and internal organ motion. These can change the patient artefacts are less of an issue. geometry compared to the initial CT scan. Motion management Motion artefacts are still a particular problem in cardiac scan- is therefore required to minimise the effects of such motions. ning due to the rapid movement of the heart. Fast rotations times, Examples of motion management include: breath-holds, beam of less than 300 ms, are now available on the latest CT scanner re-scanning, beam gating and tumour tracking (which can be models. Images are reconstructed from partial scan data to improve markerless or marker-based). Image-guided radiotherapy and temporal resolution still further, and ECG gating techniques are reducing beam delivery times can also help to mitigate against used to select the cardiac phase of least motion for reconstruction. patient motion. Dual source CT scanners are particularly suited to imaging the Related Articles: Cardiac gating, Respiratory gating, Deep heart because of the improvement in temporal resolution resulting inspiration breath-hold technique, Image-guided radiotherapy, from acquiring data simultaneously with two x-ray tubes. Fractionation Related Articles: Artefact, Beam hardening, Cone beam arte- fact, Dual source CT, Helical artefact, Image artefact, Metal arte- Motion unsharpness fact, Multislice CT scanner, Partial volume effect (artefact), Ring (Diagnostic Radiology) Image unsharpness due to movement of artefact patient during the x-ray exposure. Typical motion unsharpness is observed in chest imaging when the patient breathes during the Motion artefacts radiography (or CT scan). (Magnetic Resonance) Motion artefacts are caused by movement of the patient. These movements can be divided into two catego- Mottle, quantum ries, physiological motion such as breathing and bowel motion, (Diagnostic Radiology) Quantum mottle refers to the pattern of and gross patient movement such as small twitches. The artefact quantum noise, or random variations, in a radiographic image typically appears as a ghost along the phase-encoding direction. which is due to the statistical fluctuation of x-ray photon absorp- Flow artefacts are another type of motion artefact caused by tion and consequent light photon emission by an intensifying the flow of blood or cerebrospinal fluid within the body. Due to screen or digital radiology scintillator screen. The faster the the flow of the protons, mismapping on the image is caused lead- intensifying screen, or the higher the kV, the more light photons ing to ghosting. Flow compensation can be used to reduce this are produced and so fewer x-ray photons actually contribute to a artefact. final image of a desired optical density or pixel value. Physiological motion artefacts can be reduced by respiratory Since the emission and detection of photons are normally dis- and cardiac triggering, the use of breath-hold pulse sequences, or tributed, the noise, or randomness, associated with the number presaturation pulses, depending on their origin. of photons is proportional to the square root of the number of General patient movement cannot be avoided, but can be photons. So with fewer x-ray photons contributing to the image, reduced by ensuring the patient knows they must stay as still as the greater is the noise, or mottle, seen in the image. possible during a scan and by reducing scan times as much as is acceptable. Moving grid Abbreviation: SNR = Signal to noise ratio. (Diagnostic Radiology) The moving antiscattering grid was Related Articles: Ghost artefact, Periodic motion invented by Hollis Potter, later named Bucky–Potter, but now referred to as only Bucky grid. The advantage of the moving grid Motion management is that the lines of the grid are not seen on the image (a problem (Radiotherapy) Radiotherapy treatment is usually based on a seen in images with stationary grids, invented by Bucky). single static CT (computer tomography) scan of the patient prior The moving grid is initially displaced from the centre of the to the course of the treatment, so inter-fractional (between frac- film (i.e. is not centred). It starts its slow motion with the begin- tions) and intra-fractional (during a single fraction) uncertain- ning of the exposure and ends it at the end of the exposure, pass- ties are to be expected. For cancers in the abdomen, treatment ing through the centre at approximately half-time of the exposure. MPD (maximum permissible dose) 623 MRI-guided radiotherapy (MRIgRT) Some moving grids do have such one-directional movement, but vibrate during the exposure. As a result of this movement, the pattern of the led lamellas of the grid is blurred over the image. However this movement reduces not only the secondary (scatter) radiation, but also the pri- mary radiation. To compensate this, exposure factors (i.e. patient M dose) have to be increased by some 10%–15%. Related Articles: Grid, Bucky MPD (maximum permissible dose) (Radiation Protection) See Maximum permissible dose (MPD) MPR (multiplanar reconstruction) (Magnetic Resonance) See Multiplanar reconstruction (MPR) FIGURE M.51 The system for MRE developed by Ehman and col- MPWB leagues at the Mayo Clinic. (General) MPWB (Medical Physics for World Benefit) is an organisation created by medical physicists from Canada and the USA in 2015. The overall objective of MPWB is to provide medi- The pulse sequence used for liver MRE is typically a 2D GRE cal physics support with the goal of improving the effectiveness with short TR, short TE, and low flip angle. and safety in the use of physics and technologies in medicine, Bipolar motion-sensitising gradients are applied

similar to especially in low-to-middle income countries. those used for phase-contrast MR angiography. Moving tissues Further Reading: Van Dyk, J., Y. Pipman, G. White, D. accumulate a phase-shift when bipolar gradients are applied, but Wilkins, P. Basran and R. Jeraj. 2018. Medical physics for world stationary tissues do not. MR image acquisition is synchronised benefit (MPWB): A not-for-profit volunteer organization in sup- with the mechanical compressions, with both magnitude and port of medical physics in lower income environment. J. Med. phase data reconstructed using 4−8 different time offsets. Phys. Int. 6(1):152–156. MRE is already being used clinically for the assessment of Hyperlink: www .mpwb .org patients with chronic liver diseases and is emerging as a safe, reliable and non-invasive alternative to liver biopsy for staging MRA (MR angiography) hepatic fibrosis. (Magnetic Resonance) See Magnetic resonance angiography MRE is also being investigated for application to pathologies (MRA) of other organs including the brain, breast, blood vessels, heart, kidneys, lungs and skeletal muscle. MR elastography Related Articles: Phase contrast MR angiography (Magnetic Resonance) Magnetic resonance elastography (MRE) Further Readings: Mariappan, Yogesh K., Kevin J. Glaser and is a rapidly developing technology for quantitatively assessing the Richard L. Ehman. Magnetic resonance elastography: A review; mechanical properties of tissue. The efficiency of MRE as a diag- Muthupillai, R., D. J. Lomas, P. J. Rossman, J. F. Greenleaf, A. nostic method is based on the fact that the mechanical properties Manduca and R. L. Ehman. Magnetic resonance elastography by of tissues are affected by the presence of disease processes. direct visualization of propagating acoustic strain waves. MRE obtains information about the stiffness of tissue by assessing the propagation of mechanical waves through the tissue MRIgRT (MRI-guided radiotherapy) with a special magnetic resonance imaging (MRI) technique. The (Radiotherapy) See MRI-guided radiotherapy (MRIgRT) technique essentially involves three steps: MRI-guided radiotherapy (MRIgRT) • Generating shear waves in the tissue (Radiotherapy) Tumour localisation remains a significant chal- • Acquiring MR images depicting the propagation of the lenge in radiotherapy. Image-guided radiotherapy (IgRT) allows induced shear waves positioning of the patient based upon internal anatomy instead of • Processing the images of the shear waves to generate only external anatomy. quantitative maps of tissue stiffness, called elastograms. Typically, most linacs have either MV x-ray planar or kV x-ray (planar and CBCT) imaging capabilities. Whilst x-ray techniques A typical liver MRE exam begins by positioning a plastic dia- produce decent images for the purposes of IgRT, they also rely phragm (‘passive driver’) over the liver held in place by an elastic on ionising radiation, ultimately increasing the dose delivered to band. The passive driver is connected by a long hose to a pneu- the patient. matic pump (‘active driver’) outside the MRI room that inflates The key benefits of magnetic resonance imaging (MRI) over and deflates the diaphragm about 60 times per second. These x-ray imaging are the lack of imaging dose to the patient and mechanical impulses induce ‘seismic’ shear waves in the liver improved soft tissue contrast. MRI is now available on certain that can be detected with phase-sensitive MR sequences. Image radiotherapy linear accelerators (so called MR-linacs). MRI- acquisition is very fast, a single 2D slice obtained within a 20–25 guided radiotherapy (MRIgRT) enhances opportunities for adap- second breath-hold (Figure M.51). tive radiotherapy whereby the patient’s treatment is adjusted to As tissues become stiffer, waves propagate more rapidly and account for anatomical changes (Figure M.52). have longer wavelengths. This can be appreciated by observing There are some potential issues with MRIgRT. For example, the spacing between peak displacements in MRE wave images. any patients with ferrous implants cannot be imaged in an MR MR-Linac 624 MR microscopy M FIGURE M.52 Difference in soft tissue contrast MRI vs kV CBCT vs Planning CT (MDCT). (Acad Radiol. 2015 July; 22(7): 840–845.) field. Patients may feel claustrophobic in the bore of the MRLinac and the noise generated by the gradient coils may cause distress to some patients. The presence of a magnetic field also means that typical radiotherapy ancillary equipment may not be suitable (e.g. breast boards, motion management equipment and immobilisation equipment). Related Articles: MR-Linac MR-Linac (Radiotherapy) Unlike conventional linacs which combine MV treatments with kV imaging, an MR-linac combines magnetic resonance (MR) imaging and MV treatments to deliver image- guided radiotherapy with MR (MRIgRT). The benefits of MR imaging include excellent soft tissue contrast, no imaging dose to the patient and the potential for real-time imaging / 4D functional imaging. Using MR as the primary imaging method increases the scope of IGRT to accurately determine tumour position and intrafraction motion. This could potentially lead to improved adaptive radiotherapy and reductions in planning margins or dose escalation to the tumour target. The presence of an MR field, typically around 1.5T, means FIGURE M.53 The extent of the magnetic field in a typical MR linac. that the linac design has had to be altered in order to remove any This shows that the field can extend outside the bunker. (https :/ /do i .org /10 ferromagnetic materials. It is also necessary to account for the .1 016 /j .phro .201 7 .10 .0 02) perturbation of the electron beam in the presence of a magnetic field. MR linacs are unlike typical C arm linacs in that the treatment mm, or a little better with specialist sequences and/or equipment delivery is confined to an MRI bore. This means that certain tech- (typically 0.5 mm or above). niques such as non-conformal treatments or treatments with large However, outside the clinical arena, very high field strength isocentre shifts are not suitable for treatment on an MR linac. (>10 T), small bore MRI systems are available with high gradient MR safety in the clinic must also be addressed. MR question- field strengths, and in many non-clinical applications an imaging naires should be put in place to ensure patients are suitable for time of several hours is perfectly acceptable. This enables MR MR and any potential risks such as metallic implants or preg- microscopy – usually defined as imaging with a spatial resolution nancy are highlighted prior to the patient being exposed to the of better than 100 μm. MR field. The presence of the MR field also means that equipment The limit of resolution available to conventional MR micros- typically used in radiotherapy, such as immobilisation equipment, copy is about 1 μm. Thus the technique does not compete with needs to be made MR safe. Additionally, the MR field can extend other microscopy techniques, but rather complements them outside the bunker and interfere with equipment outside the bun- because of the capability of MRI to produce images dependent on ker. This can be seen in Figure M.53. a wide range of nuclear, physiological and chemical parameters. Related Articles: MRI only treatment planning, MRIgRT, Applications of MR microscopy have included botany, mate- IGRT, MR safety rials science and imaging of porous media. It is also possible to image animal models of relevance to medicine non-destructively, MR microscopy which has considerable potential for developmental biology and (Magnetic Resonance) The spatial resolution available to conven- genetics (Figure M.54). tional clinical magnetic resonance imaging (MRI) is limited by Recently, alternative approaches to MR microscopy have the amplitude of the switched magnetic field gradients and by the pushed the resolution limit further. Magnetic resonance force static magnetic field strength of the imaging system, which limits microscopy (MRFM) uses a sensitive cantilever to sense the force the signal to noise (SNR) achievable in an acceptable examination between a magnetic tip and individual spins within the sample. time. This combination of factors limits resolution to around 1 Resolution of around 90 nm has been achieved in this way. MR-only treatment planning 625 MR safety (new definitions) calculate the dose distribution inside the patient. The image qual- ity from CT is good, however it doesn’t have excellent soft tissue contrast, which in radiotherapy treatment planning can be crucial in order to accurately determine what is the target and what is nor- mal tissue. MR images have excellent soft tissue contrast which makes often contouring targets and OARs much easier than in M CT images. It is possible to calculate the radiation transport through, and dose deposition in, a CT data set as CT numbers are a func- tion of electron density. This electron density information is not available in an MR dataset, which is essentially a map of water molecules in the body. As MR imaging does not provide the Duke CIVM Duke CIVM electron density information required by modern treatment plan- (a) (b) ning algorithms, the direct calculation of dose on an MR data set is not possible. It is necessary then to somehow determine CT electron information from an MR data set. This is done via pseudo CT generation (see related article). Pseudo CTs allow patient plans, including dose calculations, to be created using only MR image data. Without the need for multiple imaging appointments, the patient can go from simulation to treatment more quickly. Related Articles: CT number, Pseudo CT Hyperlink: http: / /cli nical .netf orum. healt hcare .phil ips .c om / us _en /E xplor e /Whi te -Pa pers/ MRI /M R -onl y -sim ulati on -f o r -rad iothe rapy- plann ing Duke CIVM Duke CIVM MR safe (c) (d) (Magnetic Resonance) The term MR safe indicates that the device has been evaluated to show that when used in an MR environment, its use does not present additional risk to the patient. However MR safe does not assure that its use does not affect the quality of the diagnostic information. The term MR safe without specification of the MR environment to which the device has been tested should be avoided to prevent any incorrect interpretation. MR safety (new definitions) (Magnetic Resonance) The ASTM (2005) defines three classifica- tions of devices/objects from an MRI safety perspective: MR safe: an item containing no metallic parts and posing no Duke CIVM Duke CIVM known hazard in all MRI environments. (e) (f ) MR compatible: an item demonstrated to pose no known haz- ard in a specified MR environment. The ‘specified environment’ FIGURE M.54 Development of compound eye over a 16-day period defines the field strength, maximum gradient rate of change, SAR during metamorphosis of the tobacco hornworm caterpillar into a man- or other conditions under which the item has been tested. duca moth. (From Nijhout, F. and Gewalt, S., Duke Centre for in vivo MR unsafe: an item known to demonstrate hazards in all MRI microscopy, Durham, NC.) environments. These definitions have been adopted by the American College of Radiologists and by the MHRA in the United Kingdom. The Further Readings: Callaghan, P. T. 1993. Principles of definitions are designed to avoid the imprecise use of the terms Nuclear Magnetic Resonance Microscopy, Oxford University ‘MR compatible’ and ‘MR safe’. In particular the definitions Press, Oxford, U.K; Glover, P. and P. Mansfield. 2002. Limits to emphasise that the risk associated with a given device depends on magnetic resonance microscopy. Rep. Prog. Phys. 65:1489–1511; the MRI environment in which it is used: For example, a device Mamin, H. J. et al. 2007. Nuclear magnetic resonance imaging may be safe at 1.5 T but not at 3 T. ‘MR safe’ is a term appropri- with 90-nm resolution. Nat. Nanotech. Published online: April ate only for non-metallic items, safe under all conditions. Most 22, 2007; doi:10.1038/nnano.2007.105. devices designed for use in an MRI environment will be ‘MR compatible’, that is safe under defined conditions of use. The con- MR-only treatment planning ditions under which such devices are compatible must be stated on (Radiotherapy) Magnetic resonance- (MR-) only treatment plan- documentation accompanying the device. ning is when the primary imaging modality in the radiotherapy Devices used in the MRI environment should have labels planning stage is MR. Typically, patients will have a computed affixed following the convention shown in Figure M.55. tomography (CT) scan prior to undergoing radiotherapy, the aim Further Readings: ASTM International. 2005. Standard of which is to locate the treatment target, organs at risk and to practice for marking medical devices and other item for safety MRI safety 626 MRI safety MRI safety begins at system commissioning, where room layout should be designed to minimise the possibility of unauthorised access to the scan room. Warning signage

must be provided on doors leading to the unit. Rooms should incorporate oxygen alarms to monitor oxygen levels. In the event of a low oxygen M alarm, the alarm may be used to trigger an extract fan to remove helium from the room. MRI systems on the market must conform to IEC require- FIGURE M.55 Labels used for MRI safe, conditional and unsafe ments. The IEC defines three levels of MRI operation: normal devices (green, yellow and red labels respectively). mode, first level and second level (or experimental mode). The maximum static field, SAR and gradient field switching rates for each of these levels are defined. in the MRI environment. F2503-05; Kanal, E. et al. 2007. Normal mode operation will not result in whole body tem- ACR Guidance document for safe MR practices 2007. Am. J. perature rises of greater than 0.5°C and carries a low risk of Roentgenol. 188:1–27; MHRA 2007. Device bulletin: Safety peripheral nerve stimulation. In first level mode one or more of guidelines for magnetic resonance imaging equipment in clinical the MRI outputs may lead to physiological stress. Temperature use. DB2007(03). rises are restricted to less than 1°C and there is an increased risk of peripheral nerve stimulation. The MRI system will pres- MRI safety ent a warning to the user where the parameters of the examina- (Magnetic Resonance) MRI safety is concerned with the good tion result in first level operation. Improved patient monitoring management of MRI units and MRI examinations in order to may be required in first level examinations and some patient reduce the risks associated with MRI scanning. Guidance on the groups (e.g. pregnant patients) may be restricted to normal safety issues to be considered in MRI is available from a number mode exams. of sources including the ACR and the MHRA. The main safety On second level mode, one or more of the system outputs may concerns are as follows: result in a significant risk to the patient. Second level is for use only in research under strictly controlled conditions and with eth- 1. Projectile effect (see Projectile effect). ics committee approval. In routine use second level mode is not 2. Peripheral nerve stimulation (see Peripheral nerve accessible to the system user. stimulation). MRI units will have written local rules to help reduce risks. 3. Incompatible implants: Incompatible passive implants The rules will address the following: (e.g. ferromagnetic aneurysm clips) risk moving due to the strong static field of the scanner, with the potential 1. Definition of roles and responsibilities of personnel for death or injury. Implants may heat and cause local with respect to MRI safety. burns in response to the RF field. Incompatible active 2. Definition of the scan room and rooms accessing the implants such as cardiac pacemakers risk failure due to scan room as controlled areas. The ACR set out a interaction with the static, gradient and RF fields. model, multi-level zonal definition for MRI suites, list- 4. Skin burns: Interaction of the RF field with conductive ing appropriate levels of vigilance for each zone. As materials on or near the skin can induce RF heating and a basic requirement, entry to the scan room must be skin burns. For example, ECG electrodes, foil patches strictly controlled and limited to screened patients and and tattoos containing conductive materials are known personnel. to heat during RF procedures. Conductive loops formed 3. Patient screening procedures: All patients must be by incorrect positioning of the arms and legs are also screened to rule out the possibility of a contraindication known to be sufficient to cause burns. to MRI. Patients are required to complete a question- 5. RF deposition: Radiofrequency heating raises body naire on their medical history prior to scanning. Those temperature and can induce heat stress in some vulnera- with implants incompatible with the MRI procedure ble patients with reduced thermoregulatory control. RF or with metal fragments in their eyes will be excluded. energy deposition is measured in terms of SAR (specific Published data are available in commonly used implants absorption rate, W/kg). and the conditions under which they are compatible. 6. Acoustic noise: The rapid switching of the gradients Exclusion criteria for the use of contrast agents based coils sitting within the static field causes the coils to on current practice should be defined locally. vibrate and generate acoustic noise. Unprotected expo- 4. Screening of objects: No object of unknown MRI safety sure to this noise is sufficient to cause hearing damage. status may be allowed to enter the scan room. Items 7. Contrast agents: Some Gadolinium-based MRI contrast which are known to have no metallic components may agents have been implicated in inducing NSF (nephro- be safely brought into the scan room. Items tested and genic systemic fibrosis) in patients with impaired renal defined as MRI Conditional may be used within the function. scan room under defined conditions (see MRI safety, 8. Helium release: During a quench, all helium gas should new definitions). be expelled to the atmosphere via an extractor and pipe 5. Staff training: staff must be familiar with the risks system. However, in the event of a partial or complete associated with MRI and with patient set up procedures quench pipe failure some helium may be released into to reduce the risk of RF burns and heating, the provision the scan room. While not poisonous, helium is asphyxi- of hearing protection and with room evacuation proce- ating and in sufficient quantities will displace air from dures. Staff must also be familiar with the concepts of the room. normal, first level and second level system operation. MRI transparent 627 MRS voxel contamination Related Articles: MRI safety new definitions, Projectile value of a spectrum, and is usually neither immediately obvious effect, Peripheral nerve stimulation from inspection of the spectrum, nor readily correctable. Further Readings: IEC. 2010. 60601-2-33 ed.3. Particular Contamination may be broadly divided into ‘profile contami- requirements for the safety of magnetic resonance equipment for nation’, due to imperfections in the sensitivity profile of the VOI, medical diagnosis; Kanal, E. et al. 2007. ACR guidance docu- and ‘background contamination’, due to imperfect signal suppres- ment for safe MR practices 2007. Am. J. Roentgenol. 188:1–27; sion outside the VOI. M MHRA. 2007. Device bulletin: Safety guidelines for magnetic Profile contamination arises from imperfections in the slice resonance imaging equipment in clinical use. DB2007(03); profiles generated by radiofrequency pulses in the MRS pulse Shellock, F. 2009. The Reference Manual for MR Safety, Implants sequence, together with other hardware and sequence imper- and Devices, Biomedical Research Publishing, Los Angeles, CA. fections and relaxation effects. The profile typically consists of a maximum sensitivity plateau at the centre of the VOI falling MRI transparent gradually to zero sensitivity at some distance outside the nomi- (Magnetic Resonance) A material or device is ‘MRI transparent’ nal VOI. On commercial MR systems, the definition of the VOI if it does not generate detectable RF signal in an MRI scan. The size relative to this profile, and hence the degree of contamina- transparent object will appear as a signal void in an MRI image. tion arising from signal within the profile but outside the nominal Plastics and other solid non-metallic objects that do not contain VOI, is a matter for the system manufacturer (Figure M.56). mobile protons are typically transparent to conventional clinical Background contamination may occur in single-voxel MRS MRI systems. sequences in which the pulse sequence affects magnetisation in parts of the subject outside the VOI. Depending on pulse sequence details, and in some cases hardware factors, this may result in MRS (magnetic resonance spectroscopy) generation of contaminating signal from outside the VOI due, for (Magnetic Resonance) See Magnetic resonance spectroscopy example, to T1 relaxation or imperfections in signal subtraction. (MRS) Background contamination is often addressed by adding outer volume suppression (OVS) techniques to the pulse sequence. MRS voxel contamination Of the popular single voxel techniques, STEAM and PRESS, (Magnetic Resonance) The aim of single voxel localisation in for proton MRS, are single-shot techniques with little potential MRS is to obtain a spectrum that originates from within a well- for background contamination as long as the spoiler gradients are defined volume-of-interest (VOI) without contamination with sig- effective. Profile contamination is minimal, but tends to be worse nal from surrounding tissues. In practice, instrumental and pulse with PRESS than with STEAM, as it is more difficult to design sequence limitations mean that there are a number of mechanisms 180° refocusing pulses with clean selection. ISIS, for phospho- whereby a spectrum may be contaminated with extraneous sig- rus MRS, is prone to serious background contamination due to nal. Severe contamination can impact dramatically on the clinical imperfect signal subtraction, exacerbated by a mechanism known Signal loss due to Signal intensity/ incomplete refocusing arbitrary units and/or inversion and/or 700 relaxation effects, etc. 600 500 400 Profile-dependent 300 signal loss and contamination 200 100 0 15 12 10 10 8 Distance/mm 5 6 4 2 0 0 Background contamination due to incomplete suppression, subtraction and/or spoiling and relaxation effects FIGURE M.56 Typical MRS VOI profile, showing origins of profile and background contamination. ms selector 628 Multi-coil transmit as T1 smearing (see article ISIS for details), but generally shows In the pulse height analysis mode, pulses from the detec- little profile contamination. tor are sorted and stored according to their amplitude (height), Related Articles: ISIS, PRESS, Single voxel spectroscopy, enabling the pulse height spectrum to be displayed. This is STEAM achieved by feeding the pulses into an analogue-digital con- Further Reading: Keevil, S. F. 2006. Spatial localization verter (ADC) which for each pulse generates a number (in M in nuclear magnetic resonance spectroscopy. Phys. Med. Biol. binary form) proportional to the height of the pulse. This num- 51:R579–R636. ber is then used as the address of a memory location in RAM, so that each location or ‘channel’ records the number of pulses of a ms selector specific pulse height. (Diagnostic Radiology) The ms (millisecond) selector is part of In the multiscaler mode of operation, the time spectrum of the controls of the HVG of an x-ray equipment. It controls the the pulses from a radiation detector is displayed, that is a time- timer, allowing the radiographer to select the length of the expo- activity plot. sure (in milliseconds). When this time elapses the timer interrupts In nuclear medicine the MCA is often used together with a the exposure. NAI(Tl)-scintillation detector for in vivo measurements of the Related Article: High voltage generator activity uptake in different organs, for example thyroid uptake measurements or as an intra-operative probe. It is also used for measurements of radioactive samples in vitro. MSAD The MCA is also a valuable instrument together with a GeLi- (Radiotherapy) See Multiple scan average dose (MSAD) semiconductor detector for quality control of radiopharmaceuti- cals, that is control of the radionuclide purity. Two examples of MSCT (Multislice CT) isotope impurities are 110In or 114Inm in 111In-solutions and 202Tl (Diagnostic Radiology) See Multislice CT scanner in 201Tl. The MCA is sometimes a built-in module in a detector sys- MT tem, for example scintillation cameras, or used as a stand-alone (Magnetic Resonance) See Magnetisation transfer contrast PC-based, software controlled PCI-bus MCA (Figure M.57). (MTC) Abbreviations: ACD = Analogue-digital converter, MCA = Multichannel analyser and SCA = Single channel analyser. MTC Related Articles: Analogue–digital converter, Single channel (Magnetic Resonance) See Magnetisation transfer contrast analyser, Quality control, Radionuclide purity (MTC) Further Reading: Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., New York. MTF (Diagnostic Radiology) See Modulation transfer function (MTF) Multi-coil transmit (Magnetic Resonance) Multi-coil transmit or ‘parallel transmit’ is a method to divide transmission of the RF excitation pulse across MTT separate coils, each of which produces part of the RF excitation (Magnetic Resonance) See Mean transit time (MTT) pulse. Adding together contributions from these separate coils produces the desired RF field transmitted to the subject. The Multi-band simultaneous multislice (SMS) imaging major advantages of multi-coil transmit are uniformity of RF (Magnetic Resonance) See Simultaneous multislice excitation excitation across the field of view, reduction of the signal dropout seen in the centre of the field of view due to the dielectric effect Multichannel analyser and reduction the specific absorption rate (SAR) transmitted to (Nuclear Medicine) A multichannel analyser (MCA) is the stan-

the tissue. Multi-coil transmit is especially useful in the setting dard device for collation and analysis of data from a radiation of ultra-high field MRI scanners (>3T) where SAR concerns limit detector, for example scintillation or semiconductor detector. The RF power deposition. MCA is used in either pulse height analysis (e.g. gamma spectros- Related Articles: Radiofrequency pulse, Specific absorption copy) or multiscaler mode (time-activity diagram). rate (SAR), Ultra-high field MRI Delay Linear gate ADC Memory Detector and Input gate electronics Single channel (open when analyser ADC is not Ch 4 Ch 3 busy) Ch 2 Timer Ch 1 Live time Display FIGURE M.57 Function of a multichannel analyser. Multicrystal scanners 629 Multidetector CT Multicrystal scanners (Nuclear Medicine) In PET imaging the use of thin crystals improves the centre FOV spatial resolution. But the spatial resolu- tion is degraded for peripheral sources because of crystal penetra- tion by oblique incident photons. This effect is referred to as the depth of interaction or parallax effect. One common way to com- M pensate for this effect is to use several layers of crystals instead of a homogenous crystal. The crystals in the different layers have different decay times and it is therefore possible to get the depth of interaction by analysing the pulse shape. It does not necessarily need to be different crystal types in each layer. For example, the decay time of Gd SiO:Ce (GSO) crystal can be altered by chang- ing the amount of Ce (Cerium) in the crystal. Further Reading: Yamamoto, S. and H. Ishibashi. 1998. A GSO depth of interaction detector for PET. IEEE Trans. Nucl. Sci. 45(2):1078–1082. Multi-criteria optimisation (MCO) (Radiotherapy) Multi-criteria optimisation defines an optimisa- FIGURE M.58 Single slice/detector CT (left) versus multidetector CT tion process which simultaneously considers not only the dose (right). distribution in the tumour, but also the dose to selected organs at risk (OARs). Constraints are assigned to the tumour and the OARs, and the optimisation algorithm optimises the plan tak- ing these objectives into account simultaneously. In conventional optimisation, weights are assigned to these individual constraints, leading to one solution of the optimisation procedure. In contrast, MCO does not produce a single plan, but a set of plans, for which one objective can only be improved if another objective worsens. These plans define the Pareto Surface (Thieke et al., 2004). The planner can then decide and chose one of the ‘Pareto optimal’ plans. Related Articles: Pareto surface Further Readings: Literature and mathematical description: Messac, A. et al. 2003. The normalized normal constraint method for generating the Pareto frontier. Struct. Multidiscip. Optim. 25:86–98; MCO for proton therapy: Chen, W. et al. 2010. A fast optimization algorithm for multicriteria intensity modulated pro- ton therapy planning. Med. Phys. 37(9): 4938-4945; Thieke, C. et al. 2004. Beyond weight factors: New concepts for defining and analysing dose optimisation. Radiother. Oncol. 73:S75–S75. Multidetector CT (Diagnostic Radiology) Multidetector CT (MDCT) scanners have been around since the late 1990s and most scanners, if not all, currently being used in the clinical practice are of this type. MDCT scanners involve the use of multiple detector rows/arrays – e.g. 4–256 detector rows in the longitudinal direction (i.e. along the length of the patient lying on the patient table). MDCT scan- ners implement the third generation CT geometry where the arc of the detector and the x-ray source rotate together. Figure M.58 FIGURE M.59 Example of a fixed array detector size (top) and adaptice indicates the difference between a single detector and multidetec- array detector (bottom). tor CT detector bank. Prior to MDCT scanners, slice thickness in single detector scanners was determined by a combination of pre-patient and post-patient collimation. an operator is also able to choose between various reconstructed Advantages of MDCT scanners include multislice acquisi- slice widths e.g. data acquired with a 4 × 2.5 mm configuration tions, faster acquisitions times, large Field of View (FOV) cov- can be reconstructed in 4 × 2.5 mm or 2 × 5 mm or 1 × 10 mm. erage, improved isotropic imaging, improved spatial resolution One should also note that the maximum number of slices acquired and dose efficiency. Operators are able to choose between dif- by a MDCT in a single rotation is not solely determined by the ferent beam collimations and detector row combinations and number of detector arrays but by the number of channels available hence acquired slice widths (e.g. 64 × 0.625 mm or 32 × 1.25 to transfer the acquired data for image reconstruction. mm). This means that larger slice thicknesses can be generated It should be noted that, depending on the MDCT manufac- by electronically combining the signal from several detector rows turer the physical width of each detector array may vary across and that only ‘active’ detector rows are exposed. Furthermore, the detector bank (see Figure M.59). In this type of MDCT the Multi-echo 630 M ultileaf collimator central detector arrays are thinner than the outer ones. This leads Abbreviations: MFO = Multifield optimisation, OAR = Organ to a cheaper design and more geometrically efficient. at Risk. Related Articles: Multislice CT scanner Related Articles: Single field optimisation (SFO) Multi-echo Multi-gated acquisition (MUGA) M (Magnetic Resonance) A multi-echo pulse sequence utilises several (Nuclear Medicine) Multi-gated acquisition (MUGA) is a com- 180° RF pulses to recall more than one echo after a 90° RF excita- posite study of cardiac contractility and wall motion. Image tion pulse. A common version recalls two echoes with two differ- acquisition is triggered in a position of interest in the cardiac cycle ent echo times, that are collected using the same phase encoding using the signal from an ECG (electrocardiogram). The examina- and put into separate raw-data sets, hence providing two images of tion is performed on patients with suspected coronary artery dis- different contrast. A common combination is a pulse sequence col- eases or congestive heart failure. The test can be performed when lecting one proton-density-weighted image, requiring short echo the patient is at rest or exercising (called MUGA stress test). A time and long repetition time, and one T2-weighted image requir- stress test is usually performed on patients with possible coronary ing a long echo time and a long repetition time, that is a so-called artery disease. One of the important parameters to monitor is the Pd-T2-protocol. The double contrast protocol can be based on either ejection fraction. conventional spin echoes or the faster version, fast spin echoes. Related Articles: Carr–Purcell (CP), Carr–Purcell–Meiboom– Multihole collimators for radioisotope scanners Gill (CPMG) Sequence, Fast spin echo, FSE, T2-weighted (Nuclear Medicine) A multihole collimator, often called multi- channel, honeycomb or focusing collimator, is constructed of Multifield optimisation (MFO) shielding material placed between the radiation detector and the (Proton Therapy) This is an optimisation technique which allows gamma emitting radioactive source. Several holes through the different beams to cover different sections of the target. The shielding allow radiation to impinge on the detector from a speci- optimiser uses contributions from different beams to achieve the fied spatial location only. The field of view of a multihole col- required dose constraints. limator is simply the superposition of all fields of view of each of Each field has a specific volume in which it can put spots. In the several holes at the focal plane they superimpose. Everywhere MFO, each field doesn’t need to cover the entire target: as long else the superposition is only partial, depending on the distance as at least one field covers every target area, the optimiser can from the focal plane. Tapered holes must be used to make the produce a reasonable distribution. maximum amount of crystal area available to the impinging pho- This technique is especially useful for a situation where the ton fluence. treatment volume is partially obstructed by an Organ at Risk The spatial resolution is the ability to distinguish between two (OAR). Using this technique allows another beam to cover the radioactive sources close together. The depth response depends obscured part of the target volume from a different angle, com- on the focal distance. The sensitivity is its ability to detect radia- pletely sparing the OAR. tion from the radioactive source. Related Articles: Radioisotope scanner, Collimator Further Reading: Myhill, J. and G. J. Hine. 1967. Multihole collimators for scanning. In: Instrumentation in Nuclear Medicine, Vol. 1, ed., G. Hine, Academic Press, New York, Chapter 17, pp. 429–460. Multihole focused collimators (Nuclear Medicine) See Diverging collimator Multileaf collimator (Radiotherapy) Multileaf collimators (MLC) provide one of the most sophisticated collimation methods in radiotherapy, con- forming the delivered beam closely to the target shape, thereby reducing the volume of normal tissue irradiated, and reducing the dose to critical structures near the boundary. They consist of 20–80 pairs of tungsten leaves arranged in two opposing banks, thick enough to provide similar attenuation to that of conven- tional secondary collimator jaws. Each leaf is normally 1-cm wide, and can move individually to create irregularly shaped fields, which conform closely to the shape of the projected plan- ning target volume (PTV), with accuracy on the order of mil- limetre. Some MLCs have micro-MLCs within the centre of the FOV that allow for finer collimation (e.g. width 0.5 cm cf 1 cm). See Figure M.60 for an example field shape allowed with MLCs. Although the light field may appear to have very jagged edges, Multi Field Optimisation: individual proton fields with non-uni- the radiation field edges are much smoother inside the patient form doses are optimised simultaneously (lower two distributions). because of the effect of scatter. The use of MLCs has largely They combine to achieve a uniform dose across the target (upper replaced the lengthy and laborious task of creating patient-spe- distribution). cific custom blocks. Multimodality systems 631 Multi-pinhole collimator The MLC banks can either be positioned externally in addi- Abbreviations: IMRT = Intensity-modulated radiotherapy, tion to the standard collimation system, or they can replace one MLC = Multileaf collimator, MSF = Multiple static field and PTV of the secondary collimator jaws. However backup collimation = Planning target volume. will normally be needed with MLCs due to the significant inter- Related Article: Intensity-modulated radiotherapy (IMRT) leaf leakage that occurs due to the necessary mechanical clear- ance between each leaf. This is minimised by using a tongue and Multimodality systems M groove design, which is illustrated in Figure M.61. The end of (Nuclear Medicine) A multimodality system refers to an imaging each leaf is tapered so that their edges are focused towards the system where two or more scanners with a single patient table target in order to minimise the penumbra irrespective of the angle share the same system of reference. Examples of dual-modality of incidence. The positional systems vary between manufacturers, systems are PET-CT and SPECT-CT-scanners. With such a system but include the use of potentiometers or an optical system with it is possible to sequentially acquire a functional and morphologi- reflectors on each leaf. cal image. CT scans can also be used for attenuation correction. MLCs and IMRT: MLCs are used in the delivery of intensity- When combining images from stand-alone scanners the patient modulated radiotherapy (IMRT) enabling the fluence of the beam has to move between the two scanners. It is not likely that the to be modulated over time and position. There are two meth- patient can assume an identical position in the second scan and ods, the multiple static field (MSF) technique, and the dynamic also the organ orientation differs slightly between the two scans. MLC technique. The MSF technique uses the MLCs to create a These sources of error are avoided when using a multimodality sequence of 2D-shaped fields, only moving the MLC when the system and can therefore lead to higher diagnostic accuracy. For beam is off. The dynamic MLC method uses continuously mov- example, using a PET/CT system for tumour localisation and ing leaves whilst the radiation is on. Varying the velocity of each diagnosis has improved the diagnostic accuracy by 48%–60%. It leaf pair can create any required 2D profile. has affected the clinical management decisions in 12%–27% com- MLC QA: QA considerations for MLCs include the following: pared to stand-alone PET and CT and has led to a modification in up to 63% of patient external radiation treatment plans. • Leaf position accuracy Further Reading: Wernick, M. N. and J. N. Aarsvold. 2004. •

Collimator leakage (tongue and groove leakage) Emission Tomography. The Fundamentals of PET and SPECT. • Leaf speed control Elsevier, London, UK, pp. 195–197. • Interlocks Multi-pinhole collimator (Nuclear Medicine) This collimator includes several pinhole apertures instead of only a single hole. A pinhole collimator is primarily used because it provides images with high spatial reso- lution. The pinhole collimator efficiency, however, depends on the size of the apertures and the source to collimator distance. At distances smaller than ∼60 mm the pinhole efficiency (2 mm apparatus opening) is often better or equal than the efficiency of a parallel-hole collimator. In vivo distributed sources are often located further away from the collimator than 60 mm, thus giving a low efficiency for clinical studies. Low efficiency means long acquisition times and low signal to noise ratio. A way to increase the efficiency is to use multiple pinhole col- limators since multiple pinholes allow simultaneous acquisition of different projections. The increase in efficiency then is a factor of the number of pinhole apertures but also of the crystal size and thickness. When the number of pinholes increases there is also FIGURE M.60 Typical field shape achievable with MLC. an increased probability to acquire projections that overlap which FIGURE M.61 Schematic to show the use of stepped and overlapped leaf sections to minimise inter-leaf leakage, known as the tongue and groove design. Multiplanar reconstruction (MPR) 632 Multiple-image radiography M FIGURE M.62 Postmortem MIR images of a mouse thorax. Left: attenuation (or apparent absorption) image. Center: refraction angle image. Right: USAXS scattering map. Images are obtained through momenta analysis from 5 positions on the rocking curve. (Courtesy of L. Rigon, F. Arfelli and R.H. Menk.) may hamper an accurate image reconstruction. A proper collima- the rocking curve in the analyser-based imaging (ABI) technique. tor design therefore depends on a good optimisation between these Conventional radiography depicts only one object parameter, two factors. i.e. x-ray attenuation, which is eventually degraded by the effect Related Article: Collimator design of scattering, potentially obscuring image details of interest. Further Reading: Ivanovic, M., D. A. Weber and S. Loncaric. Conversely, taking advantage of the high angular sensitivity of 1999. Multi-pinhole collimator optimization for high resolution ABI, MIR is capable of providing a more comprehensive descrip- SPECT imaging. Nucl. Sci. Symp. 2:1097–1101. tion of the object in terms of attenuation (with scattering rejec- tion), refraction and ultra-small-angle scattering (USAXS). Multiplanar reconstruction (MPR) Starting from multiple (at least three) input images, MIR (Magnetic Resonance) A magnetic field can be established in any method produces three parametric images (see Figure M.62): direction in space by applying currents of the appropriate ampli- tudes to two or three of the gradient coils simultaneously. Thus it • Attenuation (also referred to as apparent absorption) is possible to collect images in any plane, as long as the net slice image is similar to the conventional radiography but selection, frequency encoding and phase encoding gradients pro- features a greater contrast due to scattering rejection duced in this way remain mutually orthogonal. performed by the analyser crystal. Nevertheless, it is sometimes useful to be able to reformat • Refraction image maps the angular deflections in the image data acquired in a particular plane for viewing in a differ- microradians range due to the refractive index varia- ent plane. This process of displaying images in different planes is tions within the sample. known as multiplanar reconstruction. • USAXS image quantifies the angular dispersion (or Where, as is often the case, two-dimensional images have divergence) of the beam caused by structures or textural been collected using a slice thickness considerably greater than properties in a spatial scale smaller than the pixel size, the in-plane resolution, reconstruction in a different plane can but much larger than the radiation wavelength. result in stair-step artefacts. This is usually less apparent with three-dimensional imaging, where there is in principle no pre- The practical implementation of MIR usually relies on the pixel- ferred plane and voxels are more likely to be isotropic. wise comparison between the rocking curves obtained without Related Article: Oblique imaging and with the presence of the sample: The reduction of the rocking curve integral due to the presence of the sample accounts for its Multiple beams attenuation, the relative shift of the centroid is due to refraction, (Radiotherapy) Multiple beams are used in conformal radiother- while the widening of the curve encodes the USAXS signal. In sta- apy, in order to conform the dose distribution as closely as pos- tistical terms, this means to measure the zeroth, first and second sible to the target volume. momenta of the rocking curve both with and without the sample, Related Article: Conformal radiotherapy which requires multiple input images to be acquired and the rock- ing curve to be analytically modelled or fit. Several approaches Multiple coulomb scattering have been introduced to tackle this problem, some of them make (Radiation Protection) When charged particles in a beam of ion- use of curve fitting using Gaussian, Voigtian and Pearson type ising radiation are incident on an absorbing medium they will VII functions, while others rely on an iterative approach using undergo not just one, but a series of scattering interactions before maximum-likelihood expectation-maximisation. The accuracy of all the energy in the incident particles has finally been absorbed. each method depends on how well the rocking curve can be mod- These scattering events are due to the interaction between the elled and on the noise content of the input images. charge of the particles, and the electrostatic (Coulombic) forces Related Articles: Analyser-based imaging, Diffraction- present in the nuclei of the atoms of the medium. These interac- enhanced imaging, Phase-contrast imaging tions are also known as Rutherford scattering. Further Readings: Arfelli, F., A. Astolfo, L. Rigon and R. H. Menk. 2018. A Gaussian extension for diffraction enhanced imag- Multiple-image radiography ing. Sci. Rep. 8(1):1–14; Kitchen, M. J. et al. 2010. X-ray phase, (Diagnostic Radiology) Multiple-image radiography (MIR) is absorption and scatter retrieval using two or more phase contrast based on the computation of multiple parametric images of the images. Opt. Express 18(19):19994–20012; Pelliccia, D., M. J. sample from multiple images acquired at different positions of Kitchen and K. S. Morgan. 2017. Theory of X-ray phase-contrast Multiple isocentre treatment 633 Multislice imaging. In Russo P. (ed.) Handbook of X-ray Imaging: Physics I > T and Technology, CRC Press, pp. 971–998; Wernick, M. N. et al. or Pitchx >1 2003. Multiple-image radiography. Phys. Med. Biol. 48(23):3875. Dmax Multiple isocentre treatment MSAD (Radiotherapy) In most radiotherapy treatments, the centre of the M tumour (target) is positioned at the isocentre. However, in some cases more complex target volumes or multiple targets located close together necessitate consideration of overlapping dose dis- z-axis tributions across the entire volume using more than one isocentre location. FIGURE M.64 MSAD for case where couch increment is greater than This multiple isocentre technique is used most often with the slice width. gamma knife treatment system. This uses a number of cobalt sources arranged in a hemisphere which are focused using a collimator helmet onto a central target and will produce small, The MSAD for a scan length of 100 mm is equal to CTDI100 spherical, high dose volumes. By combining these small radiation (the commonly used metric in CT dosimetry) when the couch spheres, positioned at multiple isocentres, it is possible to conform increment is equal to the nominal slice width, or, in helical scan- to the shape of larger and non-spherical tumours. ning, when the pitch value is equal to unity. For other cases, the CTDI must be corrected for couch increment or pitch to obtain the Multiple reflection MSAD (see the following): (Ultrasound) See Reverberation inaxialscanning MSAD = × éT ù CTDI Multiple scan average dose (MSAD) ëê I ûú (Diagnostic Radiology) The multiple scan average dose (MSAD) is used in computed tomography (CT). It is a measure of aver- é 1 ù in helicalscanning MSAD = CTDI × ê ú age absorbed dose (mGy) to the irradiated area from a series of ë Pitchx û slices. It is a consequence of the fact that the dose profile of each CT scan is not perfect (square), but is bell-shaped (for details see Related Articles: Dose profile, CTDI the articles on Dose profile). The need of MSAD manifests itself when a sequence of adjacent CT scans is made. For cases where Multiple scattering the couch increment is equal or less than the nominal slice width (Radiation Protection) A proportion of the photons or charged the bell-shaped edges of the dose profiles of each CT scan over- particles in a beam of ionising radiation incident on an absorbing lap, and result in the average (summed) dose being higher than medium will undergo not just one, but a series of scattering inter- the dose in a single CT scan. (Figure M.63). Obviously the higher actions before all the energy in the incident radiation has finally the spread of the bell shape of the dose profile, the higher the been absorbed. Such multiple scattering events can spread the MSAD. energy absorption some distance away from the primary radiation For cases where the couch increment is greater than the slice beam, having an implication for radiotherapy treatment planning, width, the MSAD is reduced and may be lower than the dose from etc., where the goal is to deliver a prescribed radiation dose to the a single scan (Figure M.64). target volume (the tumour) whilst minimising the radiation dose The formula used to calculate MSAD is to surrounding healthy tissues. Related Articles: Elastic scatter, Inelastic scatter +¥ 1 MSAD = (z) z I òD d -¥ Multislice (Magnetic Resonance) Acquisition of a single-slice image through where a body structure of interest is rarely sufficient for diagnostic pur- D(z) is the dose profile along the z-axis (scanner axis of poses. Instead, it is usual to acquire either a stack of slices or a rotation) volume of data. I is the scan increment between the adjacent slices Because of the nature of the data acquisition process in MRI, acquisition of N slices does not normally entail sequential rep- etition of the acquisition process N times. The need to allow a repetition time (TR) for recovery of longitudinal magnetisation Couch increment (I ) = Slice width (T ) or between consecutive excitations of a given slice fortuitously pro- Pitchx = 1 vides a ‘dead time’ during which other slices may be interrogated. MSAD (mGy) Depending on the TR used (i.e. the desired degree of T1 weighting) Dmax it may be possible to image a considerable number of slices with- T out an increase in the overall imaging time. A significant complication arises because the profile of an excited slice in MRI never corresponds exactly to the desired ‘top z-axis hat’ function, and there is inevitably some excitation of material lying outside the ideal slice. As a result, if consecutively excited FIGURE M.63 MSAD for case where couch increment equals slice slices (e.g. A and B in Figure M.65) are contiguous, then material width. lying close to the boundary between the slices will be repeatedly Multislice 634 Multislice CT scanner A TR A 2D M A B C Detector rings Septa 3D A B C A B C FIGURE M.65 Acquisition of signals from multiple slices within a single TR interval. FIGURE M.66 Two different acquisition modes, 2D (upper) and 3D (lower). When operating in a 3D mode the scanner has higher sensitivity excited at very short intervals. This results in heavy T1 weighting than when running in a 2D mode. of this material and severe distortion of the slice profile. There are two possible solutions to this problem, other than improving slice profiles by computational modelling. A common Further Reading: Cherry, S. R., J. A. Sorenson and M. E. approach is to leave a small gap between slices, often 10% of the Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, slice thickness, to allow for profile imperfections. This has the Philadelphia, PA. disadvantage of possibly missing clinically significant informa- tion lying in the slice gap. Alternatively, two interleaved acquisi- Multislice CT scanner tions may be performed, acquiring odd and even numbered slices (Diagnostic Radiology) For many years the detection system on respectively. This restores slice contiguity, but requires a doubling CT scanners consisted

of a single arc of detectors (Figure M.67a) of the overall imaging time. and so acquired one slice of data in a gantry rotation. In 1991 a Multislice imaging is more complicated when sequences are scanner with two parallel banks of detectors, capable of acquir- used that acquire multiple lines of k-space from the same slice ing two slices simultaneously was launched. This paved the way following a single excitation (e.g. fast spin echo). In such cases for modern multislice systems, and in 1998 the first four-slice the echo train length and the number of slices imaged in a single scanner with multiple parallel banks of detectors (Figure M.67b) acquisition may need to be traded off to ensure optimal use of the was announced. This was followed by 8-slice scanners in 2001, TR interval. 16-slice scanners in 2002, 64-slice scanners in 2004, and in Related Articles: Fast spin echo (FSE), Interleaving, Slice 2007, a scanner capable of acquiring 320 slices per rotation selection (Figure M.68). The increased length per rotation covered by these scanners, coupled with faster gantry rotation speeds, allows Multislice the scanned volume to be covered very quickly, thereby largely (Nuclear Medicine) Multislice is a camera configuration with sev- eral detector rings (or slices). Each detector ring often consists of several individual camera elements, each coupled to a signal multiplier, for example a PM tube. Since most conventional scintillation cameras have individ- ual camera heads instead of a set of rings, multislice cameras in nuclear medicine are often referred to as PET cameras. A multislice PET system can be operated in either a two- or three- dimensional mode (2D and 3D mode respectively). 2D mode involves using a separating high attenuating material between each ring, that is septa that attenuates radiation with oblique angle of incidence. When operated in a 2D mode, the PET camera will only accept annihilation photons with a near-perpendicular direc- tion relative to the detector ring surface, that is both photons must be registered in the same ring (see Figure M.66). Not all anni- hilation photons have a near-perpendicular incident angle and are therefore attenuated by the septa or simply discriminated by the coincidence processor. In a 3D system there are no septa and z-axis z-axis coincidences can be registered between all rings; hence a higher (a) (b) sensitivity. For further details about PET acquisition, see Related Article. FIGURE M.67 (a) Single-slice scanner and (b) multislice scanner (not Related Article: Data acquisition PET to scale). (Courtesy of ImPACT, UK, www .impactscan .org) Multi-vendor service 635 Multi-vendor service removing motion artefacts. On these systems the scan length is long scan times or excessive tube loading. Narrow slices result no longer limited by factors such as the patient breath-hold time. in improved z-axis spatial resolution and provide isotropic res- On single-slice scanners the acquired slice width can be varied olution images, which provide images of equal quality when between 1 and 10 mm by varying the z-axis x-ray beam collima- reconstructed in any plane (Figure M.70a). They also enable high tion. On multislice systems the width of the acquired slice, some- quality 3D reconstructions (Figure M.70b). times referred to as the ‘data acquisition width’, is governed by Multislice CT has extended the range of clinical examinations M the z-axis detector dimension. Data from multiple adjacent chan- possible with CT scanners, which has led to concerns about the nels can be combined and this results in an increase in the data radiation dose burden from these scanners, so CT examinations acquisition width. An example of a four-slice scanner is shown must always be justified. And although scanning long length with in Figure M.69. The scanner has 16 parallel, 1.25 mm detector high tube currents is possible, effort must be made towards opti- banks, but can only acquire a maximum of four data channels misation of scan parameters. simultaneously. For z-axis beam collimations in excess of 5 mm, Related Articles: Computed tomography, Helical scanning, the signals from adjacent detector banks are combined, result- Slice thickness, Image artefact ing in an increased data acquisition width. Scanners capable of acquiring a high number of data channels per rotation, such as 64-slice scanners, do not usually require this combining of sig- Multi-vendor service nals, and can therefore always acquire data with narrow acquisi- (General) Independent service organisations (ISOs) are special- tion widths, even for the widest beam collimations. ised companies that provide technical services across a wide In addition to the increased coverage per gantry rotation, the range of brands and models of medical equipment, often offering second major advantage of multislice scanners is that they can full coverage on the whole inventory of equipment of a hospital. acquire a wide coverage of data, at widths as narrow as 0.5 mm. In a third-party multi-vendor service model, the healthcare This capability means that narrow slice acquisition can be used facility stipulates a contract with an ISO to service the whole for most examination types as it no longer results in unacceptably inventory of medical equipment. These contracts are often referred to as full-risk contracts as they cover any kind of techni- cal intervention for a fixed yearly fee (thus providing a fixed and well-known yearly cost). Based on the contractual clauses defined they can include preventive maintenance, safety tests, corrective maintenance, installation and relocation of equipment, user train- ing, etc. ISOs can arrange a dedicated staff of technicians inside a lab within the hospital, or they can use technicians that travel across different healthcare facilities. When the ISO establishes an onsite presence of technicians, several advantages can be achieved: very fast interventions, immediate availability of spare parts/tools and test equipment, good knowledge of the hospital inventory and physical structure, etc. On the other side, the multi-vendor service has potential disadvantages: manufacturers may be reluc- Single slice 4 and 16 slice 64 slice 320 slice tant provide training and spare parts, ISOs may only be able to 10 mm 20 mm 40 mm 160 mm cover certain devices, the quality of interventions must be closely monitored. FIGURE M.68 Multislice scanners with increased coverage in one gan- In any case, and especially if the technical services are fully try rotation. (Courtesy of ImPACT, UK, www .impactscan .org) outsourced to a single ISO, it’s important that the healthcare Irradiated detectors X-ray beam collimation 1.25 mm: 2 × 0.63 mm 5 mm: 4 × 1.25 mm 10 mm: 4 × 2.5 mm 15 mm: 4 × 3.75 mm 20 mm: 4 × 5 mm Data channels from grouped 16 × 1.25 mm detectors detectors z-axis FIGURE M.69 Example of four-slice scanner showing data signals from different groupings of detector banks. Mu-metal 636 Myo-inositol facility establishes rules and tools to monitor the activities per- blood supply to the cardiac muscle (the myocardium), and can formed by the ISO and the key performance indicators achieved. visualise regional defects in the perfusion, caused by a stenosis in the supplying coronary vessel, or an infarction. Such perfusion Mu-metal defects can demonstrate delayed perfusion, less perfusion, or a (Diagnostic Radiology) Mu-metal is a special alloy used for mag- combination of these, in comparison to the well-perfused myo- M netic shielding. It is a nickel–iron alloy (75% nickel, 15% iron, cardium. The perfusion is often evaluated in both rest and stress, plus copper and molybdenum) with very high magnetic perme- where the stress can be either mechanical or pharmaceutical. In ability. One specific use of this alloy is the envelope of Image the clinical setting, assessment of myocardial perfusion defects is Intensifiers in fluoroscopy, where the precise internal electromag- so far qualitative. netic filed requires good shielding from external electromagnetic Myocardial perfusion is measured by single photon emission influences. computed tomography (SPECT) and positron emission tomog- raphy (PET) with good results, but these modalities expose the Muscle patient to ionising radiation. In MRI, myocardial perfusion is (General) Muscle is around 70% water, with the remaining 30% assessed by injecting a small Gd-bolus and imaging as the bolus composed mostly of protein in the form of actin and myosin. passes the heart, the so-called first-pass perfusion. The imag- Muscles can be classified into three groups: skeletal, smooth or ing sequence uses the difference in T1 introduced by the pass- cardiac. Skeletal muscle is attached to bone and is responsible ing Gd-bolus, and T1-sensitivity is achieved by using saturation for skeletal movements and posture. Smooth muscle exists within recovery. To be able to resolve the fast course of events, a read- organ walls and controls involuntary movements such as peri- out with adequate time resolution is necessary. Fast gradient echo stalsis in the oesophagus. Cardiac muscle is found in the heart. (GRE), GRE-EPI or balanced GRE readouts are used. 2–4 slices Skeletal and cardiac muscles are striated: the muscle fibres are are acquired, and all slices should be acquired in each heartbeat. divided into sections or sarcomeres, where thin filaments formed The imaging sequence takes about a minute, and for good image by actin, troponin and tropomyosin overlap to regularly varying quality, the patient should maintain breath-hold for as long as pos- degrees with thick filaments formed of myosin. During muscle sible, and at least during the bolus passage. contraction, the thick and thin filaments slide past each other to Related Articles: Dynamic susceptibility contrast MRI, shorten the muscle fibre. Perfusion imaging Properties of Muscle-Based Tissues for Medical Imaging: Muscle can be distinguished on x-ray images, although contrast Myo-inositol with other soft tissues is generally poor. Muscle has a CT num- (Magnetic Resonance) Myo-inositol is a chemical compound ber of between +10 and +40 Hounsfield Units. The distinction of that features in in vivo proton (1H) NMR spectra of the brain muscle from other soft tissues is much better in magnetic reso- (Figure M.71). nance imaging. Muscle has T1 and T2 relaxation times of around 1075 and 33 ms respectively, at a magnetic field strength of 1.5 T. Muscle has a mass attenuation coefficient of 0.0227 m2/kg1 for 50 HO OH keV photons. Related Article: Tissue contrast HO OH Myocardial perfusion imaging HO OH (Magnetic Resonance) Perfusion is an assessment of the capillary blood flow through a tissue. Myocardial perfusion scans show the FIGURE M.71 Molecular structure of myo-inositol. (a) (b) FIGURE M.70 (a) Sagittal multi-planar reconstruction of ankle bones. (Courtesy of Siemens Medical Systems.) (b) 3D volume rendered image of kidneys and associated vasculature. (Courtesy of GE Healthcare.) Myo-inositol 637 M yo-inositol The main myo-inositol resonance occurs at 3.54 ppm, and at magnetic field strengths higher than those used in clinical MRI 0.8 this is resolved into a complex multiplet. There are also smaller Myo-inositol resonances at 3.28, 3.60 and 4.05 ppm. (protons 1 and 3), Myo-inositol has a number of biological functions, mostly 3.54 ppm 0.6 related to cell signalling. High levels of myo-inositol are found M in neonatal brain spectra, and in adulthood levels are elevated in Alzheimer’s disease and reduced in hepatic encephalopa- 0.4 thy. It is widely regarded as a marker of astrocytes, as changes in concentration are also seen in certain types of tumour (Figure M.72). 0.2 0.0 ppm 4 3 2 1 FIGURE M.72 1H NMR spectrum of the human brain showing main myo-Ino resonance. N N/2 artefact further de-excitation. The delayed component causes a number (Magnetic Resonance) The N/2 artefact is another term for the of low signal amplitude events which is generally (at low and Nyquist ghosting artefact found only for the EPI pulse sequence. medium count rates) no problem since they are discriminated by N Abbreviation: EPI = Echo planar imaging. the detection electronics due to low signal amplitude. There are Related Articles: Ghost artefact, Nyquist ghosting applications in which single-electron sensitivity is needed and the result is degraded by the phosphorescence pulses. At high count N-acetylaspartate (NAA) rates the phosphorescence tends to build up when a large number (Magnetic Resonance) N-acetylaspartate (NAA) is a chemical of phosphorescence pulses overlap. compound that features in proton (1H) NMR spectra of the brain. Related Articles: Inorganic scintillators, Scintillators The methyl proton resonance from NAA at 2.01 ppm is the most Further Reading: Knoll, G. F. 2000. Radiation Detection prominent peak in a water-suppressed spectrum, and a smaller and Measurement, 3rd edn., John Wiley & Sons, Hoboken, NJ, multiplet resonance at 2.6 ppm due to −CH2 protons may also be pp. 234–238. visible (Figure

N.1). In the adult human brain, NAA is found only in neuronal Nanoparticles bodies. Its role is poorly understood, but most likely it has an (Nuclear Medicine) Nanoparticle technology is a new field that is osmoregulatory function. Despite this uncertainty, NAA has an being widely investigated in nuclear medicine for the in vivo deliv- important role in MR spectroscopy, since it is regarded as a neu- ery of radiolabelled nanoparticles to cancer cells. The approach ronal marker, and its depletion is thought to be indicative of a loss uses nanoparticles of various materials such as liposomes, iron of functional neurons. For example, depletion of the NAA peak at oxide, nanotubes, perfluorocarbon, etc. with a size between 100 2.01 ppm is found in degenerative disorders such as Alzheimer’s to 800 nm. The size of these particles is about 100 to 10,000 times disease, where this change may occur before cerebral atrophy smaller than human cells. A significant advantage of the nanopar- becomes visible on MRI. In temporal lobe epilepsy, reduced NAA ticle approach is the reduction of systemic toxicity to the patient can help to lateralise the disease in the absence of MRI findings by allowing specific targetting of the cancer cells. Additionally, – although there is also a contralateral NAA depletion that may the nanoparticles can be used for evaluating the efficacy of drug resolve following surgery to the ipsilateral focus, indicating that delivery. Nanoparticles could play a significant role in improving reduction in NAA is associated with poor neuronal function, cancer therapy in the coming years. rather than simply a loss of neurons (Figure N.2). Related Articles: Radionuclides in therapy, Radionuclide Related Article: Magnetic resonance spectroscopy imaging Further Readings: Cady, E. B. 1990. Clinical Magnetic Further Readings: Hong, H., Y. Zhang, J. Sun et al. 2009. Resonance Spectroscopy, Plenum, New York; De Certaines, J. D. Molecular imaging and therapy of cancer with radiolabeled et al., eds. 1992. Magnetic Resonance Spectroscopy in Biology nanoparticles. Nanotoday 4(5):399–413. doi:10.1016/j.nan- and Medicine, Pergamon, Oxford, UK. tod.2009. 07.001; Huang, X., I. H. El-Sayed, W. Qian et al. 2006. Cancer cell imaging and photothermal therapy in the NaI(Tl) detector crystal near-infrared region by using gold nanorods. J. Am. Chem. Soc. (Nuclear Medicine) The crystalline sodium iodine with a thallium 128(6):2115–2120. doi:10.1021/ja057254a; International Atomic impurity is the ‘workhorse’ crystal used in emission imaging. It Energy Agency (IAEA). 2012. Nuclear Data for Production of was introduced in 1948 by Robert Hofstadter and it demonstrated Therapeutic Radionuclides. Technical Report Series 473, IAEA, an exceptionally large scintillation light output compared to the Vienna, Austria; Yeong, C. H., M. H. Cheng and K. H. Ng. 2014. organic scintillators previously used. The crystal has very few Therapeutic radionuclides in nuclear medicine: Current and future obvious drawbacks (only in certain applications). NaI(Tl) plays a prospects. J. Zhejiang Univ. Sci. B. 15(10):845–863. doi:10.1631/ central role (or has) in nearly all clinical fields of nuclear medicine jzus.B1400131. and is considered as the standard scintillation material. The most favourable property of the NaI(Tl) scintillator is the Narrow beam geometry high light yield, namely 38,000 scintillation photons per MeV. As (Nuclear Medicine) Narrow beam geometry refers to situations most inorganic scintillators, NaI(Tl) shows non-linearity in light in which a radiation beam profile is typically smaller than a few yield, especially at low energies. NaI(Tl) is also hygroscopic and cm2. The opposite situation is referred to as broad beam geometry. sensitive to mechanical and thermal stress so the crystal must be Consider two different beam geometries falling on a perpendicu- sealed properly. The dominant decay time is 230 ns but there is lar surface. The dose along the central line in the two cases will also a phosphorescence component with a decay time of about differ since in the broad beam situation photons outside the cen- 0.15 s that is as large as 9% of the total light yield, that is 9% tral line can scatter inwards towards it (sometimes referred to as of all de-excitations are delayed because the electrons get caught ‘in-scatter’) and contribute to the dose. The dose along a central in energy state were further de-excitation is forbidden. The elec- line is therefore higher in the broad beam geometry compared to tron must first be excited to an adjacent upper energy state before a narrow beam of equal intensity per unit area. 639 NCRP 640 Navigator echo COCH3 Services Department in Leeds and Radiation and Environmental Monitoring Scotland at Glasgow. Jointly with the HPA division NH for Chemical Hazard and Poisons it forms the Agency Centre for –OOC CH2 CH COO– Radiation, Chemical and Environmental Hazards. The Radiation Protection Division has an Advisory Group on Non-ionising Radiation (AGNIR) and an Advisory Group on FIGURE N.1 Molecular structure of NAA. Ionising Radiation (AGIR). They are responsible for the Agency’s work on non-ionising and ionising radiations, including research to advance knowledge about protection from the risk of these N radiations. It also provides laboratory services, organises training 0.8 –CH3 of NAA, 2.01 ppm courses and provides expert services. The HPA (UK Health Protection Agency) Radiation Protection Division (formerly NRPB) has always had a leading position in 0.6 the field of medical applications and has a significant advisory role in the United Kingdom. Hyperlink: http://www .hpa .org .uk 0.4 Navigator echo (Magnetic Resonance) Navigator echo is the term for projection 0.2 images formed from slice-selective RF pulses that are not phase encoded. The absence of phase encoding creates 1D ‘line’ images that are read out in the frequency encoding direction. Because 0.0 there is no phase encoding, the line images can be acquired ppm quickly. The main use of navigators is to dynamically track motion 4 3 2 1 as navigator echoes can create high-temporal resolution images of motion versus time, similar to M-mode imaging in ultrasound. FIGURE N.2 Proton NMR spectrum from the pons in a normal human Navigator echoes are usually created by using two RF excitation brain showing NAA methyl proton resonance. pulses to create beam at their intersection, or by using a 2D spiral excitation pulse to create a cylindrical excitation. The major uses of navigator echoes are in tracking bulk patient motion, or for National Council on Radiation Protection tracking the position of the liver in respiratory-gated cardiac or and Measurements (NCRP) abdominal imaging (Figure N.3). (Radiation Protection) The National Council on Radiation Related Articles: RF pulse, Motion artefacts Protection and Measurements (NCRP) started its activity Further Readings: Ehman, R. L., and J. P. Felmlee. 1989. in 1929 as ‘The Advisory Committee on X-ray and Radium Adaptive technique for high-definition MR imaging of moving Protection’; originally it was established to represent all of the structures. Radiology 173:255–263; Liu, Y. L., S. J. Riederer, national radiological organisations in the United States. The P. J. Rossmann, R. C. Grimm, J. P. Debbins and R. L. Ehman. idea was to create in the United States the national analogue 1993. A monitoring, feedback, and triggering system for repro- of the International X-Ray and Radium Committee, created in ducible breath-hold MR imaging. Magn. Reson. Med. 30:507; 1928 (see ICRP). Due to the extreme increase in the use of ionis- Sachs, T. S., C. H. Meyer, B. S. Hu, J. Kohli, D. Nishimura and A. ing radiation, the initial informal approach was not considered Macovski. 1994. Real-time motion detection in spiral MRI using sufficient and, in 1964, the committee was reorganised and for- navigators. Magn. Reson. Med. 32:639; Wang, Y., R. C. Grimm, malised by the US Congress. J. P. Felmlee, S. J. Riederer and R. L. Ehman. 1996. Algorithms The NCRP is a US organisation with the scope to formulate and disseminate information, guidance and recommendations on radiation protection and measurements. The council closely follows the publications of the International Commission on Radiological Protection (ICRP) with the focus on data and mate- rial which can make a contribution of public interest. The council has also the mission to facilitate and stimulate co-operation among organisations dealing with radiation protec- tion. One of the main activities is to publish reports on the various topics. Hyperlink: NCRP. http://www .ncrp .org National radiation authority (NRA) (Radiation Protection) See Regulatory authority National Radiological Protection Board (NRPB) FIGURE N.3 Coronal image of the heart and liver showing a navigator (General) The National Radiological Protection Board (NRPB) beam on the dome of the liver (left). The signal from the region excited merged in 2005 with the UK Health Protection Agency (HPA) by the beam is read out as line images over time, and this signal can be and is its Radiation Protection Division. The Division con- used to track diaphragm motion and gate the acquisition to the respira- sists of Headquarter at Chilton in Oxfordshire; Occupational tory cycle. Near zone 641 Net magnetisation for extracting motion information from navigator echoes. Magn. P+ Reson. Med. 36:117. H– Near zone (Ultrasound) See Diffraction Stripping foils P+ Negative contrast media (Diagnostic Radiology) The negative contrast media used in x-ray imaging has lower absorption of x-rays (such as air, carbon diox- ide, etc.). It allows easy penetration of the x-rays and presents the N object filled with this contrast as a darker area on bright back- ground (this refers to an image on x-ray film). Negative contrast media can be used in conjunction with positive-contrast media (high absorbent media, such as iodine, barium meal, etc.) to pro- duce double contrast. Typical use of such contrast is in gastro- intestinal examinations. FIGURE N.4 Extraction of protons is accomplished by stripping nega- Digital x-ray systems allow contrast inversion as part of image tively charged hydrogen atoms of their two electrons. Protons with oppo- processing, which should not be mistaken with the use of nega- site charge will bend outwards and can then be directed onto a target. tive contrast media. Often the visual perception of darker area on bright background is better, than the reverse contrast. This human phenomenon is used, for example in the so-called positive-con- is converted from a negatively charged hydrogen atom to a posi- trast mammograms. tive proton P+. This polarity change is used to extract the proton beam since P+ have a curvature opposite to the H−. The extraction Negative contrast media is illustrated in Figure N.4 where protons are extracted at two (Magnetic Resonance) This term refers to contrast agents that different locations from different parts of the H− beam. After the make particular tissues more conspicuous by decreasing the sig- extraction, the beam is directed onto a target. nal from them. In MRI, image contrast results from interplay between the Negative pions NMR properties of hydrogen nuclei (protons) in tissue and pulse (General) Negative pions are one of the three types of subatomic sequence parameters. Negative contrast is usually achieved particles known as pions, which is an abbreviated form for the by using a T2 or T * 2 -shortening agent, together with a T2- or T * 2 pi meson. The symbol for the negative pion is the π− (the other -weighted pulse sequence. The shortened relaxation times of pro- two being πo and π+). Pions have zero spin and are composed of tons in regions receiving a high concentration of the agent result quarks. The negative pion has a mass of 139.6 MeV/c2, a mean in faster signal decay and hence decreased signal. life of 2.6 × 10−8 s and negative charge. It decays into a muon and Gadolinium chelates in sufficient concentration have a T2- neutrino. shortening effect, particularly those based on dysprosium. The use of the negative pions has been explored in radiother- However, negative contrast agents are more commonly based apy due to the depth-dose pattern of a negative pion beam and on small superparamagnetic particles of iron oxide. These affect other characteristics. relaxation primarily through outer sphere effects, resulting in dra- Related Articles: Auger particles, Beta particles, Compton matic reductions in T2. As the particles become larger, it is a matter scattering, Internal conversion electrons, Photoelectric effect of semantics as to whether the effect is regarded as a shortening of the relaxation time or irreversible dephasing due an increase in NEMA local field inhomogeneity – that is, a susceptibility effect. (Ultrasound) NEMA, the National Association of Electrical Most contrast agents are administered intravenously. Simple and Medical Imaging Equipment Manufacturers, represents the agents such as carbon dioxide, barium sulphate and perfluoro- industry’s interests in ultrasound

and other imaging modalities. chemicals may be administered orally for suppression of bowel NEMA produces guidance and recommendations for the safe and signal which may otherwise obscure structures of interest and effective application of medical ultrasound. In association with lead to motion artefacts. However, these agents may stimulate the AIUM, NEMA issued the Output Display Standards (ODS) peristalsis which is counter-productive. There are commercially safety displays and recommendations for diagnostic ultrasound. available bowel suppression agents, again usually based on super- paramagnetic particles. Related Articles: NEQ (noise equivalent quanta) Gadolinium chelate, Paramagnetic con- (Diagnostic Radiology) See Noise equivalent quanta trast agents, Ultrasmall particles of iron oxide (USPIO), Superparamagnetic particles, Superparamagnetic iron oxide, Positive-contrast media Nerve stimulation (Magnetic Resonance) See Peripheral nerve stimulation (PNS) Negative-ion cyclotron (Nuclear Medicine) A device in which negatively charged par- Net magnetisation ticles, for example H− (hydrogen atom with two electrons) are (Magnetic Resonance) When a water sample is introduced into accelerated in circular paths to several MeV in a magnetic field. a magnetic field (B0), the magnetic moment (here denoted spin) In the case of H−, the negatively charged high energy particle (in of each hydrogen nucleus in the sample experiences two effects the MeV range) loses two electrons when passing through a thin derived from quantum mechanics: The spins align with the main foil of carbon prior to extraction. Hence the total particle charge magnetic field direction (the z or longitudinal direction) in one out Network architecture 642 Neutrino Z Net magnetisation (M0) Ring Mesh Star Fully connected N Line Tree Bus FIGURE N.6 Various network connections. Neuroreceptor targeting (Nuclear Medicine) This term refers to radiopharmaceuticals that target receptors in the neurological system. Neuroligands can be FIGURE N.5 Net magnetisation. labelled with suitable radioisotopes and used to study different neurological systems, for example the dopaminergic system. Such radiopharmaceuticals can be used to study neurodegenerative dis- eases, such as Parkinson’s disease. of two possible ways, either aligned with the +z direction (spin- Related Articles: Tracer kinetic modelling, Receptor target- up, N+) or along the −z direction (spin-down, N−). The alignment ing, Antigen targeting, DNA targeting, Glycolysis targeting, is in either case made with an angle relative to the z-axis (see Apoptosis targeting, Hypoxia targeting Figure N.5). Simultaneously, the spins start to rotate around the Further Reading: Imam, S. K. 2005. Molecular nuclear z-axis (precession) with the Larmor frequency (42.6 MHz/T). It imaging: The radiopharmaceuticals (review). Cancer Biother. can be shown that a small excess (on the order of parts per mil- Radiopharm. 20(2):163–172. lion, ppm) of the spins are aligned in the +z direction. The excess amount depends upon, for example temperature and B0. As an example, the polarisation P defined as Neutral conductor (General) The conductor in a two-wire (single phase) AC elec- trical system intended for current flow back to the source. It is æ N P ABS + - N- ö = ç usually connected to the earth (ground) at the source, that is è N+ + N ÷ - ø at the low voltage site of the mains power transformer. In an ideal case, neutral and earth would be at the same potential. In reality, the neutral is at a higher potential, because of the is approximately 5 × 10−6 at 1.5 T and at body temperature. voltage drop on the neutral conductor due to the return cur- In an MRI experiment, a vast number of spins contribute to rent flow. this process in each imaging element (voxel) and since the rota- In some cases, when the local ground of the electric supply tions of the individual spins are not synchronised (not in phase), system is connected to the neutral, the voltage at conductive a net magnetisation vector (M0) is built up in the z-direction – see enclosures of equipment or appliances may become dangerous. Figure N.5. Connection of earth to neutral conductor is not allowed in medi- As mentioned, P and hence also the magnitude of M0 depends cal equipment. upon the main magnetic field, and when B0 increases M0 increases. Related Article: Earthing (ground) In MRI, M0 is used to create detectable signal by ‘flipping’ the magnetisation into the transversal plane under the influence of an external RF field and hence, an increase in B0 creates an increase Neutrino in signal-to-noise ratio (SNR). The relation between B0 and SNR (General) A neutrino is an elementary particle which has no is generally recognised to be approximately linear in the range of charge and a very small mass, which has not been measured. B There are three types of neutrino – electron, muon and tau. Their 0 values applicable to clinical MRI. symbols are respectively ve, vμ and vτ. There are also anti-neutri- Network architecture nos (symbol v* e for the electron anti-neutrino). (General) Network architecture refers to the design of a computer Electron anti-neutrinos are emitted from the nucleus when network, either a local area network or a larger installation. This protons are changed into neutrons in a nuclear transformation. can include hardware or software and protocols such as the com- An example is the beta decay of the radionuclide phosphorous-32 monly known SMTP, POP3, DHCP, HTTP, TCP/IP, FTP and into sulphur-32, by the transformation of a neutron into a pro- Telnet. ton with the emission of a beta particle (electron) and an anti- The hardware design for a LAN would include the topol- neutrino: 32 15 P ®32 * 16 S + b + ve . Often the neutrino is omitted from ogy; the pattern of linking each computer or network node. Such decay equations. designs include closed loop or ring cable connections, single line In positron decay a proton in the nucleus is converted into a or bus connections, star, mesh or tree connections or fully inter- neutron with the emission of a positron (positive electron or beta connecting designs (Figure N.6). particle) and a neutrino. Neutron activation 643 Neutrons The existence of the neutrino was first proposed by Wolfgang has an exceptionally high thermal neutron capture cross section Pauli in 1930 to explain the conservation of energy in beta decay (240,000 barn), and can cause double strand breaks of DNA from and neutrinos were only experimentally detected in 1956. high LET auger electrons after thermal neutron capture. Boron-10 Related Articles: Beta decay, Positron (19.78% natural Boron) has a high thermal neutron cross section of 3840 barn and high LET products that are highly toxic to targeted Neutron activation cells. Therefore, if boron-10 atoms are secreted into tumours (∼30 (Radiotherapy) Radiotherapy treatments involving energies ppm), then incident thermal neutrons will be captured by B-10, higher than the threshold needed to liberate a neutron may result followed by the emission of very short range (around 10 μm, the in neutron activation, particularly for low atomic number materi- order of one cell diameter), high LET and highly toxic radiation: als placed in the beam. Examples of reactions leading to neutron 10 N activation include the (γ,n) reaction for therapeutic x-ray beams B + n ® 4 He + 7 Li + 2.31 MeV (93.9%) and (p,n) reaction for proton beams. The threshold energy for (γ,n) is generally at the higher energies used for external beam treat- Thermal neutron capture therapy utilising this process is called ment, for example the threshold for O16(γ,n)O15 is 15.7 MeV (here boron neutron capture therapy (BNCT). Many different boron x-rays and γ-rays are treated synonymously). For hadron beams compounds have been tested. The compound of choice is the such as protons, a range of activation products may be created in amino acid analogue p-boronophenylalanine, solubilised at neu- addition to neutrons, including other changed particles, but these tral pH by complexation with fructose (BPA-F). do not have the path lengths of neutrons. While skin melanomas have been treated by thermal neutron Related Articles: External beam radiotherapy, Hadron ther- capture therapy, and brain tumours treated by intracavity ther- apy, Proton therapy mal neutrons, deep seated tumours require an epithermal neutron beam (∼5 keV) to give sufficient penetration for a thermal flux Neutron capture cross section of ∼109 neutrons cm−2 s−1 in the tumour. This reaction has been (General) Neutron capture is one of the two processes that a neu- studied extensively in biological frameworks, which include in tron can interact with the nucleus of an atom – the other is neu- vitro, in vivo and clinical trials. Much effort has been expended tron scattering. Neutron capture may lead to either absorption or on glioblastoma multiforme (GBM), for which epithermal NCT fission. with BPA-F was found to achieve results comparable to or better The cross section is a measure of the probability of an interac- than standard therapies. tion and can be visualised as the cross-sectional area presented Abbreviations: BPA-F = p-boronophenylalanine fructose, to an incident neutron. The neutron capture cross section is the GBM = Glioblastoma multiforme and LET = Linear energy number of neutrons captured per unit volume per second for a unit transfer. incident flux of neutrons and unit nuclear density. Related Articles: Fast neutron therapy (FNT), Boron neutron If the symbol δ (Greek delta) is used to denote the cross sec- capture, Linear energy transfer tion, then δ = πR2 cm2, where R is the radius of the nucleus. A particle which passes through a thin sheet of material of area A Neutron therapy containing NT nuclei which do not overlap will have a probability of NTδ/A of colliding with a nucleus. The number of nuclei/cm2 (Radiotherapy) In neutron therapy, fast neutron beams of typical , energy 20–50 MeV are used to treat deep seated tumours. The called the surface density, given by NT/A is equal to Nt, where N is beams are usually produced using high energy proton beams the number of nuclei/cm3 and t is the thickness of the sheet. The impinging on a beryllium target and may be shaped using a multi- collision rate of an incident beam of n neutrons/cm2, which have a leaf collimator (MLC). Compared with x-rays, neutrons have the velocity of v, is nv · NTδ/A = nvδNt. The cross section is the num- same exponential attenuation with depth, being also uncharged ber of collisions/cm2/s divided nvNt. nv is the particle flux – the particles, but the ionisation density from a neutron is higher than number of particles crossing 1 cm2 of area every second. from x-rays, resulting in greater linear energy transfer (LET). As The unit barn (b) can be used to express the cross section and 1 barn = 10−24 cm2 a consequence the relative radiobiological effect (RBE) of neu- . trons is higher than that of x-rays. Each separate capture process, denoted by i, will have an indi- Abbreviations: LET = Linear energy transfer, MLC = vidual cross section which can be summed to a macroscopic cross section: Σi = Nδ Multileaf collimator and RBE = Relative biological effect. i Related Articles: Charged particle therapy, Hadron therapy, Related Articles: Nucleus, Neutron activation Ion therapy, Proton therapy, Heavy particle beams Neutron capture therapy (Radiotherapy) Neutrons, having no charge, are highly penetrat- Neutrons ing, but lose their energy by collisions with hydrogen nuclei in tis- (General) sue. Fast neutrons (5–30 MeV) from accelerators have been used in external beam neutron therapy for cancer, and found to have Symbol n an advantage in slow growing cancers such as salivary glands. Mass 1.674 × 10−27 kg (939.56 MeV/c2 However, the hydrogen scattering leads to poorly collimated ) beams, and problems for normal tissue tolerance. Neutrons also Charge Neutral have complications arising from late radiation effects. Nuclear spin 1/2 Fast neutrons from reactors or accelerators can be moderated Gyromagnetic ratio 1.83 × 108 rad/s/T and thermalised by collisions with hydrogen nuclei. Some iso- Radiation weighting factor 5–20 depending upon the energy of the topes have very large capture cross sections for thermal neutrons neutrons (0.025 eV). For example, Gd-157 (15.68% natural Gadolinium) Nickel 644 Nitrogen-13 Neutrons are subatomic particles which were discovered in 1932 Related Articles: Magnetic resonance imaging (MRI), by James Chadwick. They are composed of two down quarks Magnet(s), Permanent and one up quark held together by the strong nuclear force. Free neutrons are unstable, decaying with a half-life of 10.3 min into Nit a proton, electron, and antineutrino via beta-minus decay. Along (Diagnostic Radiology) See Brightness with protons, neutrons form the constituent parts of the

atomic nucleus. Medical Applications: Nitrogen Boron neutron capture therapy – In (General) this form of radiotherapy, the interaction between an external N neutron beam and boron (injected intravenously) is used. External beam radiotherapy – Neutron treatment beams have Symbol N been used to treat hypoxic tumours (neutrons have a reduced oxy- Element category Non-metal gen enhancement ratio compared to x-ray photons), but due to Mass number A 14 unacceptable late normal tissue damage this practice has largely been abandoned. Atomic number Z 7 Radiation hazard – Additional shielding is required for high Atomic weight 14.0067 g/mol Electronic configuration 1s2 2s2 2p3 energy clinical linear accelerators (>10 MeV) as here neutrons are produced via both electron–neutron and x-ray–neutron Melting point 63.15 K interactions. Boiling point 77.36 K Related Articles: Atom, Half-life, Gyromagnetic ratio, Boron Density near room temperature 1.251 g/L neutron capture, Oxygen enhancement ratio, Radiation shielding Nickel History: The discovery of nitrogen is formally credited to (General) Daniel Rutherford, who published a report in 1772 describing a component of air in which combustion was not supported. Isotopes of Nitrogen: 99.634% of naturally occurring nitro- Symbol Ni gen is stable 14N. The remaining 0.366% is stable 15N. There also Element category Transition metal exist 10 synthetic radioactive isotopes, with 13N the only isotope Mass number A 58 with a half-life greater than a few seconds (the half-life of 13N is Atomic number Z 28 10 min). Atomic weight 58.694 g/mol Medical Applications: Cryogen – Nitrogen is predominantly Electronic configuration 1s2 2s2 2p6 3s2 3p6 4s2 3d8 used in liquid form as a coolant, for example as an alternative to Melting point 1728 K liquid helium in MRI cooling systems. It can also be used as a Boiling point 3186 K preservative for storing biological tissue. PET Imaging Agent – 13N-ammonia is used as a radiopharma- Density near room temperature 8.908 g/cm3 ceutical in PET imaging studies. It has a high first pass extraction rate into the myocardium, making it useful for myocardial perfu- History: Nickel was first extracted from nickel arsenide in sion imaging. 1751, by Baron Axel Fredrik Cronstedt. It has been found in Related Articles: Cryogen, PET (positron emission tomogra- bronze dating from up to 3500 BC, although it was not identified phy), PET clinical applications, Perfusion imaging until Cronstedt’s discovery. Since 1859, nickel has been used in pure form or as an alloy to make coins. Nitrogen-13 Isotopes of Nickel: Five stable isotopes of nickel are found in (Nuclear Medicine) Nitrogen-13 is a radionuclide used for the nature. The most abundant is 58Ni, with 68.1% natural abundance, assessment of myocardial perfusion with positron emission followed by 60Ni (26.2%), 62Ni (3.6%), and 61Ni and 64Ni, both at tomography. With a short half-life of 9.97 min, Nitrogen-13 must around 1%. Eighteen synthetic radioactive isotopes are known, be produced on-site, requiring an on-site cyclotron. The maxi- which have half-lives ranging from 76,000 years (59Ni) to 110 ms mum positron range of Nitrogen-13 in tissue is approximately 5.4 (78Ni). mm. Nitrogen-13 ammonia is employed to assess relative or abso- Nickel is ferromagnetic and exhibits negative magnetostric- lute myocardial blood flow. tion: slight contraction in the presence of a magnetic field. Medical Applications: Magnetic resonance imaging – Nickel, along with aluminium and cobalt, forms part of an alloy known Positron Maximum Photon Common as alnico, which is used for manufacturing permanent magnets Half-life fraction positron energy emission application used in MRI. Although superconducting magnets are now more widely used, permanent magnets may be advantageous in some 9.97 min 100% 1.2 MeV 511 keV Myocardial situations as they require no cooling and have small fringe fields. perfusion Permanent magnets are often used in ‘open’ magnet configura- imaging tions, which are patient-friendly and enable interventional MRI. Nickel, in the form of NiCl, is also used as a dopant in MR phantoms to adjust the relaxation times of the phantom contents Related Articles: Positron emission tomography, Perfusion to more closely resemble those of biological tissue. The relaxation imaging times of NiCl are largely independent of temperature and operat- Further Reading: Sharp, P. F., H. G. Gemmell and A. D. ing frequency. Murray. 2005. Practical Nuclear Medicine, 3rd edn., Springer. Nitrogen, liquid 645 Noise power spectrum (NPS) Nitrogen, liquid frequency distribution of the noise inherent to the device. The fre- (General) See Nitrogen quency dependent NEQ can now be defined as (Equation N.3) NMR (nuclear magnetic resonance) NEQ ( ) q2MTF ( f )2 (Magnetic Resonance) See Nuclear magnetic resonance (NMR) f = (N.3) NPS( f ) Noise where q is the number of incident photons per unit area (the aver- (Nuclear Medicine) In nuclear medicine noise refers to the back- age uniform input). This has replaced the term Ninc in Equation ground contribution in images, that is corrupted signal without N.1 due to NPS being a measure of the noise power per unit area. any true image information. The background contribution is a MTF( f) and NPS( f) are both linearised functions with respect to N by-product of other activities, for example electric noise in detec- input intensity. tors and background radiation. The noise contribution in an image The interpretation of the frequency dependent noise equiva- can be expressed in a signal-to-noise ratio, SNR. Images can be lent quanta (NEQ( f)) is very similar to that for standard NEQ. smoothed in the image post processing in order to raise the signal- NEQ(f) tells you the number of Poisson distributed quanta that to-noise ratio. would produce the same SNR in an ideal detector at a particular Related Article: SNR (signal-to-noise ratio) spatial frequency component. Related Articles: Noise power spectrum (NPS), Detective Noise equivalent quanta (NEQ) quantum efficiency (DQE), Modulation transfer function (MTF), (Diagnostic Radiology) Noise equivalent quanta (NEQ) Signal-to-noise ratio (SNR) describes the number of photons (quanta) which are incident Further Reading: Beutel, J., H. L. Kundel and R. L. Van on a detector as derived from the detected signal-to-noise ratio Metter. 2000. Handbook of Medical Imaging: Volume 1, Physics (SNR). In the case of an ideal detector the NEQ is proportional and Psychophysics, SPIE, Bellingham, WA. to the square of the measured SNR, that is the measurement follows photon counting statistics. For a non-ideal detector, the Noise power spectrum (NPS) NEQ is defined using the same assumption and hence an equiva- (Diagnostic Radiology) The noise power spectrum (NPS) is the lent number of incident photons can be defined assuming a per- power of noise, contained in a two-dimensional (2D) spatial fre- fect detector. quency interval, as a function of the 2D frequency. The NEQ is used in combination with the detective quantum The noise power spectrum or Wiener spectrum is used to cal- efficiency (DQE) to quantify the intrinsic efficiency of a radio- culate the magnitude and frequency of noise in medical imaging graphic imaging system such as an x-ray flat panel detector. NEQ systems. Although applicable to all imaging analysis it is used provides a quantitative measure of the quality of the image. This mostly in diagnostic x-ray imaging. A measurement of the NPS is can then be used to infer, for example the minimum x-ray dose to also needed to determine the noise equivalent quanta (NEQ) and the patient to provide images of satisfactory quality in an x-ray the detective quantum efficiency (DQE) of an imaging system. radiograph. These quantities allow quantitative evaluation and comparison of For an ideal photon counting detector the number of inci- imaging system efficiency. dent photons (N There are many sources of noise in an x-ray imaging system. inc) is proportional to the SNR by the following (Equation N.1): In a perfect detector, the number of photons measured per unit area will vary randomly due to small fluctuations in the number N of incident photons on the detector, or quantum noise. However, inc µ SNR2 (N.1) in real systems, other forms of noise will also contribute to the final image, for example imperfect detection of incident photons, In a real measuring system the detector is not perfect and not and internally generated system noise. By measuring the NPS of every photon incident on it is converted to a measured count. a system both the random and non-random noise is quantified. This reduces the SNR in comparison to an ideal detector in In 2003, the International Electrotechnical Commission (IEC) which every photon is measured. The measured SNR of a real, published an international standard for the measurement of the non-ideal detector is denoted by SNRmeas. Furthermore, there are DQE, which included a detailed description of NPS measure- additional noise sources within the detector which add to the ment, on which the following description is based. Figure N.7 noise associated with measured signal further reducing SNRmeas. depicts the basic calculation stages. As the NPS describes the The measured non-ideal SNR can thus be related to the equiva- noise in terms of the magnitude of frequency components, Fourier lent number of quanta that would be incident on the detector by analysis is used to calculate the spectrum. (Equation N.2) The NPS is measured by first taking a flat field image, an image with no test object in the beam with a standard spectrum NEQ µ SNR2 meas (N.2) and geometry. A flat field image appears dark grey and uniform to the naked eye, for with no object signal the image contains only For the case of a real medical imaging detector both the sig- noise components. Several exposures should be taken as it is rec- nal measured by the detector and the inherent noise generated ommended that there be 4 million independent pixel signals for become spatial frequency dependent quantities. Thus, the NEQ measurement. It should be noted that the taking of several images must also be frequency dependent. The signal output of an imag- assumes that there is no noise component that changes with time, ing device is related to the input signal via the modulation transfer that is the noise components are either completely random or tem- function (MTF( f)). This describes how well the imaging device porally reproducible and thus do not change over time. All images amplifies the input signal over the various spatial frequencies. must be linearised to convert the signal intensity to units of quanta The frequency dependence in the noise of the device is denoted absorbed per area. This is especially important if the DQE will be by the noise power spectrum (NPS( f)). This describes the spatial measured from the NPS as linearising the image negates the need Noise power spectrum (NPS) 646 Noise power spectrum (NPS) Spatial domain Half overlapping Detrending correction: 256 × 256 pixel ROIs Measurement ROI are created 125 × 125 mm Subtract 2D second order polynomial 256 pixels N 2D FFT of each ROI 103 Frequency domain 104 105 Sum all 2D FFTs 106 Plot as 1D or 2D y-ax 4 is fre 2 que 0 n 4 1E–04 cy –2 2 (c –4 –2 0 x-axis frequency (cycles/mm) ycles/ –4 m 1E–05 m) 1E–06 1E–07 0 1 2 3 4 5 6 7 Frequency (cycles 1 mm) Scan direction CCD direction 45° CR 25.0 45° FIGURE N.7 Measurement of NPS. (Courtesy of KCARE, UK www .kcare .co .uk) for a further gain factor in the DQE measurement. All measure- where ments are taken from a central region of interest (ROI) of area un, vk are frequency components in each axis direction in 125 mm × 125 mm. This analysis ROI is then ‘detrended’ by sub- Fourier space tracting a second-order polynomial from the image signal. This Δx, Δy are the pixel spacing detrending is used to remove some non-random image intensity Nx, Ny are the number of pixels in the individual ROIs fluctuations such as the intensity gradient caused by the anode M is the number of ROIs or ensemble averages heel effect; however it may also remove some noise components I(xi, yj) is the linearised data that are wanted be measured, such as structured noise, and some S(xi, yj) is the 2D polynomial investigators withhold this step. The analysis region is then broken up further into half-over- The NPS can be shown as either a 3D graph, or as a 2D line lapping regions of interests made of 256 × 256 pixels. For each graph depicting the horizontal and vertical NPS. The horizontal ROI the

2D Fourier transform is calculated and then the modulus and vertical NPS are found by averaging NPS at each frequency squared is taken, the final NPS is found by averaging all the 2D for ±7 points either side of the axes. Fourier transforms. Mathematically for a 2D, discrete (pixellated) Related Articles: Modulation transfer function (MTF), Noise image the NPS is defined as equivalent quanta (NEQ), Detective quantum efficiency (DQE) Further Readings: International Electrotechnical Commission NPS (un,vk ) (IEC). 2003. Medical Electrical Equipment–Characteristics of Digital X-ray Imaging Devices–Part 1: Determination of the 2 (N Detective Quantum Efficiency, International Electrotechnical xNy x y) M Nx N D D y ååå{ ( ) - ( )} i e(-2p (unxi +vky = I xi , yj S xi , y j )) j M Commission, Geneva, Switzerland, pp. 62220–62221; Dobbins, m=1 i=1 j =1 J.T., III, E. Samei, T. Ranger and Y. Chen. 2006. Intercomparison NNPS (mm2) NNPS (mm2) Nominal ocular hazard distance (NOHD) 647 Non-designated (public) area of methods for image quality characterization. II. Noise power aperture ratio – an analogue of the fill-factor of the pixels in flat spectrum. Med. Phys. 33:1466–1475. panel detectors in digital radiography. Usually this ratio is of the order of 0.9. Nominal ocular hazard distance (NOHD) The nominal size of a CRT monitor depends on the focussing (Non-Ionising Radiation) The nominal ocular hazard distance of the scanning electron beam (and the aperture grill and phos- (NOHD) for a particular laser may be calculated using the expo- phors, in case of colour monitors). In these monitors the differ- sure limit value (ELV) for the eye, given in Annex II of Directive ence between nominal pixel size and actual pixel size is larger, 2006/25/EU, and the Accessible Emission Limit (AEL) for that as the diameter of the actual pixel size is measured at 50% of the laser. The NOHD demarcates the distance away from the laser luminance profile (FWHM of the light spread around the centre beam where the ELV ceases to be exceeded; that is, where the of the pixel). Related Articles: Actual pixel size monitors, Detector fill N AEL is equal to the ELV. At distances greater than the NOHD, the ELV for the laser will not be exceeded and the beam is safe for factor viewing. The NOHD can be calculated using Equation N.4 where Further Reading: AAPM Report OR03. 2005. Assessment of all units are base SI units. Display Performance for Medical Imaging Systems. Nominal ocular hazard distance (m) Nominal value (General) Nominal value is the stated value of a physical quantity é 4.Power ù ê - initial beam diameterú (N.4) or a property of a device, under normal conditions. The actual = ë p.ELV û value may differ from the nominal for an absolute value or per- Divergence, d (rad) centage declared by the tolerances. For example, the nominal value of line voltage in Europe is 230 V, with tolerance ±10%. Divergence is a measure of how the laser beam spreads out along Resistors are produced with nominal values taken from Renard its path and can be calculated as shown in Figure N.8. series of preferred numbers and defined tolerances (20%, 10%, Equation N.4 demonstrates that the NOHD depends strongly 5%, etc.). upon the individual properties of each laser such as power, spot size, divergence and wavelength. The laser NOHD may be used to Non-coplanar beams define the controlled area for a specific laser. (Radiotherapy) Typically the fields used for radiotherapy treat- Related Articles: Exposure limit value (ELV), Accessible ment are delivered in a single plane around the patient. However, emission limit (AEL), Controlled area (Laser) on occasion it is beneficial to treat with beams incident outside a Further Reading: Council Directive 2006/25/EC on the single plane. This can be achieved through a combination of linac minimum health and safety requirements regarding the exposure movement (gantry rotation) and couch movement (couch rota- of workers to risks arising from physical agents (artificial opti- tion). Non-coplanar beams are most commonly used for brain and cal radiation) (19th individual Directive within the meaning of head and neck treatments where the target is surrounded by criti- Article 16(1) of Directive 89/391/EEC) [2006] OJ L 114. cal structures. An illustration of a non-coplanar beam arrange- ment is given (Figure N.9). Nominal output (General) See Nominal value Non-designated (public) area (Radiation Protection) One of the main methods for controlling Nominal pixel size, monitor the potential for persons to be exposed to ionising radiation from (Diagnostic Radiology) The nominal pixel size (NS) for digital a facility is to designate and demarcate areas where a significant (LCD) monitors is the active display size, dimension in a spe- hazard, and the subsequent risk of exposure, exists. cific direction, divided by the number of pixels in that direction. Areas which are considered to pose a significant hazard/risk The actual pixel size (AS) is smaller because of the electronic will be designated as ‘controlled areas’. Other areas that may components in a pixel. The ratio between the AS/NS is known as under certain circumstances, especially with changing work FIGURE N.8 Laser beam divergence. Non-homologous end-joining repair pathway 648 Non-ionising radiation of c-NHEJ are cell cycle-dependent. There are two types of c-NHEJ in G0/G1 phases: (a) fast (2–4 hrs) and (b) slow (> 8 hrs). The fast one is DNA-end-resection independent. The slow repair process in G1 also involves c-NHEJ proteins but addition- ally it requires the nuclease Artemis (i.e. protein involved in DNA repair) and DNA end resection prior to the DNA end ligation. LT The resection process in G1 is similar to that of resection during homologous recombination in G2, but the process is managed by the cell differently to suit a c-NHEJ process. The slow process is N more error-prone. In G2, only the resection-independent c-NHEJ is active. There is also the alternative-NHEJ (alt-NHEJ) pathway, which involves enzymatically-driven resection. This is slow and error- prone (similar to resection-dependent c-NHEJ) but does not seem RTLAT to contribute to repair in G0/G1-phase human cells. Since DNA free-ends must spatio-temporally co-localise to either rejoin or misrejoin (i.e. inaccurately join), it appears that the slowness of resection-dependent c-NHEJ and alt-NHEJ is FIGURE N.9 Illustration of a non-coplanar beam arrangement. what makes them more error-prone than resection-independent c-NHEJ, which is fast; i.e. the DNA free-ends have more time to move away from their initial location (possibly with Brownian patterns, pose a potential risk and need to be kept under review, motion) and meet incongruent DNA free-ends. It follows that are designated as ‘supervised areas’. a misrejoining is more likely to occur if the DSBs are initially Any other area adjacent to the facility that poses no significant closer together. hazard or risk to staff, patients and visitors, or the public, must The complexity of a DSB appears to affect the choice of repair then be ‘non-designated areas’ – i.e. public areas without the need pathway. Complex DSBs are repaired by slow pathways and sim- to restrict access. Examples include areas in and around the facil- ple DSBs by fast pathways, though it is unclear why this is the ity such as a main public corridor in a hospital, offices used by case. Accordingly, DNA free-ends from complex DSBs are more staff not directly working with ionising radiation, or areas outside likely to be misrejoined. the hospital building accessible to the public. For example, while this process is usually successful in repair- By definition there can be no restrictions placed on access to ing DNA DSBs, such as those following x-ray irradiation, if the non-designated (public) areas on the basis of any radiation dose DNA has been exposed to a form of high LET radiation, multiple that could be received. DSBs may form in close proximity. In this case, there is a chance Related Articles: Controlled areas, Supervised areas that other nearby damaged DNA molecules may be joined result- ing in severe chromosomal re-arrangements, where some chro- Non-homologous end-joining repair pathway mosome aberrations can lead to cell death. (Radiotherapy) The main repair pathway for double-strand Further Readings: A, A. S. and G. Iliakis. 2013. Nucleic breaks (DSBs) in all cell cycle phases for mammalian cells is Acids Res. 41:7589; Biehs, R., M. Steinlage, O. Barton, S. Juhasz, canonical non-homologous end-joining (c-NHEJ). This process J. Kunzel, J. Spies, A. Shibata, P. A. Jeggo and M. Lobrich. can be active in any stage of the cell cycle. It differs from the 2017. Mol. Cell. 65:671; Douglass, M. 2014. Section 8.1 in homologous recombination in that it does not require a homolo- Development of an Integrated Stochastic Radiobiological model gous template. In this process the broken ends of the DNA are for Electromagnetic Particle Interactions in a 4D cellular geom- ligated directly without the need for the template (Figure N.10). etry, PhD Thesis, University of Adelaide; Le Guen, T., S. Ragu, It also means the two DSB ends are rejoined with minimal J. Guirouilh-Barbat and B. S. Lopez. 2015. Mol. Cell. Oncol. processing. 2:e968020; Li, Y., P. Reynolds, P. O’Neill and F. A. Cucinotta. While the c-NHEJ represents the major repair pathway for 2014. PLoS One 9:e85816; Liang, Y., Q. Fu, X. Wang, F. Liu, G. two-ended DSBs in all cell cycle phases, particular mechanisms Yang, C. Luo, Q. Ouyang and Y. Wang. 2017. Phys. Med. Biol. 62:2153; Lobrich, M. and P. Jeggo. 2017. Trends. Biochem. Sci. 42:690. Non-ionising radiation (Radiation Protection) Non-ionising radiation refers to any part of the electromagnetic (EM) spectrum that cannot induce ionisa- tions in matter. In other words, unlike ionising radiation, non-ion- ising radiation does not have sufficient energy to interact directly with sub-atomic electrons. Types of non-ionising radiation that are used clinically include radio-frequency fields (in MRI) and ultraviolet (for dermatological therapy). The minimum energy for ionisation – that is, giving electrons enough energy to leave their orbits – is around 12eV, which is equivalent to the energy in an electromagnetic wave of wavelength FIGURE N.10 Non-homologous end joining. (From M. Douglass, page approximately 100 nm; this is at the far end of the ultraviolet 125.) range of the electromagnetic spectrum. Below the 12 eV threshold Non-linear dose–response curve 649 Non-linearity parameter for atomic ionisation, the radiation may still have enough energy to break molecular bonds in human tissues. The threshold for this effect may be as low as 4eV. This threshold lies in the UVB range of the spectrum. ‘Non-ionising radiation’ does not mean ‘non-hazardous’. Despite their inability to invoke ionisation events in matter, non- ionising radiation can have a damaging effect on tissue. Ultraviolet radiation can cause skin erythema, intense light or near-infrared radiation can burn the cornea or retina of the eye, whilst high levels of lower-frequency EM radiation can cause tissue heating at depth within the body. For example, the rapidly varying radio- N frequency waves that re-align the magnetic dipole of water mol- ecules leading to a signal detected in an MRI scanner, may also cause heating in the body tissues of a patient. The range of effects of non-ionising radiation on the tissues of the human body is still not fully understood. The International Commission on Non-Ionising Radiation (ICNIRP) has published guidance in this field, including proposed guideline limits for occupational and public exposure (see Further Reading). FIGURE N.11 Illustration of how a sound wave is distorted as it Related Articles: Ultraviolet radiation (UV), Infrared radia- propagates. tion (IR) Further Readings: 1985. Review of concepts, quantities, units and terminology for non-ionising radiation protection, Health and thereby also the ratio v/c (particle velocity over the sound Phys. 49(6):1329–1362; Guidelines on Limiting Exposure to speed). As a consequence the wave tends to migrate as c+v. This Non-Ionizing Radiation. A reference book based on guidelines means that in the compression phases, when the particle velocity on limiting exposure to non-ionizing radiation and statements on is at its maximum and in the propagation direction, the sound special applications, International Commission on Non-Ionizing speed is higher than in the rarefaction phases, where the particle Radiation Protection, Munich, Germany. (ICNIRP, 1999). velocity is at its maximum in the opposite direction. Non-Linear Compressibility: Second, with increased pres- Non-linear dose–response curve sure, the stiffness and bulk modulus of a fluid tend to increase. (Radiation Protection) The non-linear dose–response curve is As the sound speed is proportional to the square root of the bulk just one example

of a dose–response curve, together with non- modulus, the sound speed will also increase with pressure due to threshold dose–response curve, linear response curve, etc., which this effect. may be used either separately or in combination as models to Practical Effects: These effects manifest themselves in the describe the response of the human body to exposure to various way that the compressions (where there is high pressure) travel types on ionising radiation from both internal and external expo- faster and ‘catches up’ with the rarefactions (where the pres- sure, and at high and low doses and dose-rates. sure is low). The effect is a shocked wave front as depicted in The non-linear dose–response does not assume that there is Figure N.11, middle panels. The shocked waveform has a very a linear response between exposure at low doses/dose-rates and high harmonic content due to the asymmetric waveform, and exposure at high doses/dose-rates. as attenuation is frequency dependent, this counteracts the Related Articles: Dose–response curve, Linear dose–response build-up of harmonics. After propagating a longer distance, the curve energy transfer to higher harmonics is insufficient, and the wave again appears as a sinusoidal wave, but with lower amplitude Non-linear propagation (Figure N.11, bottom panel). (Ultrasound) Non-linear propagation of a sound wave can be Related Article: Harmonic imaging observed as a distortion of the wave as it propagates, as shown in Figure N.11. A propagating wave is usually described by the wave Non-linearity parameter equation, but the wave equation cannot explain this behaviour. (Ultrasound) The non-linearity parameter (or the B/A value) indi- This is due to the fact that as the equation is derived, a number cates how a medium supports the build-up of non-linear waves. A of non-linear terms are neglected based on the assumption that high intensity wave will generate changes in the speed of sound. the pressure perturbations are small. In diagnostic ultrasound The perturbed speed of sound can be written as however, the pressure amplitudes are on the order of a magnitude ( A B higher than the atmospheric pressure, and therefore the non-linear æ terms will not be negligible. In practical terms this has two effects ( ) B u(t ) ö (2 / )+1) c t = c0 çç1 + ÷ è 2A c ÷ 0 when a propagating wave is considered: ø Description of Figure N.11 – the top panel shows the wave as where it appears close to the transducer. The next two panels show how c0 is the speed valid for small pressure variations the waveform progressively is distorted as it propagates, and the u(t) the particle speed last panel how it appears after energy has been attenuated after a B/A is a material dependent parameter ‘long’ distance, and the amplitude is not sufficient to produce the non-linear effect. For water B/A = 5,2 (at 30°C and atmospheric pressure), Convection: First, there is a convection effect. For large whereas for air it is 0.2. For tissues B/A falls in the range of 5–11. acoustic pressures, the particle velocity will also be considerable, Also used is the non-linear coefficient β, which is β = 1 + B/2A. Non-paralysable counting system 650 Non-stochastic effects Non-paralysable counting system Non-screen film (Radiation Protection) The dead time τ in a detector is a mini- (Diagnostic Radiology) Non-screen film is film that is exposed mum time interval which must separate two events to be mea- without the use of intensifying screens. The film is placed in black sured (registered) as two separate pulses. envelope and in the past had been used for some high-resolution In non-paralysable counting system the events that occur dur- radiographs. Nowadays it is not used for human radiography ing the dead time period are lost, so that with an increasing event because high exposures are required. It is useful for some appli- rate the detector will reach a saturation rate equal to the increase cations such as specimen radiography. of the dead time. The advantage of non-screen film is that it can produce an If we assume that N is a true interaction rate and M is a mea- image with high visibility of detail because the blurring by the N sured counts rate for the detection system with dead time τ, the intensifying screen is eliminated. difference between a true and measured rate is equal to It can be useful in the measurement of geometrical parameters in the quality controls of x-ray and radiotherapy equipments. N – M = N ´ M ´ t where Mxτ is the fraction of all time when the detector is ‘dead’. Non-stochastic effects Then the true interaction rate N can be calculated from the (Radiation Protection) There are two types of biological effects following equation: (Bioeffects) of ionising radiation on human tissues categorised by M the risk of the effects being observed. These two categories are- N = Stochastic and Non-Stochastic effects. 1- M ´ t Non-stochastic effects result from cell killing due to radiation Related Articles: Dead time losses, Paralysable counting exposure, and they occur in all persons exposed to a relatively system high radiation dose above a threshold. Further Reading: Knoll, G. F., 2000. Radiation Detection The cells in many tissues and organs of the body are continu- and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, ously being replaced. However, the balance between loss and pp. 119–120. replacement rates may be affected by many factors including exposure to radiation. The result of a net loss of cells may be the failure of function of the tissue or organ. Gross reductions in the Non-scattering grid number of healthy cells in vital tissues may lead to death. For (Diagnostic Radiology) Non-scattering grid is also known as anti- each tissue type there is an absorbed dose threshold level below scattering grid. See article Grid, Bucky. which the probability of harm is very low. Above this threshold, Related Articles: Grid, Bucky pathological conditions will be observed in nearly all the irradi- ated population and their severity will be dose-dependent. Below Non-specular the threshold, it is assumed that the body is able to replace the (Ultrasound) Reflections (echoes) of the ultrasound pulse are cells at a faster rate than they are being killed – hence no dam- fundamental to generating an ultrasound image. Specular or non- age is detected. The following diagram represents the assumed specular reflections can occur to the outgoing pulse as it propa- relationship between non-stochastic effects and radiation dose gates through a medium (i.e. tissue) and particularly at boundaries received (Figure N.12). between different media. Examples of these non-stochastic effects range from eye Non-specular reflections occur when the interface surface cataracts, sterility, skin-reddening (erythema), and effects on a (sometimes called acoustic boundary) between two different foetus, all of which occur at moderate exposures, to the symp- media is ‘small’ or ‘irregular’. A ‘small’ interface surface is one toms of acute radiation syndrome such as nausea and vomiting which is less than the beam’s wavelength, and ‘irregular’ refers at exposures near lethal levels. Because of the onset of clinically to a rough/bumpy surface (i.e. one that is not smooth or flat). The observable effects of a deterministic nature, the International consequence of the ultrasound beam interfacing with such a sur- Commission on Radiological Protection now calls them ‘deter- face is the beam’s reflection path will not be at a single angle and ministic effects’. direction (as it does with a specular reflection) but will be either: Related Articles: Bioeffects, Deterministic effect, Stochastic effect • Diffuse reflect – over a range of angles • Scatter – in all directions (only if the surface size is much smaller than the beam’s wavelength) As a result, the amplitude of non-specular reflections is smaller in comparison but provide useful information about the internal texture (parenchyma) of most organs (e.g. liver and pancreas). Non-specular reflections do not adhere to the laws of reflection that govern specular interactions. The power (Ws) dispersed at these small, or irregular, targets can be related to the relationship between the target/surface size (d) and pulse wavelength (λ) as Damage well as to the target/surface size and frequency. Often referred to threshold as Rayleigh scattering, the amount of backscatter power is highly dependent on frequency and size of target/object as follows: Dose d6 W 6 s µ l4 µ d f 4 FIGURE N.12 Dose–response curve for non-stochastic effects. Severity of damage Non-threshold dose–response curve 651 Non-uniformity Further Reading: ICRP. 2008. Recommendations of the 1. Transmit/receive coil RF non-uniformity International Commission on Radiological Protection, Ann. This is perhaps the most significant determinant. In ICRP, ICRP Publication 103, 37(2–4). general, transmit coils are required to have very good uniformity so that transmission flip angles are constant Non-threshold dose–response curve throughout the volume to be imaged. For receive coils, (Radiation Protection) The curve that describes the dose– good uniformity is generally less of a priority than good response relationship for excess risk of cancer following exposure SNR. For this reason image non-uniformity is gener- to ionising radiation at low dose and/or low dose rates with the ally more dependent on RF reception rather than RF assumption that there is no threshold value below which there is transmission. no risk. 2. Nyquist limit filtration To prevent aliasing, active filters within the RF N Presently, for the purpose of radiological protection, the assumption is made that the underlying dose–response relation- receiver strongly attenuate all received frequencies ship is linear-quadratic with no threshold. above the Nyquist limit. Such filtration can introduce Related Article: LNT model non-uniformities to the image profile in the frequency encoding direction on older MR systems. Non-transparent 3. Eddy currents (Diagnostic Radiology) See Opacity Eddy currents (interaction between the metal magnet housing and the rapidly switched RF gradient fields) can cause image non-uniformities. Non-uniform activity distribution 4. Pulse sequence parameters (Nuclear Medicine) A non-uniform tracer distribution in an organ Some pulse sequence parameters can adversely or a compartment. Non-uniform distribution of tracers allows affect image uniformity. For example, if the inter-slice imaging of biological and pathological processes, for example spacing is too low, cross talk between adjacent slices FDG-compound is accumulated in regions with a high glucose can sometimes affect image uniformity. Slice profile uptake which can be an indication of tumour cells. and slice width are also important parameters. When calculating the radiation dose to patients using the 5. Dielectric resonance effects MIRD formalism one assumes a uniform activity distribution If high frequency RF pulses have wavelengths of the inside an organ. This is an obvious flaw to the MIRD formal- order of the dimensions of the human body, dielectric ism since the activity is not uniformly distributed. Future resonant interactions between the RF fields and the research with high-resolution detectors and elaborate Monte body can occur, causing intensity non-uniformity. Carlo simulations will allow for better determination of the radiation dose. Measuring Image Uniformity: Quantitative values for unifor- mity can be extracted from images of a uniform (flood field) test Non-uniformity object. (Magnetic Resonance) Intensity non-uniformity is the smooth The Institute of Physics and Engineering in Medicine (IPEM) intensity variation often evident in MR images, even those of uni- recommend the use of intensity profiles such as that shown in form media (Figure N.13). Figure N.14. Their method is as follows: Image non-uniformity in MRI is caused primarily by the fol- lowing factors: 1. On the test object image, draw a roughly central profile parallel to the frequency encoding direction, avoiding any artefacts. 2. Find the modal value of the 100 pixels at the centre of this profile. 3. Calculate the fractional uniformity: the fraction of the profile which lies within ±10% of the modal value. 3809 0 0 Horizontal profile width 0.9 at y = 120.0 239 FIGURE N.13 Image of a uniform test object. If the image represented the test object faithfully, all test object pixels would be of uniform inten- FIGURE N.14 Horizontal profile of voxel values taken through the cen- sity. In reality the voxels towards the centre of the test object are less tre of the test object in Figure N.13. For a totally uniform image this intense than those at the periphery: this is image non-uniformity. profile would be a top hat function. Value Normal database 652 Normalised noise power spectrum (NPS) 4. Repeat for 10 profiles and take a mean

to reduce the é 2 effect of noise. x - f ( x) 1 1 = ê- æ m ö ù exp ú s 2p ç ÷ 5. Repeat steps 1–4 for profiles in the phase encoding ëê 2 è s ø ûú direction. ¥ ò f ( x)dx = 1 Results: IPEM suggest that image uniformity should typically -¥ lie between 0.6 and 1.0. Further Readings: Lerski, R. et al. 1998. Quality Control The normal distribution has the following properties, which are in Magnetic Resonance Imaging (IPEM Report 80), Institute of illustrated in Figure N.15: N Physics and Engineering in Medicine, York, UK; Simmons, A. 1994. Sources of intensity nonuniformity in spin echo images at • It is symmetric about the point x = μ 1.5 T, Mag. Reson. Med. 32(1):121–128. • It has a characteristic ‘bell’ shape • The width of the curve is described by the standard Normal database deviation σ (Nuclear Medicine) The gender-specific normal database • The points x = μ ± σ are inflection points where d2f/ includes information from a number of patients with confirmed dx2 = 0 low likelihood of coronary artery disease. The gender-specific normal limits are derived from the mean of normal uptake in The probabilities that X lies between 1σ, 2σ or 3σ of the mean, a particular region and its standard deviation. The criteria for are as follows: abnormality of a specific region of the myocardium are based upon a comparison between regions for healthy and correspond- P(m - s < X £ m + s) » 68.3% ing regions for diseased patients. When the circumferential P(m - 2s < X £ m + 2s) » 95.4% profiles for individual patients match those normal limits, the P(m - 3s < X £ m + 3s) » 99.7% area of abnormal pixels that falls below them and the relative depth of severity can be calculated. If the myocardial perfusion Related Article: Gaussian distribution images are not attenuation corrected then there is no simple rela- tionship between tracer uptake and the number of counts in each Normal organ dose tolerance pixel. Image scaling is therefore necessary often relative to the (Radiotherapy) Radiation treatment inevitably affects normal tis- maximum pixel counts. sue and so may cause radiation-induced adverse effects. In radio- It is necessary to develop limits for normality and criteria therapy, it is generally the case that the total dose that can be for abnormality for the different protocols. The mean and stan- tolerated depends on the volume of tissue irradiated – the dose– dard deviation of the gender-specific database for the normal volume effect. Additionally, the tissue architecture is thought to response were calculated based on a limited number of male be important in determining the tolerance dose for partial organ and female subjects with a <5% likelihood of having CAD. The irradiation. For further information, see the article Tolerance. normal limits and criteria for detection of perfusion abnormali- Related Article: Tolerance, Adverse effects (Radiotherapy), ties were performed based on a number of male and female Long-term morbidity patients undergoing 99mTc-Sestamibi SPECT and coronary angiography. The normal limits of a gender-specific normal Normalised noise power spectrum (NPS) database in different regions of the myocardium optimise the (Diagnostic Radiology) The normalised noise power spectrum balance of sensitivity and specificity in the detection of perfu- (NPS) is a specific measure for noise in a digital image showing sion defects and the assessments of their reversibility, extent the NPS at a specific exposure value. and severity. In general terms, NPS (or Wiener spectrum) represents the Normal databases are used in other imaging applications such variation of noise amplitudes for different spatial frequencies. as brain perfusion imaging for statistical parametric mapping. Further Reading: Rozanski, A., G. A. Diamond, J. S. Forrester, D. S. Berman, D. Morris and H. J. Swan. 1984. Alternative refer- 0.4 ent standards for cardiac normality. Implications for diagnostic sd = 1 testing. Ann. Intern. Med. 101(2):164–171. 0.35 sd = 2 sd = 3 0.3 Normal distribution (General) The normal distribution is synonymous with the 0.25 Gaussian distribution, and is considered to be the most important 0.2 continuous probability distribution, since it describes the behav- iour of many random variables of interest. It can also be used to 0.15 approximate other more complicated distributions. If a random variable X follows a Gaussian distribution with 0.1 mean μ and variance σ2, we write: 0.05 X ~ N (m, s2 ) 0 –6 –3 0 3 6 9 12 The equivalent probability density function is described as FIGURE N.15 Gaussian distribution or normal distribution for mean 3 follows: and various values of the standard deviation. Normal speed tube 653 Normal tissue complication probability However, noise depends on the radiation exposure which also the fractional volume irradiated and the absorbed dose received determines the signal. The normalised NPS (NNPS) is the abso- by the volume. Many other factors are involved, particularly the lute NPS, divided to the square of the large area signal (the aver- fractionation scheme, but for this model these other factors are age signal) – LAS. held constant. It is assumed that the volume dependence of the complication probability can be represented by a power–law rela- NPS NNPS = tionship, Equation N.5, where TD(V) is the tolerance dose for a LAS2 partial volume V, TD(1) is the tolerance for full volume, and n is a fitted parameter: This assumes the signal (i.e. the photon fluence, or expo- sure) is uniformly distributed in the measured area. Typically, the increased exposure will lead to increased NPS and also to TD ( ) TD (1) V = N.5 V n ( ) N increased signal LAS (which is squared in the formula), hence the NNPS will decrease with the increased exposure. The dose dependence is represented by the integral of a normal In digital images the NPS is measured in (pixel values) × mm2, distribution which is just one of several possible representations the signal LAS is measured in (pixel values), thus the measure for of a sigmoid curve, Equation N.6: NNPS is mm2. See NNPS diagram at the article about NPS. Related Articles: Noise power spectrum t 1 NTCP = òe-t2 /2 Further Readings: Samei, E. and D. J. Peck. 2019. Hendee’s dt Physics of Medical Imaging, Wiley-Blackwell, ISBN-13 2p -¥ 9780470552209; Whitman, G. J. and T. M. Haygood. 2013. Digital where Mammography: A Practical Approach, Cambridge University Press, 9780521763721. D - TD50 (V ) t = ( . ) s(V ) N 6 Normal speed tube (Diagnostic Radiology) The speed of x-ray tube anode rotation The model utilises the TD50/5 (TD50 in Equation N.5), the dose depends on the frequency of the current supplying its stator. that would result in 50% incidence of complications in a popula- Usually normal speed is considered for stator supply with 50 Hz tion after 5 years. The sigmoid curve is completely defined by the (or 60 Hz in the United States), allowing rotation with approxi- mean TD50 (V) and the standard deviation σ(V) which for this mately 3000 rpm (in fact due to the mechanical slipping of the model was approximated as shown in Equation N.7 where m is a rotor this speed is around 2800 rpm). fitted parameter: Related Articles: Anode, Rotation anode, Anode rotational speed, Anode acceleration s(V ) = m ´TD50 (V ) (N.7) Further Reading: Hertrich, P. 2005. Practical Radiography, Siemens, Erlangen, Germany. Lyman gave estimates of TD50(V), n, and m based on the clinical data available at the time. The complication end points were not Normal tissue complication probability specified. Burman et al. (1991) obtained Lyman model parameters (Radiotherapy) The relationship between dose and normal tissue for the clinical tolerance data compiled by Emami et al. (1991). complication probability (NTCP) is shown in Figure N.16. It has In the Lyman model, when n → 1 there is a large volume effect a sigmoid (S) shape with the probability of tumour control tend- and when n → 0 the volume effect is small. Figures N.17 and ing to zero as the dose tends to zero and tending to 100% at very N.18 show the three-dimensional (3D) surfaces obtained from the large doses. Lyman model using the data of Burman et al. for lung (n = 0.87) Several radiobiological models have been proposed that relate biological effect to volume and dose distribution (dose–volume models). Probably the most widely used to date is that of Lyman (1985) in which it is assumed that NTCP is a function of both 1 0.9 0.8 1 0.7 0.75 100 0.6 0.5 0.5 0.25 80 0.4 0 60 0 0.3 0.2 40 0.2 0.4 0.1 Part 0.6 20 ial vo 0 lume 0.8 Dose (Gy) 1 0 FIGURE N.16 Relationship between the probability of normal tissue FIGURE N.17 Lyman NTCP model of dose–volume complication rela- complication and dose is sigmoid in shape. tionship for lung. NTCP NTCP Dose (gray) Normal tissue dose 654 Normal tissue reaction complication probability and TD50/5 = The dose that would result in 50% incidence of complications in a population after 5 years. Related Articles: Dose–response model, Dose–volume his- togram, Fractionation, Parallel organ, Radiobiological models, Serial organ, Sigmoid dose–response curve, Therapeutic effect, Tolerance 0.8 Further Readings: Burman, C., G. J. Kutcher, B. Emami, and 0.6 100 M. Goitein. 1991. Fitting of normal tissue tolerance data to an 0.4 0.2 80 analytic function. Int. J. Radiat. Oncol. Biol. Phys. 21:123–135; N 0 Emami, B., J. Lyman, A. Brown et al. 1991. Tolerance of normal 60 0 tissue to therapeutic irradiation. Int. J. Radiat. Oncol. Biol. Phys. 0.2 40 21:109–122; Hall, E. J. and A. J. Giaccia. 2006. Radiobiology 0.4 for the Radiologist, 6th edn., Lippincott Williams & Wilkins, Part 20 ial 0.6 Philadelphia, PA; Kallman, P., A. Agren and A. Brahme. 1992. volume 0.8 Tumour and normal tissue responses to fractionated non-uniform 1 0 dose delivery. Int. J. Radiat. Biol. 62:249–262; Kutcher, G. J. and C. Burman. 1989. Calculation of complication probability FIGURE N.18 Lyman NTCP model of dose–volume complication rela- factors for non-uniform normal tissue irradiation: The effective tionship for rectum. volume method. Int. J. Radiat. Oncol. Biol. Phys. 16:1623–1630; Lyman, J. T. 1985. Complication probability as assessed from dose-volume histograms. Radiat. Res. 104:S13–S19; Lyman, J. and rectum (n = 0.12) respectively. It must be remembered that T. and A. B. Wolbarst. 1989. Optimization of radiation therapy. the Emami data were largely the result of pooled clinical experi- IV. A dose-volume histogram reduction algorithm. Int. J. Radiat. ence and since most therapies are by definition designed to gener- Oncol. Biol. Phys. 17:433–436; Mayles, P., A. E. Nahum and J. ate small NTCPs (on the order of 5% or less, see also the article C. Rosenwald. 2007. Handbook of Radiotherapy Physics: Theory Therapeutic effect), doubt has been expressed over the credibility and Practice, Taylor & Francis Group, London, U.K; Niemierko, of those points of the sigmoid response curve corresponding to A. 1997. Reporting and analyzing dose distributions: A concept of much larger NTCP values. equivalent uniform dose. Med. Phys. 24:103–110; Niemierko, A. Additionally, the Lyman model refers to the uniform irradia- 1999. A generalized concept of equivalent uniform dose (EUD). tion of an organ where part or all of it receives a uniform dose (Abstract). Med. Phys. 26:1100; Niemierko, A. and M. Goitein. and the rest none. This is clearly not the case in practice. Models 1991. Calculation of normal tissue complication probability and for inhomogeneous dose distributions include dose–volume his- dose-volume histogram reduction schemes for tissues with a criti- togram (DVH) based approaches to the calculation of the com- cal element architecture. Radiother. Oncol. 20:166–176; Yorke, E. plication probability such as those of Lyman and Wolbarst (1989) D., G. J. Kutcher, A. Jackson and C. C. Ling. 1993. Probability of and Kutcher and Burman (1989), generally known as the LKB radiation-induced complications in normal tissues with parallel model, and the equivalent uniform dose concept of Niemierko architecture under conditions of uniform whole or partial organ (1997). The latter was originally introduced for tumours but has irradiation. Radiother. Oncol. 26:226–237. been generalised to normal tissue, Niemierko (1999). It should be noted however that irrespective of what method is used to address Normal tissue dose the effect of inhomogeneous dose distributions, the resulting cal- (Radiotherapy) Radiation treatment inevitably affects normal culation of NTCP

still utilises empirical models that are limited tissues and the dose they receive is a contributing factor to the in that they lack a biological basis. Indeed, for most normal tissue induction of adverse effects. For further details on the response end-points, the biological interpretation of the sigmoid shape of of normal tissue to radiation see the articles on Adverse effects, the relationship is not obvious. Dose–response model and Tolerance. Improved formulae describing the dose–response for par- Related Articles: Adverse effect, Dose–response model, tial organ irradiation may result from a better understanding of Tolerance the underlying mechanisms of normal tissue complications. For example, the tissue architecture is thought to be important in Normal tissue dose–response determining the tolerance dose for partial organ irradiation (see (Radiotherapy) Radiation treatment inevitably affects normal tis- articles on Tolerance, Parallel organs and Serial organs). Yorke sues and their response depends on a number of factors including et al. (1993) have described the biological data needed as input the dose received, the volume of tissue irradiated, the tissue archi- for NTCP calculations for tissue with a parallel architecture, and tecture and the fractionation regime. For further details on the Niemierko and Goitein (1991) have done the same for tissue with response of normal tissue to radiation see the articles on Adverse a serial architecture. The relative seriality model of Kallman et al. effects, Dose–response model, Normal tissue complication prob- (1992) attempts to take explicit account of the architecture by the ability and Tolerance. introduction of a parameter s which reflects the degree to which Related Articles: Adverse effect, Dose–response model, the tissue is considered to be serial (s = 1) or parallel (s → 0). Fractionation, Normal tissue complication probability, Tolerance However, no model currently exists that fully accounts for all the factors that may influence radiation tolerance and all should be used with caution. A comprehensive review of NTCP models may Normal tissue reaction be found in Mayles et al. (2007, Chapter 36). (Radiotherapy) Radiation treatment inevitably affects normal tis- Abbreviations: DVH = Dose–volume histogram, LKB sue and so may cause radiation-induced adverse effects. For fur- = Lyman–Kutcher–Burman model, NTCP = Normal tissue ther details on radiation-induced normal tissue reaction, see the NTCP Dose (gray) Normal tissue toxicity 655 Nuclear activation analysis articles on Adverse effects, Dose–response model, Normal tissue Related Articles: Normalisation point, Treatment plan complication probability and Tolerance. normalisation Related Articles: Adverse effect, Dose–response model, Normal tissue complication probability, Tolerance Notification (Radiation Protection) National laws and regulations include the Normal tissue toxicity duty to notify to the national regulatory authority the intention to (Radiotherapy) Radiation treatment inevitably affects normal carry out a practice (see below and Further Reading). A person, or tissue and so may cause radiation-induced toxicity. For further their legal representative, intending to carry out any practice, as details see the articles on Adverse effects and Tolerance. defined by the regulatory authority, or other actions specified by the Related Articles: Adverse effect, Tolerance regulatory authority, shall notify the regulatory authority and sup- N ply the details and documents specified by the regulatory authority. Normalisation point A practice is any human activity that introduces additional (Radiotherapy) The normalisation point is a specific point to sources of exposure or exposure pathways or extends exposure to which the 3D matrix of dose values calculated by the TPS is nor- additional people or modifies the network of exposure pathways malised, that is a relative dose of 100% is assigned to this point. from existing sources, so as to increase the exposure or the likeli- There are two principally different ways of normalisation hood of exposure of people or the number of people exposed. point selection: Related Article: Regulatory authority Further Reading: IAEA (International Atomic Energy 1. The normalisation point position depends on the Agency). 1996. International Basic Safety Standards for details of the treatment plan or on the calculated dose Protection against Radiation and for the Safety of Radiation distribution. Sources. Safety Series No. 155, International Atomic Energy The isocentre, the intersection of several beam axes, Agency, Vienna, Austria. the minimum or maximum dose value in a slice or in a specified volume (entire volume, PTV, OAR, …) are Nozzle examples of such normalisation points. When two or (Radiotherapy) The Nozzle is the final piece of equipment tra- more competing treatment plans are then compared for versed by a proton beam, before that beam reaches the patient. a given patient, they are in general normalised to differ- It is mounted on the rotating gantry. In scanning proton therapy, ent normalisation points. the nozzle typically has pairs of scanning magnets which steer 2. The normalisation point position is independent of the the proton beam. In scattered proton therapy, the nozzle has a treatment plan. series of components including a range modulation wheel, a range In this approach, a fixed normalisation point is shifter, collimators and scatterers to conform the radiation field to defined in a clinically relevant location, the ‘ICRU ref- the tumour. The nozzle also monitors the beam uniformity, align- erence point’ being an example. The ICRU introduces ment and the amount of monitor units delivered. in its reports the concept of ‘ICRU reference point’ Related Articles: Gantry, Monitor unit. to which 100% relative dose should be assigned. The Further Reading: Leibel, S. A. and T. L. Phillips. Chapter ICRU point should be located firstly at the centre, or in 69 Proton Therapy. In Textbook of Radiation Oncology, 3rd edn., the central part, of the PTV, and secondly on or near the Radhe Mohan PhD, Michael T. Gillin PhD, Shiao Y. Woo MD, central axis of the beam(s). In this approach, competing FACR, Andrew K. Lee MD, MPH. treatment plans of a particular patient are normalised to the same point in the patient body. NPL (General) The National Physical Laboratory at Teddington is the Abbreviations: ICRU = International Commission on United Kingdom’s National Measurement Institute. Radiation Units and Measurements, OAR = Organ at risk, PTV Hyperlink: www .npl .co .uk = Planning target volume and TPS = Treatment planning system. Related Articles: Treatment planning system, Isodose curve, NRA Critical structures, Organ at risk, Planning target volume, (Radiation Protection) National regulatory authority. For further Prescribed dose information, see Regulatory authority. Further Readings: ICRU (International Commission on Related Article: Regulatory authority Radiation Units & Measurements, Inc.) 1993. Prescribing, recording and reporting photon beam therapy. ICRU Report 50, NRC Bethesda, MD; ICRU (International Commission on Radiation (Radiation Protection) See Nuclear Regulatory Commission Units & Measurements, Inc.) 1999. Prescribing, recording and (NRC). reporting photon beam therapy (Supplement to ICRU Report 50). ICRU Report 62, Bethesda, MD; ICRU (International Nuclear activation analysis Commission on Radiation Units & Measurements, Inc.) 2004. (Nuclear Medicine) Activation analysis depends on the observa- Prescribing, recording and reporting electron beam therapy. tion that almost every element can be converted into a radioactive ICRU Report 71, Bethesda, MD. isotope by bombardment with neutrons or charged particles. The induced radioactivity is highly specific of the radionuclides con- Normalised treatment dose (NTD) tributing to it. By studying the induced radioactivity, the constitu- (Radiotherapy) This refers to the treatment dose at the normalisa- ent elements in the sample may be identified and quantitatively tion point in a treatment plan. Please see the article Normalisation estimated. This apparently round-about procedure is, for more point for further information. than half of the chemical elements, the most sensitive analytical Nuclear binding energy 656 Nuclear forces technique available. The gain in sensitivity by comparison with Nuclear emissions other methods can be on the order of 106. (Radiation Protection) ‘Nuclear emissions’ is a collective term Related Articles: Induced radioactivity, Radioactivity, covering all particulate and electromagnetic radiation emitted Cyclotron, Nuclear reactor during radioactive decay or other nuclear event. The most com- Further Reading: Lenihan, J. M. A. 1967. Nuclear activation mon particles involved are alpha (α) and beta (β) particles, neu- analysis. In: Instrumentation in Nuclear Medicine, Vol. 1, ed., G. trons, protons and neutrinos. The term ‘gamma rays’ is used for Hine, Academic Press, New York, Chapter 13, pp. 309–325. the high energy electromagnetic radiation emitted from a nucleus during radioactive decay. X-rays and electrons can be emitted Nuclear binding energy from the atom involved as a consequence of interactions between N (General) Nuclear binding energy is the energy to separate a the nucleus or emitted particle and the electrons in the shells sur- nucleus into its unbound neutrons and protons. rounding the nucleus. In interactions of very high energy particles The mass of a nucleus is less than the sum of the masses of its with the nucleus other sub-atomic particles can be emitted. constituent protons and neutrons. The difference in mass (mass The term ‘nuclear emissions’ is sometimes loosely used instead defect) is the binding energy in accordance with Einstein’s for- of the term ‘nuclear discharges’ – the release of radionuclides into mula E = Δmc2, where E is the binding energy, Δm is the mass the environment (sea, air, river, etc.) from a nuclear facility. defect and c is the velocity of light. Related Articles: Alpha particle, Beta particle, Neutrinos, Nuclear binding energy is derived from the nuclear force. Nuclides, Radioactive decay The mass defect per nucleon varies with the mass of the nuclide is as shown in Figure N.19. Nuclear fission Related Articles: Nuclear force, Nucleus, Nucleon (General) Nuclear fission is when a nucleus splits into two sepa- rate nuclei accompanied by the emission of particles and gamma Nuclear chain reaction rays. In most cases the particles emitted are neutrons but in some (General) A self-propagating nuclear reaction, in which the occur- instances either alpha (α) particles or light nuclei are emitted. rence of a nuclear reaction causes further nuclear reactions. The Tertiary fission (three fission fragments) occurs very rarely. term is used predominantly to describe nuclear fission reactions, Nuclear fission can occur spontaneously in high mass radio- where a neutron is absorbed by a heavy fissile nucleus, causing the active nuclides. An example is californium-252, which has a nucleus to split into a number of fission fragments and releasing half-life of 2.6 years. It decays by alpha decay in 97% of transfor- energy (usually in the high MeV range). In a nuclear chain reac- mations and by spontaneous fission in 3%. tion, neutrons produced through the fission of one fissile nucleus Nuclear fission is normally induced by a neutron and an exam- go on to initiate further nuclear reactions by absorption in the ple is the induced fission of uranium-235 in a nuclear reactor. surrounding material. Various fragment pairs may be a result but an example is Uses in Medicine: Nuclear medicine – Nuclear chain reac- tions in a nuclear reactor play a part in the production of many 235 1 137 97 1 1 U + n ¾¾fissi¾on® Cs + Rb + n + n radionuclides used in nuclear medicine. The most common 92 0 55 37 0 0 nuclear medicine radionuclide, 99mTc, is generated by the decay of + Energy (about 200MeV) unstable molybdenum, which can itself be produced through the fission of uranium nuclei in a nuclear reactor, generating neutrons The sum of the masses of the product nuclei must be 234, whilst the which take part in further reactions: sum of the atomic numbers of the products must equal the atomic number of uranium for the equation to balance. Figure N.20 illus- trates the relative amounts of the products of the fission of U-235 U + n ® 236U* ® 134Sn + 99Mo + 3n + g – one fission fragment from each of the peaks. Over 60 primary products have been detected. Most of the primary products are themselves radioactive and decay into a wide range of nuclides. 10 Related Articles: Alpha particle, Nuclear reactor, Radioactive decay 8 Nuclear forces (General) The nuclear force acts between the nucleons (nucleon– nucleon interaction) in the nucleus of the atom and holds the 6 nucleons (protons and neutrons) together, being much stronger than the electrostatic repulsion between protons. The nuclear force, also referred to as the strong force, strong 4 nuclear force or residual strong force, is one of the four basic forces of nature – the others are gravity, the electromagnetic force and the weak nuclear force. 2 The force is short range, essentially acting only within the nucleus, and is mediated by mesons. A

full explanation and current knowledge of the nuclear force 10 50 100 200 is complex as the nucleons are themselves composed of quarks Mass number with gluons involved as the carriers of the force. Quantum chro- modynamics is a theory of the interaction between quarks and FIGURE N.19 Mass defect per nucleon as a function of mass number. gluons. Data have been smoothed. Related Article: Nucleons Mass defect per nucleon (MeV) Nuclear instability 657 N uclear magnetic resonance (NMR) 10 137 o many N and P 90Sr 99Mo Cs To 132Te N = Z 1 Too many N 10–1 Too few N N 10–2 10–3 Number of protons (N ) FIGURE N.21 Schematic diagram showing the relative position of 80 100 120 140 160 stable and unstable nuclides. The unstable radionuclides can be classified Mass number into three groups, those with excess neutrons, those deficient in neutrons and those with an excess of protons and neutrons. FIGURE N.20 Distribution of the yield of fission products from U-235 as a function of mass number (data smoothed). Note log scale for the ordinate. energy carried by the photoneutrons must be provided in addition to their binding energy, additional shielding is considered to be unnecessary for beam energies of 10 MeV and below. Nuclear instability Proton and Neutron Nuclear Interactions: At typical clini- (General) Nuclides which do not have a stable combination of cal energies, the cross-section for bremsstrahlung interaction neutrons and protons (the nucleons) transform spontaneously into of protons is suppressed due to their high rest mass, although a stable (or more stable) nuclide through the process of radioactive inelastic nuclear Coulomb interactions without bremsstrahlung decay. A stable nuclide is one that requires the addition of energy will still occur. Protons and neutrons can also interact via the to transform it into another nuclide. strong nuclear force, opening up additional interaction pathways Stable nuclides with low mass (apart from hydrogen) have with atomic nuclei. Neutrons, although electrically neutral, can similar number of neutrons and protons but with increasing be directly scattered by the nuclear potential of a nucleus, as can atomic number there is an increasing excess of neutrons over protons. neutrons and the ‘line of nuclear stability’ deviates from the Direct nuclear transmutations can also occur, in which a new N = Z line (Figure N.21). The reason for this lies in the char- compound nucleus is formed through the absorption of the proton acteristics of the nuclear force and the electrostatic repulsion or neutron into the nucleus. Similarly to photonuclear interac- between protons. The nuclear force has a very short range, less tions, the absorption of an energetic proton or neutron will often than that of a large nucleus, so additional neutrons are required result in the immediate ejection of one or more nucleons e.g. as in large nuclei to overcome the electrostatic repulsion between (p,n) or (n,p). the protons. Alternatively, radiative capture can occur in which case the Related Articles: Nuclear force, Nucleons, Radioactive decay, compound nucleus settles to its ground state through photon emis- Stable nuclei sion e.g. (n,y), with this new nucleus often remaining unstable to nuclear decay. Through such processes, components in the linac Nuclear interaction head in particular will become radioactive (termed as activation), (Radiotherapy) For radiation (whether electrically charged or potentially with long half-life, such that even with the linac de- neutral) incident on some medium, there exists the possibility for powered, suitable safety precautions will be required when car- direct interactions with that material’s atomic nuclei. rying out maintenance work on certain head components. Note Photonuclear Interactions: Photons with energy equal to or that photonuclear interactions can also lead to activation of linac in excess of the binding energy of a nucleon (approximately 6–7 head components (through loss of nucleons), and thus contribute MeV) can be absorbed by the nucleus, with subsequent ejection to these same safety considerations. of one or more nucleons, usually singular neutrons or protons at Further Readings: Mayles, P., A. Nahum and J. C. Rosenwald. radiotherapy energies. These interactions can be represented as 2007. Handbook of Radiotherapy Physics: Theory and Practice, (y, n) and (yp). Taylor & Francis, New York; Smith, F. A. 2000. A Primer in The former typically has a higher cross-section, particularly Applied Radiation Physics, World Scientific, Singapore. in high atomic number (Z) materials such as the lead and copper commonly found in linac heads. This can present a difficulty in Nuclear magnetic resonance (NMR) the shielding of linac bunkers, as additional and more complex (Magnetic Resonance) The phenomenon of nuclear magnetic res- materials/designs are required in order to sufficiently attenu- onance (NMR) was first detected in 1945 independently by Felix ate the emitted neutrons. In practice, however, since any kinetic Bloch and Edward Purcell (1,2). Both were in 1952 awarded the Fission yield (%) Number of neutrons (N ) Nuclear medicine 658 Nuclear medicine imaging Nobel prize ‘for their development of new methods for nuclear magnetic precision measurements and discoveries in connection therewith’. An NMR signal can be detected in odd–odd or odd–even nuclei (i.e. nuclei possessing a magnetic moment or spin) when a sample of the nucleus is exposed to a magnetic field. In the absence of an external magnetic field, all spin states have the same energy but when introduced into a magnetic field B0, a split- ting into different possible energy levels occurs, in parallel with N a precession of each spin around the z-direction (the direction of B0). As an example, the very simple hydrogen nucleus splits into two different energy levels (spin-up and spin-down) and the spins rotate at a frequency of approximately 42.6 MHz/T around z. In a sample, a large number of spins contribute to this process and since the rotations of the individual spins are not synchronised (a) (b) (not in phase), a net magnetisation vector (M0) is built up in the z-direction (see Net magnetisation). Since the rotation frequency FIGURE N.22 (a) Normal patient bone uptake in a bone scintigraphy. (the Larmor frequency) is well defined, it is possible to exchange The high uptake in the patients left arm is due to an ‘extravasal’ injection energy with the system of spins in a sample consisting of a cer- when administering the radioactive tracer. (b) Pathological uptake in a tain kind of nuclei, such as water. If a water sample is exposed bone scintigraphy. Patient demonstrates an increased relative uptake of to a magnetic field perpendicular to the B0 field (generated by a the tracer in metastases from a prostate cancer. transmitting coil) and rotating at the exact Larmor frequency (i.e. on resonance), energy is transferred to the sample. The macro- scopically observable effect in the sample is that M0 is ‘flipped’ radioisotope, the radiopharmaceutical can be used for imaging from its original z-direction down towards the transverse plane, and/or therapy. Radioisotopes with high penetrative radiation are where it can be detected in a receiver coil using the induction phe- suitable for imaging if the radiation can be acquired with an exter- nomenon. In the early experiments, the frequency of the transmit- nal imaging detector. For therapeutic purposes, like tumour tar- ter was swept over a frequency interval in order to establish the geting, radioisotopes with short range are suitable. An increased frequency at which the nuclear magnetic resonance phenomenon uptake in an organ or structure might be an indication of patho- occurred. logical condition (see the two examples in Figure N.22). The NMR technique has rapidly evolved into an important tool Nuclear medicine examinations are used in a variety of fields, within chemistry and biochemistry, but has also given rise to the for example oncology and neurology. In oncology 18F-FDG-PET is clinically well-established magnetic resonance imaging (see MRI) used world wide to locate tumours with excessive glucose metab- and magnetic resonance spectroscopy techniques. Spectroscopic olism. SPECT examinations provide valuable information when techniques in clinical as well as non-clinical applications rely on trying to diagnose diseases like Alzheimer’s and Parkinson’s at the fact that the Larmor frequency of a specific nucleus depends an early stage. not only upon the main magnetic field B0 from the magnet, but also Abbreviations: CT = Computed tomography, MRI = Magnetic upon the immediate atomic and molecular structures surrounding resonance imaging, PET = Positron emission tomography and the nucleus since particles such as electrons in the near vicinity SPECT = Single photon emission computed tomography. of the nucleus have a screening effect. Hence, NMR-detectable nuclei of the same type but situated at different places in a larger Nuclear medicine imaging molecule show slightly different Larmor frequencies and the cor- (Nuclear Medicine) Nuclear medicine is a specific branch in responding NMR signals can therefore, if frequency resolved medicine that uses radioactive substances for therapy and imag- (spectroscopically detected), be used to deduce, for example the ing. Nuclear medicine imaging refers to the technique of using shape and orientation of the molecule. radionuclides in order to study the functionality of biological sys- Related Articles: Magnetic resonance imaging (MRI), Net tems rather than provide morphological information. Because the magnetisation nuclear medicine images represent the biological process at cellu- Further Readings: Bloch, F., W.W Hansen and M. Packard. lar and sub-cellular level it is also considered a part of molecular 1946. The nuclear induction experiment. Phys. Rev. 70:474–485; imaging. The fundamental idea of nuclear medicine imaging is Purcell, E.M., H. C. Torry and R.V. Pound. 1946. Resonance to label a biological substance with a radioactive isotope, a radio- absorption by nuclear magnetic moments in a solid. Phys. Rev. pharmaceutical. When radioisotopes used in medical imaging 69:37–38. decay, photons are emitted as a direct or indirect (i.e. following positron annihilation) result of the decay. The radiopharmaceuti- Nuclear medicine cal is introduced into a biological system where the radiophar- (Nuclear Medicine) Nuclear medicine is a special discipline in maceutical kinetics (i.e. uptake, retention and wash out) can be medicine and medical imaging and it involves the use of specific studied using a nuclear medicine imaging modality like PET, radioactive isotopes. Nuclear medicine imaging differs from most scintillation camera and SPECT to image the emitted photons. other imaging modalities like MRI and CT because the images In diagnostics one takes advantage of the fact that the kinet- primarily show the physiological function of biological processes ics of different substances differs between pathologic tissue and rather than morphological structures of the body. healthy tissue. For example, an increased local uptake of the The fundamental idea of nuclear medicine is to target specific radiopharmaceutical 18F-FDG is an indication of tumours growth. biological processes with radiopharmaceuticals (pharmaceuti- FDG is a glucose analogue that is retained in cells with high cal + radioisotope). Depending on the decay properties of the metabolism, for example tumour cells. Nuclear reactor 659 Nucleus The two most common radionuclides used for nuclear medi- Transformation can be spontaneous, if the nuclide is unsta- cine imaging are 18F (preferably PET scanner) and 99mTc (scintil- ble, through the process of radioactive decay, or induced when a lation camera and SPECT scanner). 99mTc is the most common nuclide is bombarded by a particle of sufficient energy or by very radiopharmaceutical and it can be ordered with a range of pre- high energy gamma rays. pared pharmaceutical kits, each designed to participate in a Related Articles: Nuclide, Radioactivity specific biological process. Radiopharmaceuticals can be admin- istered intravenously as a liquid or in aggregate form, via inges- Nucleon tion or inhalation of a gas or aerosol. (General) Nucleon is the collective term applied to protons and Newer acquisition techniques allow cardiac gated imaging. neutrons. These are now considered as composite particles, called Abbreviations: PET = Positron emission tomography and baryons, composed of quarks and gluons but previously were pre- SPECT = Single photon emission computed tomography. N sumed to be elementary particles. Atomic nuclei are composed of Related Articles: PET, SPECT, Cardiac gating protons and neutrons. Related Articles: Baryons, Neutrons, Nuclear forces, Nuclear Nuclear reactor binding energy, Nucleus, Protons (General) A nuclear reactor is a device in which a nuclear chain reaction is initiated, controlled and sustained. The chain reaction Nucleus is induced by neutrons and to achieve stability, only one neutron (General) The nucleus is the central structure of an atom and is on average, emitted during a fission event, must induce a further composed of protons and neutrons, which are

collectively known fission. as nucleons. The nucleus is surrounded by electrons which cir- The main components of a nuclear reactor are as follows: culate around it in orbits, called shells. The nucleus is positively Fuel: The most common chain reaction is the induced fission charged due to the protons and is extremely small compared with of uranium-235, and the uranium is normally in the form of ura- the overall size of the atom. The nucleus has a diameter on the nium oxide, encased in tubes to form the fuel rods. order of 10−14 m compared with 10−10 m for an atom. The nucleus Moderator: The cross section for fission of U-235 is low for is held together by the nuclear force that is much greater than the the fast neutrons emitted during the fission of U-235 but high for electrostatic force repelling the protons. slow neutrons. To slow or ‘moderate’ the neutrons, a material The simplest nucleus is the nucleus of the hydrogen atom that which does not capture significant numbers of neutrons is incor- has one proton and no neutron. Atoms with a small mass have porated into the design of the reactor. Water, heavy water, D2O (D a similar number of protons and neutrons but with increasing = Deuterium), and graphite are suitable and widely used. atomic number there is an increasing excess of neutrons over pro- Control rods: These are made of neutron-absorbing materials, tons (Figure N.23). such as boron or cadmium, to control the rate of the chain reaction A simple model of the nucleus is to consider the nucleons by insertion or withdrawal. arranged in shells, somewhat analogous to the arrangement of the Coolant: A liquid or gas to transfer the heat from the core electrons outside the nucleus but only in terms of energy levels. to outside the reactor. In reactors used for power generation the Electron shells are full when they have 2, 8, 18 and 32 in the K, coolant is used, directly or indirectly, to generate steam to for the L, M and N shells respectively. In the nucleus the corresponding turbines generating electricity. numbers are 2, 6, 12 and 8; the protons and neutrons have sepa- Core: It contains the fuel, the moderator, the control rods and rate shells. When a nucleus has only filled shells an extra nuclear the coolant. stability is conferred on it, an example is the helium nucleus, with Pressure vessel: The vessel usually made of steel or concrete, two neutrons and two protons. It is the special stability of the containing the core. helium nucleus which explains why it is ejected as a particle in Shield and containment: The structure, usually concrete, alpha decay (Figures N.24 and N.25). to contain the pressure vessel and to provide a shield from the Normally nuclei are in the ground state but can, in special situ- intense radiation. ations, exist for a limited time in an excited state. When a nucleus Most reactors are designed as a heat source to produce elec- in an excited state falls to the ground state, a gamma ray is emit- tricity but other uses include the production of fissile mate- ted. In most cases the lifetime in an excited state is in the range rial for making nuclear weapons. For the medical physicist 10−6–10−4 s but in some situations it can be minutes or hours – in and others involved in the use of radioisotopes in medicine, these cases the levels are known as metastable states. the main interest in nuclear reactors is their use as an intense source of low- and high-energy neutrons for the production of radionuclides. The term ‘nuclear reactor’ may also be used for a device in which a nuclear fusion reaction can be initiated, controlled and sustained – a fusion reactor. Related Articles: Cross-section, Nuclear fission, Reactor Nuclear Regulatory Commission (NRC) (Radiation Protection) The Nuclear Regulatory Commission is an independent agency of the USA government. Hyperlink: www .nrc .gov Hydrogen Helium Lithium Proton Neutron Electron Nuclear transformation (General) Nuclear transformation is the change of one nuclide FIGURE N.23 Simplistic representation of the atoms of Hydrogen (1 1H), into another with a different number of protons or neutrons. Helium ( 4 2 He) and Lithium ( 7 3Li). Nuclide 660 Nyquist theorem Nude mouse (Nuclear Medicine) A mouse that lacks a thymus and hence has no mature T cells. Xenografts of human origin, such as tumours N can easily grow in such immune-deficient mice. There also exist nude rats which also are immune-incompetent. These animals L are mostly used as experimental animals especially in cancer K research in the development of new therapeutic drugs. The nude mouse and rat is hairless, hence the term ‘nude’. M N Number of excitations (NEX) (Magnetic Resonance) To improve the signal-to-noise ratio (SNR) in magnetic resonance imaging (MRI) it is possible to use two or more excitations/acquisitions of the same k-space line and aver- FIGURE N.24 age the signals together to yield an improved SNR image. The Simplistic representation of an atom showing the central nucleus surrounded by the electron shells. number of used averages is often referred to as the number of excitations (NEX) or the number of acquisitions (NSA). The SNR in the image will be improved with the square root of the number of excitations; however, the scan time will be prolonged with a factor equal to the NEX value. 120 Related Article: Signal-to-noise ratio (SNR) Nyquist frequency 100 (Magnetic Resonance) The Nyquist frequency is the maximum frequency νmax that is correctly encoded by discrete sampling at a frequency νsample: 80 νmax = νsample/2 for real (one channel) signals (as stated by the Nyquist theorem). 60 Quadrature detection yields also negative frequencies of νmin = – νsample/2 Frequency components exceeding the Nyquist frequency 40 are folded back (aliasing). Related Article: Nyquist theorem 20 Nyquist limit (Ultrasound) In a sampling system, the Nyquist limit is always half of the sampling frequency. Frequencies above this limit will 20 40 60 80 not be represented correctly but will appear as a lower frequency. Number of protons (N ) A frequency which, for example is 0.6 fs, where fs is the sampling frequency, will appear as 0.4 fs. The term ‘folding’ is sometimes FIGURE N.25 Relationship between the number of protons and the used, which refers to that the spectrum is ‘folded’ around fs. In number of neutrons for all stable toms. The nuclei would lie on the line at Doppler measurements, ‘folding’, or aliasing, appears somewhat 45° if they had equal number of the two nucleons. different due to quadrature demodulation, see Aliasing. Nyquist theorem Related Articles: Alpha decay, Gamma rays, Protons, (Magnetic Resonance) The Nyquist theorem (aka Nyquist– Neutrons, Nucleons, Stable nuclei Shannon theorem) states that a band-limited signal of ν < νmax has to be sampled at a sampling frequency νsample > 2 νmax. Nuclide In other words, digital sampling at a given sampling frequency (General) The term nuclide is used to denote a species of atom (or rate) νsample can correctly encode frequencies below νsample/2, with a specific atomic number Z, neutron number N and in a the so-called Nyquist frequency. defined nuclear state. Usually it is assumed that the nucleus is in Complex signals acquired by quadrature detection in MRI and the ground state and the same symbolism is used as for an atom MRS allow negative frequencies to be discerned, thus yielding a – for example, 131 53 I represents a nuclide/atom of iodine with total total bandwidth bw = νsample = 1/τsample, that is the inverse of the mass (represented by the symbol Z) of 131 and with 53 protons dwell time. Higher or lower frequencies will be folded back below (represented by the symbol N). If the nucleus is in a metastable the Nyquist frequency (aliasing). state then this represented by the superscript m, thus 99 43Tc is a dif- The Nyquist frequency and theorem are named after Harry ferent nuclide from 99 43Tcm. Nyquist (1889–1976), Swedish-American engineer. Related Articles: Atom, Nucleus Related Article: Nyquist frequency Number of neutrons (N ) O Object–film distance Oblique incidence (Diagnostic Radiology) The term object–film distance (OFD) is (Radiotherapy) Typically, radiotherapy beam data (e.g. depth used in planar x-ray radiology to describe the distance between dose, etc.) is measured with the beam incident at right angles the central axes of the object being imaged and the surface of to a flat and uniform surface. However, in reality, for patient the film. The distance influences parameters such as geometric treatments this situation is unlikely, and so this data cannot be magnification and image distortion. In modern digital radiology directly employed to calculate the distribution of dose within where film is no longer used, the distance between the object and the body. O the detector is called the object–detector distance (ODD) or the There are two main approaches to compensate for this change: object–image receptor distance (OID) (Figure O.1). (a) the use of wedges, bolus, or compensators, and (b) a calcu- Related Articles: Focal–film distance (FFD), Target–film dis- lation technique. Calculation techniques vary from the complex tance (TFD), Magnification (requiring computer processing) to those which may be calculated manually. Some examples of simple manual calculation correc- Object image receptor distance tion techniques are the effective SSD method, the TAR method, (Diagnostic Radiology) See Target–film distance and the isodose shift method. (For further detail on these cor- rection methods, see Further Readings.) These methods apply for Object recognition angles of incidence up to 45°. (General) Object recognition is the ability of a computer program Abbreviations: SSD = Source-to-skin distance and TAR = to recognise a particular object within an image. In medical imag- Tissue air ratio. ing, this may be a particular organ such as the liver, lungs, kidneys, Related Articles: Electron oblique incidence, Incidence angle, etc. This can be achieved by comparing the features of this object Obliquity, Obliquity effect, Bolus, Compensator, Wedge (such as shape, size, tracer uptake) with those of a known object. Further Readings: Podgorsak, E. B. 2003. Review of Radiation Oncology Physics: A Handbook for Teachers and Students, International Atomic Energy Agency (IAEA), Vienna, Object scatter events Austria; Williams, J. R. and D. I. Thwaites. 2000. Radiotherapy (Nuclear Medicine) An object scatter event refers to a photon Physics in Practice, 2nd edn., Oxford University Press, Oxford, scattered in the object (i.e. patient or phantom) prior to detection. UK. Consider the situation where a photon is emitted with a direction non-parallel to the collimator holes. In an ideal situation, the pho- ton would travel through the object and be attenuated in the col- Obliquity limator. But the photon can undergo Compton scatter and change (Radiotherapy) This is the situation where the patient surface its propagation direction so that it is parallel to the collimator is not perpendicular to the incident beam. The obliquity will holes (Figure O.2). therefore cause the dose distribution to be distorted due to the Object scatter events will cause a loss in contrast because of varying distance at which the beam is incident on the skin the mis-positioning of events. The mis-positioning can be sev- surface. eral centimetres from the original event location. The photon Related Articles: Oblique incidence, Electron oblique inci- loses energy in the scatter process and it is therefore possible to dence, Incidence angle, Obliquity effect separate the object scatter events using an energy discrimination window. A scintillation camera with good energy resolution is Obliquity effect therefore important. (Radiotherapy) See Oblique incidence Further Reading: Simon, R. C., J. A. Sorenson and M. E. Related Articles: Oblique incidence, Electron oblique inci- Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, dence, Incidence angle, Obliquity Philadelphia, PA, pp. 222–223. Occupancy factor Oblique (Radiation Protection) When designing the shielding for a radia- (General) There are a series of terms used to describe the position tion facility dose, constraints are used. These determine the of an individual when undertaking different imaging examina- annual dose in any area outside the facility. Currently, an annual tions. Oblique means turned usually to a specific angle. For exam- dose constraint of 0.3 mSv is adopted for an area outside the ple, a 45° angle for an oblique chest projection. facility. Related Article: Patient position The dose constraint can be modified (increased) under cer- tain conditions. For example, if

the area adjacent to the radiation Oblique imaging facility is a controlled area, the dose constraint may be increased. (Magnetic Resonance) Oblique imaging is performed in non- Another permissible modification is to consider occupancy. orthogonal planes (or slice orientations) where required by anat- Some areas of building, e.g. corridors, toilets, etc., would not be omy (e.g. of the foot or heart). expected to have full occupancy. Thus an occupancy factor may 661 Occupational dose limits 662 Occupational exposures 1. Effective dose to whole body: 20 mSv/year averaged X-ray tube housing over a period of 5 years with a maximum effective dose Focal spot of 50 mSv in any single year Target 2. Effective dose to the embryo or foetus: 1 mSv Internal 3. Annual equivalent dose to the lens of eye: 150 mSv collimators 4. Annual equivalent dose in the skin and extremities: 500 mSv Focal–film distance (FFD) or For occupational exposure to radionuclides, with risk of internal Target–film distance contamination, there are tables (for the various radioisotopes) (TFD) which give ingestion and inhalation dose coefficients. This means Focal–object distance (OFD) that it is possible to evaluate the committed effective dose per unit intake via ingestion corresponding to different gut transfer O factors, for various chemical forms, and the committed effective dose per unit intake (via inhalation) for the lung absorption, con- Object sidering also the component cleared from the lung to the gastro- intestinal track. Anti-scatter Object–film grid distance (OFD) In the case of apprentices from 16 to 18 years old, who are under training with activity involving exposures, the following Film or detector limits shall be respected: an effective dose to whole body of 6 surface mSv/year, an annual equivalent dose to the lens of eye of 50 mSv and an annual equivalent dose to the extremities and skin of 150 FIGURE O.1 Distance metrics in planar radiography. mSv. There might be ‘special circumstances’ under which a tempo- rary change in dose limitation is required and approved. In these cases, the 20 mSv shall be averaged over a period up to 10 years, Crystal with the maximum dose per year remaining 50 mSv (circum- stances shall be reviewed when any worker reaches 100 mSv). Collimator Alternatively, the annual limit shall not exceed 50 mSv with a temporary change period not exceeding 5 years. Dose Limits for Public Exposures: The averaged estimated Object doses to critical groups of the public, attributable to practices shall not exceed the following limits: (1) effective dose to the whole body: 1 mSv/year, in special circumstances averaged over a period of 5 years with an effective dose of 5 mSv/year, (2) annual FIGURE O.2 Compton scatter in object prior to detection. The dotted equivalent dose in the lens of the eye of 15 mSv and (3) annual line represents the true line of response and the black arrows are the pho- equivalent dose to the skin of 50 mSv. ton pathway. Object scatter events lead to less contrast in areas with high Dose Limits for Comforters and Visitors of Patients: The radioisotope concentration. dose of any voluntary comforter or visitor of patients shall be con- strained so that it is unlikely to exceed 5 mSv during the period of diagnostic investigation or treatment. The dose to children visit- be assumed that increases the planning dose constraint. The fol- ing patients who have ingested radioactive material shall be lim- lowing occupancy factors are in general use: ited to less than 1 mSv. Related Articles: International Commission for Radiological Corridor: 20% Protection (ICRP), Occupational exposure Toilet (washroom): 10% Further Readings: IAEA. 1996. International Basic Safety Stairway: 5% Standards for Protection against Radiation and for the Safety Thus, for example, if the area adjacent to the radiation facility was of Radiation Sources, Safety Series No. 115, International a corridor, the dose constraint for that area could be increased by Atomic Energy Agency, Vienna, Austria; ICRP. 1991. 1990 a factor of 100/20 (5). Instead of 0.3 mSv, a planning dose con- Recommendations of the International Commission on straint of 1.5 mSv could be used. Radiological Protection, ICRP Annual Report, ICRP Publication Related Articles: Dose limits, Dose constraints 60, ICRP 21. Occupational dose limits Occupational exposures (Radiation Protection) The occupational dose limits are set by the (Radiation Protection) Exposures of workers incurred in the International Commission for Radiological Protection (ICRP). course of their work are considered occupational exposures. The following dose limits apply to exposure from practices, with Responsibilities: Different main responsibilities are attrib- exception made for exposures from medical practices (diagnostic uted to registrants, licensees and employees. Licensees shall and therapy) and from natural sources. ensure that for all workers (1) occupational exposures are lim- Dose Limits for Occupational Exposures: Occupational ited and optimised; (2) suitable and adequate facilities, equipment exposures of any worker shall be controlled in order not to exceed and services for protection are provided; (3) appropriate protec- the following limits: tive devices and monitoring equipment are provided and properly OER (Oxygen enhancement ratio) 663 On-line portal imaging used; (4) appropriate training is provided as well as periodic depending on the k-space trajectory. In general, spin echoes retraining and updating; (5) adequate records are maintained. reduce off-resonance effects in comparison to gradient echoes. Employees shall follow any applicable rules for protection, use In ultrafast imaging (where EPI-readouts are particularly sensi- the monitoring devices, protective equipment and clothing pro- tive), reducing the readout time will reduce these effects. Other vided properly and co-operate with the licensee with respect to sequences vulnerable to off-resonance effects are so-called bal- protection. anced gradient echo sequences and spectroscopic sequences, both Conditions of Service: Special compensatory arrangements requiring adequate shimming for achievement of optimal results. (preferential treatment, salary, vacation, etc.) shall be neither Related Article: Eddy currents granted nor used as substitutes for the provision of proper protec- tion and safety. The condition of work of pregnant women should Offset be adapted in order to ensure the compliance with the dose limit (General) In electronics, offset means offsetting of a signal from for the embryo/foetus. zero. It is considered to be direct current voltage measured at the No person under age 16 years shall be subject to occupa- input or output of an electronic device. In operational amplifiers, tional exposures, and no person under the age of 18 years shall input offset voltage and input offset current are specified for each be allowed to work in a controlled area unless supervised (under type. O training). Suitable alternative employment should be provided to Direct current (DC) offset is usually undesirable, but can be workers who may no longer continue in employment involving eliminated by trimming or by auto-zero amplifiers in electronic occupational exposure. devices. See Dose limits for the specification of the values. Further Reading: IAEA. 1996. International Basic Safety On-line portal imaging Standard for Protection against Ionizing Radiation and for the (Radiotherapy) On-line portal imaging involves the use of a por- Safety of Radiation Sources, Safety Series No. 115, International tal imaging device to make measurements of treatment accuracy, Atomic Energy Agency, Vienna, Austria. in terms of patient positioning, during a fraction of radiotherapy Hyperlink: http://www .IAEA .org and to act on that information to reduce errors for that treatment fraction, hence the name ‘on-line’. Often this is achieved by OER (Oxygen enhancement ratio) delivering a small test dose at the start of the fraction, imaging (Radiotherapy) See Oxygen enhancement ratio (OER) it, determining the set-up correction needed and making that cor- rection before delivering the rest of the treatment. Oersted On-Line versus Off-Line Set-Up Correction: The alternative (General) Oersted (Oe) is a cgs unit for magnetic field strength H. to on-line imaging is to image a treatment fraction and determine 1 Oe = 1000/4π AT/m (cgs – the centimetre, gram second system the set-up error off-line, after the fraction has finished, and cor- of units, the cgs is now replaced by SI). rect subsequent fractions on this basis. In practice several frac- Oe is related to the cgs unit gauss (G) through the magnetic tions are imaged before any corrective action is taken for the permeability μr (relative to permeability of vacuum): off-line process so that random errors in each image are averaged out and the systematic error is determined and corrected. Various B(G) = mr * H (Oe) empirically derived methods exist for deciding how many images to take and how often to take those images to arrive at the most accurate treatment with the minimum workload for off-line set-up The Oersted is named after the Danish physicist Hans Christian correction. Oersted. Errors and Margins: The overall error in a treatment has Related Article: Gauss many contributing sources. The set-up errors detected with on-line portal imaging arise from several of these sources. Radiotherapy OFD (object–film distance) treatment errors may be described in terms of random and sys- (Diagnostic Radiology) See Object–film distance tematic components. Crudely, the random component detected in portal imaging is the day-to-day variation in set-up and the sys- Off-resonance tematic component is the difference between the average position (Magnetic Resonance) Off-resonance occurs when spin-iso- at treatment and that at planning. On-line correction addresses chromats have a frequency different from that expected, i.e. the both types of errors, whereas off-line correction addresses only Larmor precessional frequency. For example, the Larmor fre- systematic errors. quency of protons in fat differs from that of protons in water due Various numerical simulations have been carried out to deter- to different magnetic shieldings. However, even if the Larmor mine the treatment margins necessary to achieve a certain prob- frequencies of all protons were equal, differences in magnetic ability of target coverage as a function of the size of the two types susceptibility throughout the object would result in local mag- of errors. These reveal the effects of systematic errors to be three netic field gradients within the object, leading to differences in times more important than random errors. the Larmor frequency. Other sources of off-resonance effects can be eddy currents arising in the conducting material of the scanner when gradients are switched on and off, or concomitant gradi- EXAMPLE: MARGIN FORMULAE ents (Maxwell terms), e.g. a magnetic field gradient applied in the z-direction results in additional gradients in the x- and y-direction Stroom et al. derived the following margin formula for as well since all magnetic fields must obey Maxwell’s equation. 99% of the CTV to receive 95% dose: The effects of off-resonance include spatial distortions (e.g. chemical shift artefacts or artefacts occurring in the surrounding M = 2S + 0.7s of metal implants), signal losses (c.f. T2* relaxation) and blurring, One-day protocol 664 Open field laboratory to alternate between stress/distribution 201Tl and stress/ Van Herk et al. calculated margins for a minimum CTV rest 99mTc-Sestamibi or 99mTc-Tetrofosmin. This sequence is also dose of 95% in 90% of patients: a good choice for patients with low likelihood of coronary artery disease, since the second study is not always needed if the stress M = 2.5S + 0.7s study is normal. The disadvantage of this sequence, on the other hand, is that the count rates in the stress image set are not ade- where quate for accurate assessment of defect reversibility. If the stress/ M is the margin size rest sequence is chosen, a longer interval between the two injec- Σ is the size of the systematic error tions is required so that the background contribution from the first σ is the random error set of images is minimised. Related Article: Two-day protocol Further Readings: Taillefer, R., A. Gagnon, L. Laflamme, J. Abbreviation: CTV = Clinical target volume. Gregoire, J. Leveille and D. C. Phaneuf. 1989. Same day injec- Related Articles: Electronic portal imaging, Electronic portal tions of Tc-99m methoxy isobutyl isonitrile (hexamibi) for myo- O imaging device, Treatment verification, Clinical target volume cardial tomographic imaging: Comparison between rest-stress (CTV) and stress-rest injection sequences. Eur. J. Nucl. Med. 15(3):113– Further Readings: van Herk, M., P. Remeijer, C. Rasch and 117; Van Train, K. F., J. Areeda, E. V. Garcia, C. D. Cooke, J. J. V. Lebesque. 2000. The probability of correct target dosage: Maddahi, H. Kiat et al. 1993. Quantitative same-day rest-stress Dose-population histograms for deriving treatment margins in technetium-99m-Sestamibi SPECT: Definition and validation of radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 47:1121–1135; stress normal limits and criteria for abnormality.

J. Nucl. Med. Stroom, J. C., H. C. J. de Boer, H. Huizenga and A. G. Visser. 34(9):1494–1502. 1999. Inclusion of geometrical uncertainties in radiotherapy treat- ment planning by means of coverage probability. Int. J. Radiat. One-way rectifier Oncol. Biol. Phys. 43:905–919. (Diagnostic Radiology) One-way rectification (also known as half-way rectification) is an electric power supply which uses One-day protocol either the positive or the negative AC wave. Half-way rectifi- (Nuclear Medicine) The one-day protocol refers to myocardial ers use either one diode (single-phase power supply) or three perfusion imaging where the rest and stress studies are performed diodes (three-phase power supply). This type of rectification is on the same day. relatively inefficient, because it blocks half of the input signal Irrespective of the stress/rest order, the first study requires an (Figure O.3). injection with a low activity of 300–400 MBq 99mTc-Sestamibi or Related Article: Rectifier 99mTc-Tetrofosmin followed by a higher activity of 800–900 MBq to overcome the residual activity from the first study. Opacity The one-day low-dose rest/high-dose stress protocol is the (Diagnostic Radiology) Opacity is the characteristic of an object most commonly used 99mTc-based SPECT acquisition protocol. or substance to reduce the penetration or passage of radiation such Advantages of choosing the one-day protocol are as follows: as light. It is the inverse of transparency. The degree of opacity can be measured and expressed in terms of the optical density. 1. The stress injection can be administered immediately Optical density is used extensively to express the ‘opacity’ of after the acquisition of the rest image and may be con- areas within a radiographic film. venient for patients. 2. It provides quicker diagnostic information. Open field 3. A true rest study may improve ability to detect defect (Radiotherapy) Open field is the term generally applied to a field reversibility due to higher administered activity used used with no beam modifier present in the beam, such as wedge, for stress study resulting in higher activity ratio in the compensator, tray or shielding blocks. The open field obtained normal to abnormal myocardial territories. with secondary collimator or multileaf collimator may be square, rectangular or irregular. The disadvantage is that there might be a potential reduction in Related Articles: Secondary collimator, Multileaf collimator, the defect contrast due to residual activity from the initial study. Block design, Wedge, Compensator, Irregular field The one-day low-dose stress/high-dose rest sequence, on the Further Reading: ESTRO. 1997. Physics for Clinical other hand, has the advantage of requiring image acquisition Radiotherapy, Booklet No.3: Monitor Unit Calculation for High durations similar to those used for 201Tl, making it easy for a Energy Photon Beams, ESTRO, Neuilly-sur-Seine, France. RL 0 0 FIGURE O.3 Example of a one-way rectifier. Open magnet 665 Optical density Open magnet R3 (Magnetic Resonance) The open magnet design of MR scanners + features a magnet top and bottom and is open on the sides (see Vout Figure E.19 in the Electromagnet article). This design decreases V R in 1 the risk of claustrophobia and allows parents to stay close to very – R2 young patients, obtaining their collaboration. Open magnets are limited in field strength; the maximum is obtained in scanners based on superconducting technology and it FIGURE O.6 Electrical schematic of inverting amplifier. does not exceed 1 T. The increased bore dimension of the conven- tional MRI systems and the clinical request of having a magnetic field equal to 1.5 T or higher results in a reduced presence of open magnet systems in the marketplace. Related Articles: Magnet, Electro-magnet, Permanent mag- V – R GR in net, Resistive magnet, Superconductive magnets – V + V in R out O Open-core transformer + (Diagnostic Radiology) See Transformer R Vo = G(V + in – V – G in ) Operating mode (Diagnostic Radiology) See Acquisition modes for digital image OV Operational amplifier (General) An operational amplifier is high-gain DC-coupled FIGURE O.7 Differential amplifier. electronic voltage amplifier. It amplifies the voltage difference between its differential inputs (+ input and − input). It has a sin- gle voltage output. Operational amplifiers have very high input in many standard IC housings, typically in 8-pin dual in-line impedance at input terminals and very low output impedance packages. (Figure O.4). Operational amplifiers are used in a number of different con- Optical density figurations, but primarily as amplifiers. The voltage gain of an (Diagnostic Radiology) Optical density is the opaqueness of operational amplifier is determined by negative feedback. Typical translucent film and has assigned numerical values related to the configurations of DC voltage amplifiers are shown in Figures O.5 amount of light that penetrates the film. Increasing film density and O.6. decreases light penetration. The relationship between density val- The gain of the non-inverting amplifier is determined by G = ues and light penetration is exponential, as shown in Figure O.8. 1 + R2/R1 and the gain of the inverting amplifier by G = −R2/R1. A clear piece of film that allows 100% of the light to penetrate In Figure O.7, the differential amplifier configuration is pre- has a density value of 0. Radiographic film is never completely sented. Differential amplifiers are often used in sensing circuits to clear. The minimum film density is usually in the range of 0.1–0.2 amplify low-level voltage difference signals whilst rejecting any density units. This is designated ‘the base plus fog density’ and is common voltage. the density of the film base and any inherent fog not associated Operational amplifies are most widely used analog integrated with exposure. circuits (IC) with applications in industrial, professional, con- Each unit of density decreases light penetration by a factor of sumer and medical electronic devices. They are encapsulated 10. A film area with a density value of 1 allows 10% of the light to penetrate and generally appears as a medium grey when placed on a conventional viewbox. V+ +V A film area with a density value of 2 allows 10% of 10% cc (1.0%) light penetration and appears as a relatively dark area when + Vout viewed in the usual manner. With normal viewbox illumination, V– – –Vcc Optical density 0 0.5 1 1.5 2 2.5 3 FIGURE O.4 Operational amplifier symbol. Vin + Vout R1 100 (%) 10 (%) 1 (%) 0.1 (%) – R Light penetration 2 FIGURE O.8 Optical density and light penetration. (Courtesy of FIGURE O.5 Electrical schematic of non-inverting amplifier. Sprawls Foundation, www .sprawls .org) Optical distance indicator 666 Optically stimulated luminescence it is possible to see through areas of film with density values of up Optical transfer function to approximately 2 units. (Diagnostic Radiology) The optical transfer function is the Optical densities of 3 (or even the maximum 4) cannot be dis- Fourier transform of the (real) point-spread function. Its magni- tinguished visually. They all appear completely black, but can be tude is the modulation transfer function. The point-spread func- measured with optical densitometer. tion has the advantage of visualising the effect of an imaging well, but is less convenient to express the effect of a sequence Optical distance indicator of components. The Fourier transform of a convolution being a (Radiotherapy) The optical distance indicator (ODI) allows the product, the OTF makes these combinations very easy in the fre- operator to know the distance from the source of radiation to quency domain, the same holds for its magnitude, the MTF. The the skin (or phantom) surface. It is commonly a scale pattern OTF of the radial example shown in the PSF article is simply the projected by a light source in the head of the linac, and is tilted function with 1 at the sampled values in frequency domain, zero at an angle such that the appropriate point on the scale is in focus otherwise. on the central axis at the corresponding SSD. Therefore, at the The difference between OTF and MTF is the difference O isocentre (SSD = 100 cm) the 100 indicator is focussed on the between a complex number and its magnitude: central axis, while at 10 cm above and below the isocentre the 90 and 110 indicators, respectively, are focussed on the central axis. MTF(f ) = OTF(f ) or OTF(f ) The scale typically runs from around 80 to 150 cm, to allow for a range of different SSD set-ups, e.g. isocentric treatments and = MTF(f )* exp(i * phase(f )) extended SSD treatments where fields of size greater than 40 cm are needed. The level of agreement of the ODI with SSD namely the phase. This does not always play an important role, (and couch vertical movement) is usually of the order of 3 mm but in MRI, e.g. the imaging system can introduce phase changes, towards the extremes of the range and 1 mm at the isocentre and the PSF is not necessarily real, in which case phases must be (100 cm). taken into account. Related Article: Source-to-skin distance (SSD) Abbreviation: OTF = Optical transfer function. Related Articles: Point-spread function, Modulation transfer Optical radiation hazard function (Non-Ionising Radiation) Optical radiation can only penetrate in Further Reading: Harrison, H. B. and K. K. Myers. 2004. the skin up to a few millimetres, and therefore only superficial Foundations of Image Science, Wiley Series in Pure and Applied parts of the body, such as the skin and the eyes are affected by Optics, Wiley-Interscience Publication, Hoboken, NJ. it. (Unless of course we consider a therapy light source attached Hyperlink: http: / /www .imag ingec onomi cs .co m/ /is sues/ /arti to endoscopes which are designed to be guided within the body). cles/ /MI _2 001 -0 5 _10. asp The hazards in optical radiation can be caused by absorption of both the direct beam and scattered light, and depend on the Optically stimulated luminescence wavelength of the irradiation which is linked to both the energy (Radiation Protection) Optically stimulated luminescence (OSL) and the depth of propagation. That is to say, the longer the wave- is a physical phenomenon of the emission of optical radiation in length, the less the energy, and the deeper the penetration. the form of prompt fluorescence, stimulated by illuminating a pre- Some effects can arise after even brief exposure and even small viously irradiated sample. Many crystals, e.g. quartz, aluminium harm can lead to long-term effects caused by repetitive exposure. oxide and feldspars, can absorb energy of the ionising radiation The International Commission on Non-Ionising Radiation and store it by elevating electrons from the valence to the conduc- (ICNIRP) has been studying and reviewing studies on exposure to tion band, then capturing it at trapping centres (metastable states) NIR and defining safe limits of exposure to artificial optical radi- within the bandgap, below the bottom of the conduction band. ation. These limits are also reported in the European Directive on Holes, created in the valence band, can move through the crys- Artificial Optical Radiation (AORD, 2006). tal and be trapped by trapping centres situated above the top of Related Articles: AORD, Blue Light Hazard, ICNIRP, Eye, the valence band. The lifetime of the electron or hole traps can Skin, Thermal Light Hazard, UV Light Hazard be large (up to hundreds of years). If the trap is a luminescence Further Readings: Coleman, A., F. Fedele, M. Khazova, centre, then the energy of the optical radiation from a laser or P. Freeman and R. Sarkany. 2010. A survey of the optical haz- an LED (light-emitting diode) will cause de-excitation (recombi- ards associated with hospital light sources with reference to the nation) and, consequently, emission of light (luminescence) will Control of Artificial Optical Radiation at Work Regulations 2010. occur. The emitted light is detected using a photomultiplier tube. J. Radiol. Prot. 30(3):469; Icnirp. A closer look at the thresholds The equipment needed in OSL dosimetry consists of a readout of thermal damage: Workshop Report by an Icnirp Task Group. light source (laser or LED) for exciting charges from traps and Health Phys. 111(3):300–306; 2013. ICNIRP Guidelines on limits initiating recombination, a set of optical filters to shield the pho- of exposure to incoherent visible and infrared radiation. Health tomultiplier tube from the direct or reflected light from the read- Phys. 105(1):74–91; 2013. ICNIRP Guidelines on limits of expo- out source, a photomultiplier tube measuring the light emitted by sure to laser radiation of wavelengths between 180 nm and 1,000 the previously irradiated sample, associated

electronics and a PC µm. Health Phys. 105(3):271–295; 2004. ICNIRP Guidelines on (Figure O.9). limits of exposure to ultraviolet radiation of wavelengths between The readout light source can be used as a continuous excitation 180 nm and 400 nm (Incoherent Optical Radiation). Health Phys. source or as multiple pulses of light. The readout time is of the 87(2):171–186; 2000. ICNIRP Revision of the guidelines on limits order of several seconds. of exposure to laser radiation of wavelengths between 400 nm The sensitivity level is down to a few μGy for OSL radiation and 1.4 µm. Health Phys. 79(4):431–440; Sihota, Ramanjit and dosimetry using aluminium oxide. Radhika Tandon. 2011. Parsons’ Diseases of the Eye. Elsevier, The OSL technique can also be applied to optical dating of India. ancient materials. Optically stimulated luminescence (OSL) dosimeter 667 Optimisation Light source, Optical e.g. laser or LED Previously filter Electronic PC irradiated processing (output by ionizing of data presentation) radiation crystal, O e.g. quartz Photomultiplier FIGURE O.9 A scheme of a set-up for OSL radiation dosimetry. Related Articles: Dosemeter, Dosimeter, Thermoluminescent +Y dosimeter Further Reading: Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, p. 735. Optically stimulated luminescence (OSL) dosimeter (Radiation Protection) Certain crystalline structures such as OSL materials (e.g. beryllium oxide ceramic, aluminium oxide doped with C, calcium fluoride phosphor doped with Tm, quartz, feld- –X +X spars) contain defects in their crystal structure that trap electrons released by an exposure to ionising radiation. In OSL, the trapped electrons are subsequently freed by stimulation with light. After stimulation by light, the detector releases the stored energy in the form of light, i.e., the crystal is stimulated to emit light. The light output measured with photomultipliers can be used to determine the crystal absorbed dose. –Y Optimal incident beam profile (Radiotherapy) Conventional linear accelerators are designed to FIGURE O.10 Multisegment IMRT field showing variation in intensity produce a flat cross-sectional beam profile at a depth in a water within the field. Darker grey indicates an increased dose. equivalent phantom. Forward planned, conventional conformal radiotherapy plans use a number of these beams, appropriately shaped with shielding blocks or multileaf collimators. While many of the treatment plans, particularly when there is Further Readings: Bortfeld, T. 1999. Optimised planning a concave target volume, can be improved by modifying the shape using physical objectives and constraints. Semin. Radiat. Oncol. of the beam profile, the ability to do this is limited to using fixed 9:20–34; Webb, S. 2001. Intensity-Modulated Radiation Therapy, wedge angles or designing missing tissue compensators. With the Institute of Physics Series in Medical Physics, Institute of Physics introduction of intensity-modulated radiotherapy (IMRT) tech- Publishing, Bristol, UK. niques, not only the geometric field shape can be engineered but the intensity at any point in the beam profile can be varied within Optimisation the shaped field, as shown in Figure O.10. One of many reviews of (Radiation Protection) The second principle of protection against IMRT techniques is given by Webb (2001). ionising radiation for workers, patients and members of the pub- The determination of the optimal beam profile for any given lic, specified by the International Commission on Radiological beam entry angle is non-trivial and requires the treatment plan- Protection, is optimisation. ning system to be capable of inverse planning techniques using Once a practice, carried out by an employer involving the sophisticated optimisation algorithms. Common examples of exposure of staff, patients or the public, has been justified in these include the gradient descent method and the simulated terms of the net benefit outweighing the risks associated with the annealing method. Various optimisation techniques for radiother- exposure (Justification), then it must be optimised – i.e. the expo- apy planning have been reviewed by Bortfeld (1999). sure must be minimised whilst the benefit of the practice is still Related Articles: Multileaf collimator, Intensity-modulated achieved. This is often called the ALARP Principle (that doses radiotherapy, Treatment planning system should be as low as reasonably practicable). It used to be called Optimisation of, detector 668 Ordered subset expectation maximum method the ALARA Principle (as low as reasonably achievable). The radi- effect is more pronounced in indirect detectors, since the voltage ation employer is entitled to consider economic and social fac- applied in direct detectors enforces a narrower propagation path tors when deciding what measures to implement to try to reduce to the electron-holes. doses to staff, patients or the public – hence the use of the word Noise: The statistical fluctuations producing image noise ‘practicable’. are affected by the number of radiation photons absorbed in the Finally, once a practice has been justified and optimised, the detector. A high-Z and high-thickness detector exhibits less noise employer must also ensure that any radiation doses received by for a specific x-ray exposure to the receptor. staff or the public are under statutory dose limits – i.e. limitation. Pixel Size: The pixel size is a factor determining the smallest Related Articles: ALARP principle, Justification, Limitation objects that can be visualised. A small pixel enhances visibility of Further Reading: ICRP. 2008. Recommendations of the detail; however, because the area needed for the electronic circuit International Commission on Radiological Protection, ICRP components has a smaller pixel has a smaller effective area, thus Annual Report, ICRP Publication 103, ICRP 37, pp. 2–4. the system has lower sensitivity. Conclusion: Sensitivity, unsharpness and noise are affected O Optimisation of, detector by the detector material, its structure, the thickness and the (Diagnostic Radiology) There are three basic technical designs applied voltage (for direct detectors). These parameters should for receptors in digital radiography. One uses a photostimulable be chosen for optimal detector performance (i.e. high DQE) per phosphor material that is first exposed by x-radiation to form a X-ray modality. latent electrostatic image and then later processed to convert the Related Articles: Image quality metrics, Detective quantum electrostatic image into a digital image. This method has often efficiency, Direct digital radiography been designated as ‘computed radiography’. The other method Further Readings: International Atomic Energy Agency. 2014. is generally designated as ‘digital radiography’ and captures Diagnostic Radiology Physics: A Handbook for Teachers and the x-ray image in a digital format. There are two methods of Students, Vienna, Austria; Kandarakis, I. S. 2016. Luminescence digital radiography, designated as either ‘direct’ or ‘indirect’. in Medical Image Science. J. Lumin. 169b:553–558; Zahangir The difference is in the type of material used to capture and Kabir, M. and Safa Kasap. 2017. Chapter 45 Photoconductors for absorb the x-radiation. The direct method uses a semiconduc- X-ray imaging detectors. In: Springer Handbook of Electronic tor, typically selenium (a-Se), in which the x-radiation produces and Materials, 2nd ed. Kasap Capper Editors, ISBN: 978-3-319- an electric charge that is picked up by the CCD array. The indi- 48931-5, e-ISBN: 978-3-319-48933-9, doi:10.1007/978-3-319- rect method uses scintillators like CsI:Tl or Gd2O2S:Tb as the 48933-9, Library of Congress Control Number: 2017944229 © x-ray absorber. The x-radiation is first converted into light. This Springer International Publishing AG. is passed on to a photosensitive device (e.g. photo diode), and converted into an electrical charge picked up by the CCD array. Optocoupler These detectors are configured into a matrix of individual pixel (General) An optocoupler is an electronic device used to elec- elements. trically isolate two parts of a system, while allowing data trans- The design of the x-ray detector elements, including selection mission from one side to the other. Optocouplers are available of material and size, are major factors effecting image quality and in integrated circuit packages as shown below. An LED trans- patient exposure and should be taken into consideration in pro- mitter on the input side converts input signals to light and trans- ducing optimised detectors. mits this light across a transparent barrier to a phototransistor Sensitivity: The scintillator or the semiconductor should be receiver. The phototransistor senses this light and an output sig- made of a high-Z material like CsI:Tl or Gd2O2S:Tb for indi- nal can be generated in external circuitry. Optocouplers can be rect detectors and a-Se for direct detectors. The high-Z material configured to transmit either analog or digital data from input enhances photoelectric absorption and reduces the possibility of to output. scatter in the detector material. The latter may be a significant The barrier between the transmitter and receiver in an opto- source of unsharpness. Another way for enhancing absorption is coupler is non-conducting and electrically separates the two sides. creating thick detectors. The detector thickness should be related Optocouplers find application in ECG monitoring to electrically to the related energy spectra of choice. For instance, mammo- isolate the ECG amplifier connected to the patient from the rest of graphic detectors are thinner due to the lower energy of the the device. This safety measure eliminates any potential electrical mammographic X-ray spectra, compared to radiography and CT paths from the machine to the patient leads (Figure O.11). detectors. The material of the detectors should have a high efficiency Ordered subset expectation maximum method in transforming the absorbed X-ray energy to optical photons (Nuclear Medicine) The main problem with the maximum like- (indirect detectors) or electron-hole pairs (direct detectors). The lihood expectation maximisation (MLEM) method is the long columnar structure of CsI:Tl and the grain size in Gd2O2S:Tb time required to obtain acceptable accuracy in the reconstructed should be such that the optical photons suffer little losses. In addi- image. A successful method to accelerate the convergence rate tion, the voltage applied between the output and the input of the is the ordered subset expectation maximum (OSEM) method. semiconductor in direct detectors should be optimised so that the This method is identical to the MLEM algorithm in its prin- recombination effects resulting in signal loss are minimal. The ciples but differs at the stage where the image is updated. In recombination is also affected by the amount of the electron-hole the MLEM algorithm, the image is updated (i.e. multiplied by created, since a larger amount of signal carriers increases the the correction matrix) only after all projection angles have been recombination rate. processed. Unsharpness: Detector based unsharpness or blurring is In the OSEM algorithm, the image is instead updated after a mainly due to the spread of the secondary carriers (optical pho- subset of projections has been processed. For example, a com- tons or electron-hole pairs) while propagating to the output. A mon number of subsets are 16 for a 64 projection angle acqui- high thickness detector usually exhibits higher unsharpness. The sition. Hence, the image is updated after four angles have been Organ at risk (OAR) 669 O rganic liquid scintillators LED 1. Parallel organs can be treated using a ‘critical volume’ model; the organ architecture is such that each individ- IC pins ual element is not crucial to the overall function of the organ. It is important to ensure that a significant volume of the organ is not damaged by the radiation. Tolerances may vary with percentage volume irradiated. Examples include the liver, kidney and lung. 2. In contrast, the spinal cord and rectum are serial organs, following a ‘critical element’ model, each element is crucial to the function. If one element is damaged then the organ is damaged irreparably. Abbreviation: OAR = Organ at risk. Related Articles: Adverse effects (Radiotherapy), Tolerance Further Readings: ICRU Report 50. 1993. Prescribing, O Recording and Reporting Photon Beam Therapy, International Commission on Radiation Units and Measurements, Washington, DC; ICRU Report 62. 1999. Prescribing, Recording and Reporting Photon Beam Therapy (Supplement to ICRU Report 50), Phototransistor International Commission on Radiation Units and Measurements, Washington, DC. FIGURE O.11 Integrated circuit (IC) package for a basic optocoupler. The LED on the left is electrically isolated from the phototransistor on Organ dose the right. (Radiation Protection) Organ dose is otherwise known as the equivalent dose, H, and is a product of the absorbed dose D to an organ and the relative biological effectiveness (RBE) for the Iterations irradiation: 1 2 3 4 5 10 20 100 H = D ´ RBE MLEM For radiation protection purposes, the RBE is taken into account OSEM by the use of the radiation weighting factor. Since RBE = 1 for the vast majority of diagnostic and therapeutic radiations, organ dose is usually equivalent to the absorbed dose to the organ. FIGURE O.12 Example using images with different iteration numbers

Risks from ionising radiation are based on calculations of for MLEM and OSEM. organ doses. In therapeutic applications, the organ dose to critical tissues (peripheral tissue of high radiosensitivity, e.g. bone mar- row) is the limiting factor, and the efficacy of the treatment is processed. The acceleration in this method is roughly propor- usually determined by this. tional to the number of subsets. The order of the projection angle Methods of estimating organ doses vary depending on the is not linear but follows a particular pattern for an optimal recon- purpose of irradiation. Very accurate dose estimates are made struction. An iteration is defined when all subsets are processed. in radiotherapy where large doses prescribed to tumours must be The process is then repeated for the desired numbers of iterations. carefully optimised to minimise surrounding critical-organ doses. Figure O.12 shows an example using images obtained with differ- In diagnostic procedures, where the accuracy is less critical, ent iteration numbers for MLEM and OSEM. organ doses are usually estimated by using the scan parameters Related Article: Maximum likelihood expectation maximum in a calculation that assumes a standard-sized patient phantom. (MLEM) Related Articles: Radiosensitivity, Relative biological effec- tiveness, Radiation weighting factors Organ at risk (OAR) (Radiotherapy) Any normal tissue whose radiosensitivity sig- Organic liquid scintillators nificantly influences the treatment planning or prescribed dose is (Radiation Protection) Liquid scintillation counting is used known as an organ at risk (OAR). As for the target volume, it for detection of a weak penetrating radiation like beta or alpha is important to consider the physiological movement and set up particles. errors and incorporate this variability within a margin. There are several types of organic scintillators, e.g. crystals, The use of OARs as planning volumes was proposed by the liquid organic solution, plastic, thin plastic films and ones loaded ICRU in Report 50 (with addendum 62). This report provided a with high atomic number elements, such as Pb or Sn. Organic common framework on prescribing, recording and reporting ther- scintillators are transparent to their own fluorescence emitted apies, with the aim to improve the consistency and inter-site com- radiation. parability. This report details the minimum set of data required to Liquid organic scintillators consist of an organic scintillator, be able to adequately assess treatments without having to return e.g. anthracene, solvent and sometimes a wavelength shifter if it is to the original centre for extra information. necessary to adjust the light wavelength to the spectral sensitivity OARs can be classified into two categories – parallel or serial of the photomultiplier. They can be used up to very high expo- organs: sures, e.g. 105 Gy. Organisational structure 670 Orthogonal films The counting efficiency, i.e. the scintillation efficiency, is • 180 degrees (beam pointing at the ceiling) – 0.25 defined as a ratio of the number of particles whose energy is con- • 90, 270 degrees (beam pointing horizontally at the verted in the light to the number of all incident particles. If the walls) – 0.25 radioactive material, e.g. C-14 or H-3, is dissolved in a liquid scin- tillator, the counting efficiency attains 100%. For other treatment modalities such as tomotherapy or cyberknife, Liquid organic scintillators with a hydrogen-containing other (lower) orientation factors may be used. organic solvent are applied in the detection of fast neutrons using Related Articles: Instantaneous dose rate (IDR), Workload proton recoil. factor (W), Occupancy factor (T) Related Article: Liquid scintillation (LS) counting Further Readings: 2017. Design and Shielding of Further Reading: Knoll, G. F. 2000. Radiation Detection and Radiotherapy Facilities. IPEM Report 75, 2nd edn., IPEM; 2007. Measurement, 3rd edn., John Wiley & Sons, Inc., New York, pp. Structural Shielding Design and Evaluation for Megavoltage X- 223–226, 558. and Gamma-Ray Radiotherapy Facilities. NCRP Report 151, NCRP. O Organisational structure (General) The organisational structure in healthcare facilities is Ortho pan tomography (OPG) the backbone of a company and is the result of choices and divi- (Diagnostic Radiology) Ortho pan tomography, or orthopan- sion, management and coordination of work to pursue and achieve tomography (OPG), is a dental x-ray procedure that scans the a target. curved jaws and teeth and produces a panorama-type image on a Building an organisational structure involves: the choice of the flat film or digital receptor. most appropriate formal structure with respect to the corporate An OPG system is shown in Figure O.13. OPG equipment purpose; defining the levels of responsibility among the various includes x-ray tube with slit beam restrictor, which produces thin figures present in the company; the creation of a stable organisa- vertical beam with width of the order of 4–7 mm (at the film/detec- tional structure for the assignment of specific functions and roles; tor surface) and height sufficient to cover the mandibles and max- the realisation of the organisational asset configuration (organisa- illa regions. The x-ray tube rotates around the patient head and the tion chart); and the definition of rules and standard procedures to x-ray beam exposes the x-ray film/detector positioned opposite to be adopted to carry out specific operations. the x-ray tube. In front of the x-ray film/detector is another slit diaphragm, which allows only a thin x-ray beam to expose part of the film/detector. During the rotation of the x-ray tube around the Orientation Factor patient head, the film/detector moves behind this slit diaphragm; (Radiation Protection) In designing the shielding for a room this way, different parts of the film/detector record the x-ray beam (bunker) housing an external beam radiotherapy LINAC, it is passing through different areas of the patient head. The produced necessary to consider the primary beam and secondary radiation panoramic radiograph of the jaws and teeth (Figure O.14) usually (from scatter and from treatment head leakage). delivers high dose over the scanned area, as the rotation around The secondary radiation varies in different directions relative the head (the x-ray exposure) usually takes several seconds. to the primary beam, but it is incidental to all the internal surfaces of the walls, ceiling and floor all the time whilst the radiation beam is ‘on’. Orthogonal films However, the primary beam can only point in one plane (Radiotherapy, Brachytherapy) around the isocentre (centre of rotation of the treatment head). Definition of Applicator and Source Positions: Radiographs Thus only some areas of the walls, ceiling and floor will have to are commonly used for reconstruction of applicator/source posi- be considered as primary barriers – i.e. that the primary beam is tions when image-guided brachytherapy techniques are not avail- incidental to them, and the barrier needs to be thick enough that able. At least two radiographs are needed for the 3D reconstruction. the beam is attenuated sufficiently to protect anyone on the other Orthogonal radiographs, frontal (anterior to posterior, AP) and side of that barrier. -lateral, are the most commonly used imaging arrangement. In considering the thickness of primary barrier required, a number of factors will need to be considered: • Firstly, the maximum incident instantaneous dose rate (IDR) that the beam could deliver at the position where someone could be standing outside the barrier. • The workload factor – i.e. the proportion of a working day that the beam is ‘on’. • The occupancy factor – i.e. the proportion of a working day that a person may be at a position just beyond the barrier. However, one other factor is also of importance – the orienta- tion factor. It is the proportion of the total time that the radiation beam is ‘on’ and is pointed towards that particular barrier. The IPEM and NCRP reports on the design of radiotherapy facilities (see Further Readings section) broadly concur that the following orientation factors may be used: • 0 degrees (beam pointing at the floor) – 0.3 (i.e. 30% of the total beam on time) FIGURE O.13 A typical OPG x-ray system. Orthogonal films 671 O rthogonal films Two orthogonal radiographs are shown in the article entitled Internal reference point. These are used to determine rectum and bladder reference points for a cervix cancer application. Reconstruction devices, such as the jig shown in Figure O.15a, have also been implemented in some brachytherapy treatment planning systems. A reconstruction jig, with its markers, has a stable predefined geometry. The geometry (e.g. magnification) of the radiographs taken is thus determined from the projection of the jig, and in this case there is no requirement on exact film posi- tioning for a good reconstruction. The applicator/source positions are in turn reconstructed in relation to the jig. In the example shown in Figure O.15b, a treatment of an epi- pharynx cancer with a two-channel intracavitary applicator, two ‘almost’ orthogonal images using a jig are presented. FIGURE O.14 Panoramic radiograph of the jaws and teeth (OPG Related Articles: Implant dose distribution, Internal reference O radiograph). point (a) (b) FIGURE O.15 (a) Lateral radiograph (one of two orthogonal films) with marker wires inserted in the two treatment channels. The spinal cord is indi- cated by a thin pencil line by the oncologist. The reconstruction jig to be used is shown to the right, a stable 90° jig arm with two boxes, each with two crosses. (b) Lateral and frontal radiographs with the jig in place showing the two crosses on each box. Marker wires are inserted in the two treatment channels. Lead shot marks the medial and lateral corners of the right eye. Also shown are the isocentre cross and square belonging to the x-ray unit. Orthogonal pair of x-rays 672 Oscilloscope Orthogonal pair of x-rays t (Radiotherapy) An orthogonal pair of x-rays, either megavoltage RF 0 (MV) or kilovoltage (kV) planar images, can be used to check anatomy in three dimensions as part of patient setup before treat- ment. By imaging at two angles, at 90 degrees from each other, Gslice it is possible to calculate the necessary geometrical shifts, so that a patient can be re-positioned to match the digitally recon- structed radiographs (DRRs) associated with their treatment plan. Gx Orthogonal pairs of x-rays can also be used as part of image- guided radiotherapy (IGRT), instead of Cone Beam CT (CBCT). Gy O Echo TD FIGURE O.16 Sequence diagram for a spiral EPI sequence. A Comparing orthogonal x-rays to digitally reconstructed radiographs in order to verify a patient’s set-up, immediately prior to treatment. β Further Reading: Ali, I. et al. 2010. Setup accuracy of a stereo- tactic immobilization mask. J. Appl. Clin. Med. Phys. 11(3). FIGURE O.17 Block diagram of an oscillator. Orthovoltage radiation (Radiotherapy) Orthovoltage therapy describes treatment with medium energy kilovoltage x-ray beams (accelerating potential C2 R4 160–300 kV; 0.5–0.4 mm Cu half-value layer [HVL]). There is no sharp division between the various voltage ranges and they can vary slightly in different documents. + C1 V Abbreviation: HVL = Half-value layer. out R Further Reading: Bomford, C. K. and I. H. Kumkler, eds. 3 – 2003. Walter and Miller Textbook of Radiotherapy, 6th edn., R2 Churchill Livingstone, Oxford, UK, p. 143. R1 Oscillating gradient (Magnetic Resonance) An oscillating gradient is used to fill k-space using a spiral trajectory. The spiral EPI sequence uses two simultaneously applied magnetic field gradients, and the data are FIGURE O.18 Electrical schematic of Wien-bridge oscillator built with collected continuously during an acquisition time T an operational amplifier. D. The method offers reduced sensitivity to motion and some reduction in imag- ing time, and does not require such strong gradients (Figure O.16). β. In order to maintain steady oscillations, the Barkhausen equa- Related Articles: k-space, Spiral scanning tion must be satisfied: Further Reading: Haacke, E. M., R. W. Brown, M. R. Thomson and R. Venkatesan. 1999. Magnetic Resonance A ×b = 1Ð0° Imaging. Physical Principles and Sequence Design, Wiley-Liss (John Wiley & Sons), New York. Feedback oscillators are often realised with operational ampli- fiers. RC circuits (C2R4 and C1R3) form the positive feedback net- Oscillator work and the gain is determined by the resistor network R1 and R2 (General) An electronic circuit that converts DC power into AC (Figure O.18). power at a frequency defined by values of electronic components Related Article: Operational amplifier used in the circuit. Basic types of oscillators are harmonic and relaxation oscillators. Harmonic oscillators generate sinusoidal Oscilloscope waveforms, while relaxation oscillators generate non-sinusoidal (General) Oscilloscope is an instrument that enables the dis- waveforms such as square wave, sawtooth, etc. play and measurement of signal parameters from a screen. Most

The block diagram of an oscillator in Figure O.17 comprises of oscilloscopes have linear time bases that enable measurement an amplifier having gain A with external positive feedback circuit of the parameters against time, but many enable measurements OSLD (Optically stimulated luminescence dosimeter) 673 Over voltage with respect to other electrical quantities as well. Analogue the net flow of water molecules from a compartment with high oscilloscopes display the waveforms on the screen of a cathode water potential (i.e. a measurement of the tendency of water to ray tube. Digital oscilloscopes retain the basic idea of viewing move between compartments) through a semipermeable barrier the signal waveform on the screen, but use a microprocessor or to a compartment with low water potential. DSP computational power to calculate a vast number of signal parameters. Digital oscilloscopes usually can store the wave- Output factor form and the measurement results in a memory. Their input sig- (Radiotherapy) Linear accelerators are calibrated to deliver a nal range is regularly from a few millivolts to a few tens of volts, known dose at a specific point in a given field (under reference and the frequency range from DC to gigahertz (Figures O.19 conditions). For example, it is common to calibrate electron beams and O.20). at the reference depth (dref) for a 10 cm × 10 cm field at SSD = 100 cm. All other dose data, as used in dose calculations, can be then OSLD (Optically stimulated luminescence dosimeter) calculated from a ratio of the current output to the output at the (Non-Ionising Radiation) See Optically stimulated luminescence reference depth under reference conditions. This ratio is known as (OSL) dosimeter the output factor, and it is used to calculate the length of time of beam-on, or monitor units (MUs), that must be used to deliver the O Osmosis prescription dose. (Nuclear Medicine) The diffusion of water molecules through a For example, if the output factor at the prescription point is semipermeable barrier, e.g. a cell membrane, is called osmosis. 0.5, then the number of MUs needs to be doubled in order to One prerequisite for osmosis is two compartments with different deliver equal dose to the PTV, as it would have done at the refer- solute concentrations leading to a flow of water molecules from ence depth under reference conditions. one compartment to the other. Osmosis can also be defined as The output factors are usually defined as the ratio of the dose for non-reference conditions to the dose under reference condi- tions. They may be determined as the ratio of corrected dosimeter readings measured under a given set of non-reference conditions to that measured under reference conditions. The measurements are usually done at the depth of maximum dose or at the refer- ence depth and corrected to the depth of maximum dose using PDD (or TMR). This method minimises the dosimetric error due to positioning and reduces any electron contamination effect. Reference conditions depend on the type of ionising radiation (photon beams, electron beams, kilovoltage x-ray beams, etc.) and on the implemented concept of the absorbed dose determination. The output factors are measured and tabulated at commissioning. They take into account the field size and the field shape, wedges, off-axis positioning, various SSD, etc. Abbreviations: MU = Monitor unit, PDD = Percentage depth dose, PTV = Planning target volume, SSD = Source surface dis- tance and TMR = Tissue maximum ratio. Related Articles: Percentage depth dose, Electron contamination Output screen (Diagnostic Radiology) See Image intensifier FIGURE O.19 Analog oscilloscope. Overall survival (Radiotherapy) See Clinical trial endpoints Over-table radiography (Diagnostic Radiology) Over-table (over-couch or overtable) radi- ography uses x-ray tube mounted above the patient. This is the most typical setup of x-ray radiographic equipment. However, for fluoroscopic equipment, this set-up produces considerably more scattered radiation (from the patient) to the staff, and in this case, under-couch x-ray tube positioning is the preferred option. Some sources also name these mounts according to the detector (e.g. under-couch film holder, or over-couch image intensifier). Related Articles: Radiography, Fluoroscopy Over-table tube (Diagnostic Radiology) See Over-table radiography Over voltage FIGURE O.20 Digital oscilloscope. (Diagnostic Radiology) See Overvoltage protection Overbeaming 674 O verload protection Overbeaming exposure. For most medical procedures using radiation, there is (Diagnostic Radiology) In CT scanning, the term ‘overbeam- usually an optimum amount of exposure that produces the neces- ing’ is used to refer to the additional extent of the z-axis x-ray sary image quality. In film radiography, overexposure results in beam profile over the nominal z-axis beam profile, as shown in excessive film density or darkness and a possible reduction in con- Figure O.21. trast. In digital radiography overexposure of the receptor (above Overbeaming usually occurs on multislice CT scanners and the optimum value) results in unnecessary exposure to the patient. is due to the x-ray beam penumbra. In multislice CT, all detector Radiographs produced with three different exposures are com- banks used for image acquisition should receive a uniform photon pared in Figure O.22. flux if a uniform image quality is to be achieved. The x-ray beam extent is therefore increased so that the penumbra lies outside the Overload active detectors, shown in dark grey in Figure O.21. The extent (Diagnostic Radiology) See Overload protection of penumbra on CT scanners is typically in the order of 2–3 mm. Overbeaming results in a reduction in dose efficiency as not all Overload protection the photons that the patient is exposed to contribute to the image. O (Diagnostic Radiology) Overload is the excess of energy imparted Related Articles: Geometric efficiency, Multislice CT scanner to the anode of an x-ray tube. Each tube has its maximal tube load parameters (kV, mA and overall watts or heat units). These are Overexposed specific for each focal spot. Loading over of the thermal capac- (Diagnostic Radiology) See Overexposure ity of the specific actual focal spot can lead to the local melting of anode. This can further lead to destruction of the whole x-ray Overexposure tube. Each x-ray tube has a special circuit for overload protection. (Diagnostic Radiology) Overexposure is a condition in which There are different ways to assure the protection of the x-ray something has received more than the optimum radiation tube anode against overloading. One of the simplest systems is a rubber membrane in the x-ray housing allowing expansion of the insulating/cooling oil (due to excessive heat) to activate a cut-off switch. However, the protection has to be related to the specific actual Nominal focal spot. In this case, a calculator estimates the maximal kV beam and mA s, compares these with the x-ray tube rating chart and width in case of potential overloading does not allow for the exposure. Similarly, different charts will be used for overload protection in multiple exposures (e.g. in angiographic series). Older x-ray equipment use special capacitor whose charge rep- Overbeaming from resents the heat units of the anode. The capacitor charges during penumbra the exposure and after it discharges through a system with time constant similar to the cooling curve of the x-ray tube. In this way, a system measuring the charge of the capacitor will effec- tively show the accumulated heat in the anode. This system can be used to disallow (or interrupt) exposures which could overload the z-axis x-ray tube. In a similar way, contemporary x-ray equipment use microprocessor simulator, which works in parallel with the x-ray FIGURE O.21 ‘Overbeaming’ on a multislice CT scanner. (Courtesy of tube and shows its current thermal load. The simulator provides ImPACT, UK, www .impactscan .org) signal for the overload protection system. Normally, replacing of 70 kVp, 25 mA s 70 kVp, 50 mA s 70 kVp, 80 mA s FIGURE O.22 Radiographs produced with different exposures. The one on the right appears to be overexposed. Oversampling 675 Oxygen the x-ray tube requires replacement of memory (or firmware) stor- Overshoot ing information about the thermal capacity of the tube. Related Articles: Heat units, Heat capacity, Maximum load, Cooling curve, Tube rating charts, Actual focal spot, Overload Rectangular pulse Oversampling (Magnetic Resonance) Oversampling denotes sampling of a band-limited signal at a rate faster than required by the Nyquist theorem. Since aliasing of unwanted components exceeding the FIGURE O.24 Overshoot of a rectangular pulse. Nyquist frequency is avoided, these signals can be removed after Fourier transform. In MRI, oversampling (usually by a factor of 2) is routinely performed during readout. In this direction, the field Overvoltage protection of view may be chosen to be smaller than the object. In 3D MRI, (Diagnostic Radiology) Overvoltage (over voltage) is the excess of oversampling may be performed in the slice-direction to avoid high voltage to the x-ray tube. This can lead either to overload or aliasing of signals from the slope of the selected slab. Such phase- to insulation problem. O oversampling requires time through additional repetitions and is Usually the overvoltage protection circuit of the high-voltage usually restricted to 10%–20% of the partitions. generator measures the kV and provides corrective feedback. In Related Articles: Nyquist frequency, Fourier transform high-frequency generators this reduces the frequency of the cur- rent supplying the high-voltage transformer. Additionally high- Overscanning voltage semiconductor devices ensure avoiding the voltage from (Diagnostic Radiology) In CT, ‘overscanning’ refers to the addi- exceeding a predetermined limit. tional irradiation in helical scans from x-ray tube rotations beyond Related Articles: High voltage generator, High voltage control the extent of the nominal selected scan length (Figure O.23). Some overscanning is required by necessity in a helical scan, Oxine so as to obtain sufficient data for interpolation of axial images (Nuclear Medicine) Oxine is a non-specific blood cell label- from the helical data set (see Spiral interpolation). ling agent used for infection/inflammation scintigraphy and is The increase in dose due to overscanning is more significant attached to 111In. To detect sites of infection, white blood cells on multislice scanners due to the greater extent of the z-axis x-ray (WBCs) are labelled with 111In-oxine. The WBCs are extracted beam profile (Figure O.23b). It is also more significant for short from the patient and labelled. Labelled WBCs are then re-injected scan lengths where it may be more dose efficient to employ a nar- into the patient. Delayed imaging (at approximately 24 hours) is rower z-axis x-ray beam despite the reduced dose efficiency of performed to assess if any infection is present. narrow collimations (see Geometric efficiency). Alternatively, 111In-oxine has been used since the mid-1970s and is supplied where appropriate, sequential (axial) scanning may be employed as a solution in a vial (i.e. ready-to-use radiopharmaceutical). which does not require the additional irradiation. Nowadays, 111In-oxine has in many departments been replaced Recently, some manufacturers have adopted a method of by WBC labelled with 99mTc-HMPAO. Nevertheless,111In-oxine is ‘adaptive collimation’ to minimise the extent of the additional known to have a higher labelling efficiency. irradiation from overscanning. 111In-oxine penetrates through the cell membrane and the 111In Related Articles: Spiral interpolation, Geometric efficiency attaches to proteins in the cell. In a healthy human the 111In-oxine is mainly observed in spleen, liver and bone. In the initial four Overshoot hours after injection, pulmonary activity can be observed. (General) Overshoot is a characteristic of a realistic rectangular Related Articles: Tc-99m-Labelled Leukocytes, electrical pulse, when the instantaneous value of the pulse rises Tc-99m-HMPA to a value larger than amplitude (height) of the pulse and then Further Readings: Roca, de Vries, Jamar, Israel and Signore. decays to the amplitude value usually in a damped oscillatory way 2010. Guidelines for the labelling of leucocytes with 111In-oxine (Figure O.24). - Inflammation/Infection Taskgroup of the European Association Related Article: Pulse of Nuclear Medicine. EJNMMI 37(4):835–841; Sharp, P. F., H. G. Gemmell and A. D. Murray. 2005. Practical Nuclear Medicine, 3rd edn., Springer. Nominal Nominal Oxygen scan scan (General) length length Symbol O Element category Non-metals Mass number A 16 Atomic number Z 8 Atomic weight 15.994 g/mol Electronic configuration 1s2 2s2 2p4 (a) (b) Melting point 54.8 K Boiling point 90.2 K FIGURE O.23 Overscanning due to additional irradiation beyond the Density near room temperature 1.43 kg/m3 nominal scan length (a) on a single slice scanner and (b) on a multislice scanner. (Courtesy of ImPACT, UK, www .impactscan .org) Oxygen-15 676 Oxyhaemoglobin At standard temperature and pressure, oxygen is a diatomic gas, such as α-particles. With low-LET radiation, the addition of oxy- with molecular formula O2, which

is both colourless and odour- gen increases radiosensitivity, and the precise value of this factor less. Oxygen is used by all living cells in the process of respiration, is the OER. For low-LET radiation the OER is 2.5–3.5 and for combining glucose and oxygen to produce carbon dioxide, water high-LET radiation it is unity. For intermediate LET radiation, and stored energy in the form of adenosine triphosphate (ATP). such as neutrons, it is about 1.6. Isotopes of Oxygen and Their Medical Applications: Oxygen Studies of the effects of oxygen on the response of rapidly has three stable isotopes all of which have medical applications. growing cells cultured in vitro to x-rays suggest that the OER has Radioactive 13N, used in positron emission tomography (PET), is a value of 2.5 at lower doses such as those typical of the daily produced using 16O. 17O can be used as a tracer to study cerebral dose per fraction delivered in fractionated radiotherapy regimes oxygen utilisation. 18O is used in cyclotrons to produce 18F which (around 2 Gy). This is believed to be due to a variation of OER can be incorporated into fluorodeoxyglucose (FDG), a com- with the phase of the cell cycle. The OER is lower for cells in the mon PET radiopharmaceutical. In addition, cyclotron produced G2 phase (value about 2.3) than in those in the S phase (value about 15O can be used in the form of 15O-Water to assess myocardial 2.8). The G2 cells are more radiosensitive and therefore dominate perfusion. the low-dose region of the survival curve. Therefore, the OER of O In functional MRI, the relative change in blood oxygenation an asynchronous cell population is slightly smaller at low doses level in different areas of the brain during mental tasks is utilised than high doses. This has been demonstrated in vitro where pre- as a form of contrast sensitive to brain activation. cise survival measurements are possible but would be difficult Related Articles: Blood oxygenation level–dependent con- to show in tissue. Although this is an interesting radiobiological trast, Cyclotron, Functional magnetic resonance imaging, Isotope, observation, it is of little or no clinical significance in radiotherapy. Myocardial perfusion imaging, Oxygen enhancement ratio, Abbreviations: LET = Linear energy transfer and OER = Oxyhaemoglobin, Positron emission tomography, Reoxygenation, Oxygen enhancement ratio. Radioactivity Related Articles: Cell cycle, Linear energy transfer (LET), Linear quadratic (LQ) model, Radiosensitivity, Redistribution, Oxygen-15 Reoxygenation (Nuclear Medicine) A radionuclide used for in vivo PET imaging. Further Reading: Hall, E. J. and A. J. Giaccia. 2006. Radiobiology for the Radiologist, 6th edn., Lippincott Williams & Wilkins, Philadelphia, PA. Maximum Positron positron Photon Common Oxyhaemoglobin Half-life fraction energy emission application (Magnetic Resonance) Haemoglobin is an oxygen-transporting protein attached to red blood cells. The essential active compo- 123 seconds 1.00 1.72 MeV 511 keV Cardiac perfusion nent of haemoglobin is iron, a transition element with variable imaging electronic configuration. As such, the iron within haemoglobin allows the transformation of the protein from an oxygen-rich diamagnetic substance (oxyhaemoglobin) to an oxygen-poor 15O-labelled water is a cardiac perfusion agent, used for the paramagnetic substance (deoxyhaemoglobin). In the diamagnetic quantitative assessment of myocardial blood flow. It is widely substance there are no net unpaired electrons (the electrons pair in considered to be the gold standard of non-invasive quantitative spin-up/spin-down partnerships according to the Pauli’s exclusion myocardial blood-flow imaging. A disadvantage with the use principle). The paramagnetic substance has the six outer electrons of 15O water is the need to subtract ‘early’ from the ‘late’ PET of the iron ion distributed across the five 3d orbitals in the form of images. The relatively short half-life of this agent requires an on- one pair and four single electrons of identical spin. It is these four site cyclotron. These difficulties make the use of this agent rela- unpaired electrons of identical spin which cause the deoxyhae- tively rare in routine clinical practice. moglobin to be paramagnetic. When the haemoglobin molecule Related Articles: Positron emission tomography, Radionuclide is oxygenated, there is a slight structural change which alters the imaging lowest energy electronic configuration: the electrons then have no Further Readings: Cherry, Sorenson and Phelps. 2012. net spin-up/spin-down imbalance. Physics in Nuclear Medicine, 4th edn., Elsevier; Mettler and In many areas of the body, a healthy reserve of oxyhaemoglo- Guiberteau 2012. Essentials of Nuclear Medicine Imaging, 6th bin can be stored in muscle, but in the brain, there is insufficient edn., Elsevier; Saraste, Kajander, Han, Nesterov and Knuuti. muscle for this to be the case: storage is poor. Thus, to protect the 2012. PET: Is myocardial flow quantification a clinical reality? brain’s supply of oxygen, oxyhaemoglobin is always oversupplied J. Nucl. Cardiol.; Zeissman, O’Malley, Thrall and Fahey 2014. (even in periods of rest). Nuclear Medicine, 4th edn., Elsevier. During periods of neuronal stimulation, the oversupply is esca- lated so that even after the brain has taken the oxygen it needs, the Oxygen enhancement ratio venous blood flow is still more oxygenated during activity than (Radiotherapy) The oxygen enhancement ratio (OER) is the ratio during the resting state. of the radiation dose given under hypoxic conditions to the radia- As oxyhaemoglobin has no net electron spin imbalance, it tion dose given under fully oxygenated conditions that achieves dephases more slowly than deoxyhaemoglobin. This means that the same biological effect. increased levels of oxyhaemoglobin cause increased T2 and T2* It is known that tumour cells can be hypoxic and therefore less at sites of neuronal activity. fMRI pulse sequences are usually sensitive to radiation causing DNA damage indirectly through T2*-weighted EPI, such that increased oxyhaemoglobin leads to free radicals produced by the ionisation of oxygen. Therefore, the increased MR signal. presence or absence of oxygen dramatically influences the biolog- Related Articles: fMRI (functional magnetic resonance ical effect of sparsely ionising radiations (low-LET) such as x-rays imaging), Blood oxygenation level–dependent contrast, BOLD, but there is no effect for densely ionising radiations (high-LET) Haemodynamic response function P P-32-sodium orthophosphate with the nuclear field. The entire photon energy is absorbed to (Nuclear Medicine) Phosphorus-32 is a reactor produced radionu- produce an electron–positron pair. As the rest mass of an electron clide through the reaction or positron is equivalent to 511 keV, there is an energy threshold for pair production of 1.022 MeV. 31P(n,g)32 P ®b - 32S. The emitted positron quickly annihilates with an electron 14,26d within the medium, producing two photons each of energy It is a β-emitter with maximum beta energy of 1.711 MeV. 511 keV. The half-life of 32P is 14.26 days. It is obtained as a solution of Above 2.044 MeV, triplet production may occur. In this case, 32P-sodium phosphate suitable for oral or intravenous administra- the photon interacts with the field of an electron, giving rise to an tion. The solution is clear and colourless with pH 5–6. It contains electron–positron pair plus the interacting electron. P isotonic saline and a sodium acetate buffer. In nuclear medicine The cross section for pair production varies with Z2. There is 32P is mainly used for radionuclide therapy of polycythemia vera also a weak dependence on E. (a rare blood disease characterised by an elevation of the imma- Related Article: Electron–positron pair ture red blood cells) and other neoplastic haematologic diseases. It is frequently in use in biomedical research as a tracer substitute Palliative treatment for phosphor, e.g. in nucleic acid sequences. (Radiotherapy) Where cure from radical treatment is unlikely, The 32P-sodium phosphate activity recommended for therapy palliative treatment can alleviate painful or distressing symptoms of polycythemia vera is 74–111 MBq/m2 body surface area with a and restore a higher degree of life quality for patients. Palliative maximum activity of 185 MBq. treatments tend to deliver lower doses than radical treatments The ICRP assumes that 30% of administered 32P phosphate but can have a most favourable impact to bring relief in the cases is permanently accumulated in the mineral bone and 30% is dis- where disease has compressed nerves (cord compression) or other tributed in soft tissues, where it is excreted with half-times of 12 structures (oesophagus). h (20%), 2 days (20%) and 19 days (60%). The absorbed dose for Related Articles: Parallel opposed fields, Beam arrangement bone surfaces is 11 mGy MBq−1 and other organs receive 0.74 mGy MBq−1. The effective dose for 32P phosphate is approxi- Paper and thin layer chromatography mately 2.2 mSv MBq−1. (Nuclear Medicine) Chromatographic methods are very suit- Related Article: Phosphorus-32 able for the separation of compounds of similar chemical struc- Further Readings: Saha, G. B. 2004. Fundamentals of Nuclear ture. The appropriate chromatographic method has to be chosen Pharmacy, 5th edn., Springer, New York; Kowalsky, R. J. and S. according to the physicochemical properties of the compounds. W. Falen. 2004. Radiopharmaceuticals in Nuclear Pharmacy and In paper chromatography, the solution of the compounds is put Nuclear Medicine, 2nd edn., American Pharmacists Association, in a small droplet at the end of a specially prepared paper. After Washington, DC; Tennvall, J. and B. Brans. 2007. EANM pro- the spot is dried, the paper (with the end including the spot) is put cedure guideline for 32P phosphate treatment of myeloprolifera- in a suitable solvent. Different compounds can then be separated tive diseases. Eur. J. Nucl. Med. Mol. Imaging 34:1324–1327; due to their different migration velocities. Firestone, R. B. 1999. Table of Isotopes, 8th edn., Update with Thin layer chromatography is similar to paper chromatogra- CD-ROM. http://nucleardata .nuclear .lu .se. (accessed June 2012); phy although a glass plate is used. The plate is prepared with a ICRP (International Commission on Radiological Protection). thin layer of adsorbent or porous material. 1992. Radiological protection in biomedical research. Addendum In liquid chromatography a solution of the compounds is 1 to ICRP publication 53. Radiation dose to patients from radio- dropped on the upper end of a column of adsorbent material. The pharmaceuticals, Annals of the ICRP, ICRP Publication 62, Vol. column is developed by adding a solvent to the upper end of the 22. Pergamon Press, Oxford, UK. column. In the aforesaid methods the radioactive distribution is deter- PA (posteroanterior) projection mined and a radio chromatogram is produced. The locations of (General) See Posteroanterior (PA) projection the activity peaks can then be used for identifying the compounds in the sample. Prior to measurement, the system has to be cali- brated with known samples. PACS (picture archiving and communication system) Abbreviations: PC = Paper chromatography, TLC = Thin (General) See Picture archiving and communication system layer chromatography, LC = Liquid chromatography and HPLC (PACS) = High-performance liquid chromatography. Further Reading: Berthold, F. and M. Wenzel. 1967. Pair production Radiochromatographic counting techniques. In: Instrumentation (General) Pair production dominates photon interaction processes in Nuclear Medicine, ed., G. Hine, Academic Press, New York, at high energies (above 50 MeV). The incident photon interacts Vol. 1, Chapter 11, pp. 251–273. 677 Paraffin phantom 678 Parallel imaging Paraffin phantom results in undersampling of spatial frequencies. On transforma- (General) A tissue-equivalent phantom for medical imaging and tion to the spatial domain, undersampling becomes apparent as therapy, made from paraffin wax. fold over or ‘aliasing’ of the image. Uses in Medicine: Dosimetry – Paraffin phantoms can be Parallel imaging allows acquisition of a reduced number of used in the estimation of absorbed dose in a region of the body k-space lines, while avoiding the aliasing problem. The accelera- resulting from external exposure to ionising radiation. An appro- tion in scan time achieved is given by the factor R. Parallel imag- priate radiation sensitive material (radiographic film, OSL or ing employs multiple coil elements, each with a known spatial TLD material, etc.) is placed within the phantom and the phan- sensitivity to signal (Figure P.2). tom is exposed. The total dose to a point from the exposure can There are two methods of reconstructing an image using par- be read from the radiation sensitive material in the usual manner. allel imaging. In SENSE imaging (Figure P.3), reconstruction Since the phantom has similar attenuation and scatter character- takes place after the Fourier transform, in the spatial (image) istics to tissue, this dose is assumed to be representative of that domain. Data is collected simultaneously from all coil elements which would be given to a tissue or organ located in the equivalent and undersampled k-space is filled for each coil. A Fourier trans- position in a patient. form generates an

aliased image from each coil. As coil spatial Radiotherapy: Paraffin phantoms can also be used in radio- sensitivities differ, each image generated will be unique. If the therapy to compensate for ‘missing tissue’. Radiation travelling spatial sensitivity of a given coil j is Cj(x,y), then the image Ij(x,y) through the phantom will exit with a modified energy spectrum for the jth coil element of N coils is and spatial distribution, allowing the use of a standard radiother- P apy treatment plan when oblique beams are used, or when the I j (x,y) = C j (x,y)r(x,y) + C j (x,y + Dy)r(x,y + Dy) beam must pass through an inhomogeneous medium or an uneven skin surface. + C j (x,y + 2Dy) ´ r(x,y + 2Dy) (P.1) Parallel acquisition technique (PAT) + C j (x,y + RDy)r(x,y + RDy) (Magnetic Resonance) Parallel acquisition technique (PAT) is a vendor term (Siemens) for partial parallel imaging, i.e. MRI where ρ(x,y) represents signal intensity corresponding to point acquisition techniques using multiple receiver coils to allow (x,y) in the ideal, unaliased MRI image of the object. Each suc- k-space undersampling. The receive profiles of the coils must be cessive term aforesaid represents an alias term modulated by the spatially dependent and complement each other in at least one coil sensitivity profile Cj(x,y) for that particular coil j. The number direction. If these are known, the origin of the signals can be of alias terms equals the sense factor R. The function ρ(x,y) is reconstructed from the (otherwise aliased) MR images (SENSE) periodic over a spatial extent of the field of view (FOV) and the or the undersampled k-space signal (SMASH). spatial shift in each term Δy is FOV/R. Related Articles: SMASH, SENSE Each of the N coil elements generates an expression in the form of (P.1). Expressing in matrix notation Parallel connection (General) Elements of an electric circuit are considered to be connected in parallel if the current divides between them at one æ I1 ( x, y) ö æ C1 ( x, y) C1 ( x, y + Dy) . C1 (x, y + RDy) ö side of the connection and reunites at the other side of the con- ç ç 2 ( , ) ÷ ç I x y ÷ ç C2 ( x, y) ( , + ) ÷ C2 (x, y + Dy) . C2 x y RDy nection. If two or more components are connected in parallel ÷ ç ÷ = . ç ç ÷ ç . ÷ they have the same potential difference (voltage) across their ÷ ends. ç è IN ( x, y)÷ ç ø èCN ( x, y) CN (x, y + Dy).CN ( x, y + RDy)÷ ø Related Article: Serial connection (P.2) æ r( x, y) ö ç ÷ Parallel imaging r D ´ ç ( x, y + y) ÷ (Magnetic Resonance) Parallel imaging is a technique enabling ç . ÷ faster MRI scanning without the need to reduce image resolu- ç ÷ ç èr( x, y + RDy)÷ tion. In conventional MR imaging, the number of lines in k-space ø acquired must equal the number of voxels in the phase encoding direction. For a fixed field of view, each line added in the phase encode direction improves resolution but costs time. Reducing scan time by skipping the acquisition of lines in k-space (Figure P.1) Coil sensitivity profiles Coil 1 Lines in Coil 2 FoV phase Imaged encode direction object Object k-space Image of object with rows shows Coil 4 Coil 3 unfilled aliasing FIGURE P.1 Skipping acquisition of k-space lines reduces scan time but FIGURE P.2 Parallel imaging uses multiple coil elements, each with a causes aliasing in the image. different spatial sensitivity to signal. Parallel imaging 679 P arallel imaging Acquired k-space data Reconstructed, for each coil aliased image from Unaliased each coil image ky Coil 1 FFT SENSE reconstruction kx ky FFT Coil 1 Coil 2 kx Coil 2 ky Coil n Coil n FFT k Preacquired x coil sensitivity maps FIGURE P.3 Acquisition and reconstruction of an image using the parallel imaging SENSE method. P which can be simplified to: Beam 1 I = Cr (P.3) where I and ρ are vectors C is a matrix representing the coil sensitivities The expression can be solved for the unaliased image data ρ once the number of coils N is greater than or equal to the number of ‘unknowns’ in the rows of the vector ρ˜. That is, the number of coil elements must equal or exceed the acceleration factor R. It is also necessary that the individual coil spatial sensitivities are sufficiently different over the volume of interest. In an alternative realisation of parallel imaging called ‘SMASH’ (also called GRAPPA), the ‘missing’ k-space lines are filled prior to image reconstruction by the FFT. For an individual coil, the k-space value (i.e. the signal intensity) corresponding the Beam 2 spatial frequencies kx, ky is FIGURE P.4 Illustration of a parallel opposed field arrangement. Sj (kx ,ky ) = òòC j ( x, y)r( x, y)exp(-ikx x,-ikyy)dxdy (P.4) By weighting individual coil elements, j by a weight wj, the overall profile can be established with spatial frequency Δky through the profile C(x,y) can be controlled. With weightings set to w j 0 to give use of appropriate weighting factors w j Dk (this assumes the indi- an overall uniform sensitivity profile (i.e. combined C(x,y) = 1) the vidual coil profiles are suitable for this purpose): lines in k-space are filled with values S(kx, ky): C x y w j Dk ( , ) = å Dk C j ( x, y) = exp(iDky ) (P.7) N j S (k k w j x , y ) = å 0 S j (kx ,ky ) (P.5) j =1 Rewriting (P.6) for the case of weightings w j Dk chosen to give a Substituting (P.4) into (P.5) sinusoidal sensitivity, the signal now takes the form N òòr( x, y)exp(-ikx x,-ikyy + iDky )dxdy S (k k w j x , y ) = åòò 0 C j ( x, y)r( (P.8) x, y)exp(-ikx x,-ikyy)dxdy j=1 (P.6) This is simply the FT of ρ(x,y) as in (P.6) mentioned previously, = òòr(x, y)exp(-ikx x,-ikyy) but shifted in spatial frequency by Δky. The missing lines in dxdy k-space can then be constructed through weighted combination of the signals from each coil as From the definition of the Fourier transform, this represents the FT of ρ(x,y), F(ρ). This data is used to fill lines as shown in the cen- N tre panel of Figure P.4. The ‘missing’ lines displaced by multiples S (k j x ,ky + Dky ) = åwDkS j (kx ,ky ) (P.9) of Δky remain to be filled. A combined coil sinusoidal sensitivity j =1 Parallel opposed fields 680 Parallel plate ionisation chamber For both SENSE and SMASH techniques the coil sensitivities In radiotherapy, it is generally the case that the total dose Cj(x,y) must be known. The sensitivities can be measured in a that can be tolerated depends on the volume of tissue irradiated prescan or in some variants of the techniques by acquiring addi- – the dose–volume effect. It has been suggested that groups of tional calibration k-space line during scanning. cells within an organ are organised into collective bodies called Implementation of parallel imaging reduces SNR by minimum functional subunits (FSU). The arrangement of the FSUs within factor of R , all other factors remaining equal. In practice, the the tissue is thought to be an important factor in determining the SNR reduction with parallel imaging is somewhat worse due to volume dependence of an organ. In parallel organs, the FSUs are imperfections in coil geometries, given by a geometry factor ‘g’: arranged in parallel so the inactivation of a small number of FSUs does not lead to loss of organ function. Inactivation of a criti- SNR cal number of FSUs is required for functional damage to occur, SNRparallel = (P.10) meaning that there should be a threshold volume of irradiation g R below which no functional damage will develop even after high- With parallel imaging, SNR is variable across the image. dose irradiation. Above this threshold there is a graded rather Further Readings: Dietrich, O. et al. 2002. iPAT: Applications than a binary response: Functional impairment increases in sever- for fast and cardiovascular MR imaging. Electromedica 2:122– ity with increasing dose. 146; Kellman, P. 2004. Parallel Imaging: The Basics, ISMRM When planning radiotherapy treatment, assessment is always Educational Course–MR Physics for Physicists, International made of the dose to normal tissues and modern commercial treat- ment planning systems have a number of tools to aid this process P Society for Magnetic Resonance in Medicine (ISMRM) 12th Scientific Meeting, Kyoto, Japan. including dose–volume histograms (DVH). For parallel organs, complications are likely to be dependent on the average dose to the whole volume. Reduction of the volume of normal tissue Parallel opposed fields irradiation increases the dose that can be received without induc- (Radiotherapy) This form of treatment involves two coaxial ing an adverse effect. For some organs this is quite a significant beams entering the patient at 180° to one another (Figure P.4). effect. For example, Emami suggests that for the kidney reduc- It is typically used where it has not been possible to accurately ing the volume irradiated from the whole organ to only a third define the tumour, or for palliative treatments involving a low increases the tolerance dose from 23 to 50 Gy. dose. These treatments tend to be relatively straightforward plans Examples of parallel organs are kidney, lung, liver. It should to calculate, and a manual calculation may be performed with the be noted however that most real normal tissues have a mixed par- dose distribution being calculated from simple data tables and allel and serial architecture. knowledge of the separation between the two beam entry points Abbreviations: FSU = Functional sub-unit and DVH = Dose– on the skin surface. The dose is generally prescribed to the mid- volume histogram. point of the target volume. Related Articles: Adverse events, Dose volume histogram, The summation of the two opposing percentage depth dose Dose response model, Serial organs, Sigmoid dose–response curves yields a plateau-shaped region as demonstrated in curve, Tolerance Figure P.5 for two 6 MV beams separated by 20 cm. Further Readings: Emami, B., J. Lyman, A. Brown, L. Coia, Related Articles: Palliative treatment, Beam arrangement, M. Goitein, J. E. Munzenrider, B. Shank, L. J. Solin and M. Geometric field separation Wesson. 1991. Tolerance of normal tissue to therapeutic irradia- tion. Int. J. Radiat. Oncol. Biol. Phys. 21:109–122; Hall, E. J. and Parallel organs A. J. Giaccia. 2006. Radiobiology for the Radiologist, 6th edn., (Radiotherapy) Radiation treatment inevitably affects normal tis- Lippincott Williams & Wilkins, Philadelphia, PA; Withers, H. R., sue and so may cause radiation-induced adverse effects. The toler- J. M. G. Taylor and B. Maciejewski. 1988. Treatment volume and ance of normal tissues to radiation depends on the ability of the tissue tolerance. Int. J. Radiat. Oncol. Biol. Phys. 14:751–759. clonogenic cells to maintain a sufficient number of mature cells suitably structured to conserve organ function. The tissue archi- Parallel plate ionisation chamber tecture is thought to be important in determining the tolerance (Radiotherapy) Parallel plate ionisation chambers may have dose for partial organ irradiation. cylindrical or plane parallel plates. The latter form are typically used in radiotherapy and are mainly designed for low energy x-rays and electron beams that feature steep depth dose gradients. According to most dosimetry protocols, parallel plane chambers 120 are recommended to be used at all electron energies, and below 110 10 MeV their use is mandatory. They are suitable for photon beam 100 reference dosimetry when a calibration in terms of absorbed dose 90 80 to water is available at the relevant beam quality. They are also 70 suitable for reference dosimetry for proton and heavy ion beams. 60 In high energy photon beams, plane-parallel chambers are useful 50 for measurements in the dose build-up region. Figure P.6 shows a 40 30 diagram of a parallel plate ionisation chamber. 20 The design of a chamber with parallel plates is relatively com- 10 Beam 1 Beam 2 Total dose plex due to the presence of a guard electrode and a

large amount of 0 0 2 4 6 8 10 12 14 16 18 20 back-scatter material. A guard electrode is an important element Distance (cm) of the construction of a parallel plate chamber because it is useful not only to obtain a homogeneous electrical field in the sensitive FIGURE P.5 Demonstration of the central axis dose plateau. volume of the chamber (Figure P.7) but also to avoid secondary PDD Parallel-hole collimators 681 Parallel-hole collimators a 3 1 2 3 d m g FIGURE P.8 Typical parallel plate ionisation chamber. (Image courtesy of Standard Imaging Inc.) P Further Reading: Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, p. 149. FIGURE P.6 Diagram of a parallel plate ionisation chamber. (Image courtesy of Standard Imaging Inc.) Parallel-hole collimators (Nuclear Medicine) A specific collimator design used to attain a spatial resolution with a scintillation camera. The collimator consists of a high absorbing material with parallel holes drilled Polarizing Sensitive or cast in it. Photons with propagation near-parallel to the holes’ (biasing) volume direction will pass through and reach the detector. Photons with electrode an oblique angle of incidence will most likely be absorbed when passing through the lead septa separating two holes and will therefore be prevented from reaching the detector. A parallel-hole collimator will produce an image with the same size and orienta- tion as the object size. By altering the length and width of the holes it is possible to Guard V electrode construct collimators with different features, e.g. long holes with Measuring a small width will produce an image with high spatial resolution (collecting) but low count rate since a lot of photons will be absorbed by the electrode collimator and vice versa when having short broad holes. A It is necessary to increase the septa width when using radionu- clides emitting high energy photons to prevent septal penetration. Thick septa on the other hand will absorb too much of the incident FIGURE P.7 Schematic drawing of a parallel plane ionisation chamber. radiation and therefore decrease the collimator efficiency. Septal penetration yields a decrease in spatial resolution and it is there- fore important to consider the photon energy when selecting a electrons scattered from the chamber wall being counted in the collimator. The parameters for a number of different collimator chamber. types can be seen in Table P.2. The electrical field lines of plane parallel ionisation chamber A collimator is one of the degradable factors affecting the are parallel to the direction of the incident radiation (Figure P.8). system spatial resolution. The contribution from a parallel-hole The secondary electrons are predominantly forward-directed, collimator is and therefore, they can gain energy if the entrance window is neg- atively charged and the collecting electrode is positively charged. æ l This may lead to more ionisation events than in the opposite R eff + b ö coll » d çç ÷ (P.11) è l ÷ polarity. Therefore, in plane parallel ionisation chambers, the eff ø polarity effect is more pronounced than in cylindrical chambers. where Parallel plate chambers are generally not well suited for scanning b is the source to collimator distance measurements in water because they create relatively strong dis- d is the hole diameter turbances in the water as they move and it is difficult to set water leff is the effective length of the holes, which includes septal surface level accurately due to the adhesion to the water menis- penetration cus. Therefore, parallel plate chambers are more commonly used in combination with solid slab phantoms. The characteristics of The spatial resolution is best when the source is located as some parallel plate chambers are given in Table P.1. close to the collimator face as possible and when the holes are Related Article: Ionisation chamber long and thin (small width). Parallel-hole collimators 682 Parallel-hole collimators TABLE P.1 Some Characteristics of Ionisation Chambers Collecting Guard Ionisation Chamber Window Electrode Electrode Ring Recommended Type Materials Thickness Spacing Diameter Width Phantom Material NACP01 (Scanditronix) Graphite window, graphited rexolite 4 mm–90 2 mm 10 mm 3 mm Polystyrene, graphite, Calcam-1 (Dosetek) electrode, graphite body (black mg/cm2 water (with wall), rexolite housing waterproof housing) NACP 02 (Scanditronix) Mylar foil and graphite window, 0.6 2 mm 10 mm 3 mm Water, PMMA Calcam-2 (Dosetek) graphite rexolite electrode, graphite mm–104 body (black wall), rexolite housing mg/cm2 Markus chamber PTW Graphited polyethylene foil window, 0.9 2 mm 5.3 mm 0.2 mm Water, PMMA 23343 NA 30-329 NE graphited polystyrene collector, mm–102 2534 PMMA body, PMMA cap mg/cm2 (incl. cap) P Scdx-Welhofer PPC 05 Window and body C-552, graphited 1 mm–176 0.5 mm 10 mm 3.5 mm Water (PEEK) electrode mg/cm2 Holt chamber (Memorial) Graphited polystyrene wall and 4 mm–416 2 mm 25 mm 5 mm Polystyrene (phantom NA 30-404 electrode, polystyrene body mg/cm2 integration) Capintec PS-033 Aluminised mylar foil window, 0.004 2.4 mm 16.2 mm 2.5 mm Polystyrene (phantom carbon impregnated air equivalent, mm–0.5 integration) plastic electrode, polystyrene body mg/cm2 Exradin 11 Conducting plastic wall and P11: 1 2 mm 20 mm 5.1 mm P11: polystyrene electrodes Model P11: polystyrene mm–104 water equivalent Model A11: C-2552, air mg/cm2 equivalent Model T11: A-150 tissue equivalent Roos chamber PTB FK6 PMMA, graphited electrodes 1 mm–118 2 mm 16 mm 4 mm Water PMMA PTW 34001 Scdx mg/cm2 Welhofer PPC 35 Scdx Welhofer PPC 40 Attix chamber RMI 449 Kapton conductive film window, 0.025 1 mm (0.7 12.7 mm 13.5 Solid water graphited polyethylene collector, mm–4.8 mm mm solid water body mg/cm2 reported) Source: Data courtesy of the manufacturers listed in the first (left column). The geometric efficiency is defined as the fraction of incident TABLE P.2 photons registered by the detector. When using a parallel-hole Collimator Parameters for Some Commonly Used collimator the geometric efficiency, g, is expressed as follows: Collimators 2 2 ù Comment Diameter Septa Length Hole Shape g 2 æ d ö é d » K çç ÷÷ ê ú 2 (P.12) è leff ø ëê (d + t ) ûú Low energy 0.250 0.030 4.100 Hex general where purpose K is a constant depending on the hole shape Medium energy 0.340 0.140 4.200 Hex t is the septal thickness general purpose The geometric efficiency is not dependent on the distance between the source and the collimator face because even if the Low energy 0.180 0.030 4.000 Hex individual hole efficiency is decreased by 1/b2 the irradiated detec- high resolution tor area is increased by b2 so the total count rate is unchanged. But High energy 0.400 0.320 4.000 Hex since the spatial resolution is dependent on the source to collima- general tor distance it is important to place the camera as close to the purpose object as possible. Low energy 0.340 0.050 3.600 Hex Related Articles: SPECT, Collimator, Parallel-hole collima- high sensitivity tor, Diverging collimator, Converging collimator, Collimator, Collimator design, Collimator parameters Paralysable counting system 683 Pareto surface Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Gadolinium in a chelate form is often used as a contrast agent Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, in MR imaging because of its T1-shortening effect. Gadolinium Philadelphia, PA, pp. 239–244. by itself is toxic. Examples of gadolinium chelates are gado- pentetate dimeglumine (Gd-DTPA) and gadoterate meglumine Paralysable counting system (Gd-DOTA). (Radiation Protection) The dead time τ in a detector is a mini- Related Articles: Susceptibility, Ferromagnetism mum time interval which must separate two events to be mea- sured (registered) as two separate pulses. Parent radionucleus In a paralysable counting system the events that occur dur- (Nuclear Medicine) The first radioisotope in a subjectively chosen ing the dead time period are not just missed but will also restart decay chain. In a decay chain the initial radioisotope disintegrates the dead time, so that with increasing event rate the detector will into another radioactive isotope. The second radioisotope is called reach a saturation point where it will be incapable of recording the daughter radioisotope, progeny or decay product. The daughter any event at all. The measured rate M can be expressed as equal to radioisotope will eventually decay to a second decay product, or grand-daughter radioisotope. This chain will continue until one of M = Nx exp(-Nxt) the decay products is stable. In nuclear physics, a specific part of a decay chain can be of clinical and/or research interest (the parent where radioisotope can be chosen subjectively). For instance 99Mo disin- N is a true interaction rate tegrates to the meta-stable state of 99mTc. The parent radionucleus τ is a dead time of the detector system in this example is 99Mo. The relations between parent and daughter P activation is described by the Bateman equations. Related Articles: Deadtime losses, Non-paralysable counting Related Articles: Grand-daughter radionucleus, Daughter system radionucleus, Bateman equations Further Reading: Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, Parent–daughter decay pp. 119–121. (Nuclear Medicine) Parent–daughter decay refers to the situation in which a radioactive daughter nucleus is produced by a decaying Paramagnetic contrast agents parent nucleus. A parent–daughter pair commonly used in nuclear medicine is 99Mo and 99mTc where 99Mo decays to 99mTc. 99m (Magnetic Resonance) Agents based on paramagnetic ions are Tc the most common form of contrast agent used in MRI. A number is continuously produced and eluted from a generator containing 99 of different ions have been proposed (e.g. Mn3+, Dy3+). However, Mo and used for imaging applications. If the 99mTc is not eluted gadolinium (Gd3+) remains the mainstay of commercially avail- regularly the activities of the two radionuclides will reach a tran- able agents. sient equilibrium. Because of the toxicity of heavy metal ions in vivo, for clini- A parent–daughter can be modelled by the Bateman equations cal use paramagnetic contrast agents invariably take the form of which describe a decay sequence with two or more radionuclides: a metal–chelate complex, in which the chelate molecule is added for safety reasons. ìïé l Ad = íêA ( ) d ´ ( p p 0 e-l t - e dt )ù ü -l Paramagnetic agents exert their effect on water protons ú ´ ï BRý through two complementary mechanisms. Innersphere relax- îïë ld - l p û þï (P.13) ation occurs when a water molecule becomes temporarily + A ( d d 0)e-l t coordinated with the metal ion and undergoes relaxation due to magnetic field variations caused by motion of the complex and where relaxation of electrons within the ion. These temporarily bound Ap(t), Ad(t) are the activities of the parent and the daughter water molecules undergo fast exchange with the bulk water, respectively at time t so the relaxation effect is propagated. Outersphere relaxation λp, λd are their respective decay constants occurs as randomly moving water molecules experience vary- BR is the branching ratio for situations where there is more ing magnetic fields due to the presence of the paramagnetic ions than one decay mode for the parent in solution. The relative effects on T1 and T2 relaxation vary The last term in Equation P.13 is the residual daughter-product between different ionic species, but in general these agents are activity that might be present at time t used as T1-shortening positive contrast agents. Related Articles: Contrast agent, Gadolinium chelate, Positive For a 99Mo-99mTc generator t = 0 is set to the first elution; hence, contrast media the daughter activity at t = 0, Ad(0) can be assumed to be zero. Related Article: Transient equilibrium Paramagnetism (Magnetic Resonance) In a paramagnetic material, the electrons Pareto chart are unpaired and bulk magnetic properties are only apparent in (Nuclear Medicine) A special kind of bar chart where the plot- the presence of an externally applied field. When the paramag- ted values are arranged in descending order. The bars are usually netic material is removed from the externally applied field no accompanied by a line graph that represents the cumulative value permanent magnetisation exists in the material. Paramagnetic of the bars. The Pareto chart is used as one of the many tools in materials have a positive magnetic susceptibility χ > 0. Magnetic the quality assurance process. field is strengthened by a paramagnetic material. Examples of paramagnetic materials are: aluminium [13], barium [56], cal- Pareto surface cium [20], gadolinium [31] and magnesium [12]. The

numbers in (Radiotherapy) During Multi-criteria optimisation, an optimi- the parenthesis are the atomic number of the material. zation algorithm calculates a set of plans for which a constraint Paris system 684 Partial Fourier imaging (PFI) (on tumour or organ at risk) can only be improved at the cost of worsening another one. These plans lie on the so-called Pareto surface. Subsequently, the planer can then decide on one plan of this Pareto surface. See Multi-criteria optimisation for more information and literature. BD = (BD1 + BD2 + BD3)/3 Further Reading: C. Thieke et al. “Beyond weight factors: New concepts for defining and analysing dose optimisation,” Radiother. Oncol. 73, S75–S75 (2004). Link: Multi-criteria optimisation FIGURE P.10 Volume implant with a triangular pattern. Paris system (Radiotherapy, Brachytherapy) The Paris system is a predictive defined by perpendicular bisector lines projected from the sides dosimetry system for interstitial brachytherapy originally devel- of the triangles formed by the intersections of the line sources oped for 192Ir wires. This system involves the use of rules and has with the central plane. For a perfect square, the midpoint is used. proved to be simple, reliable and clinically efficacious. In clinical situations these ideal source patterns are difficult The Paris system is based on the following principles: to realise. The implant is acceptable if the elementary basal dose rates satisfy the following relationship (and no triangles have any • Sources must be straight, parallel and arranged so that obtuse angles): their midpoints are located in the same plane, perpen- P dicular to the line sources. This plane is called the cen- 0.9BD £ BDi £ 1.1BD tral plane, and it is the mid-plane of the application. • The reference linear air kerma rate, the apparent linear Dose is specified in relation to the basal dose, and the reference activity, along each line source must be uniform and isodose rate (RD) surface used in the Paris system corresponds to identical for all wires. 85% of the basal dose: • Adjacent line sources must be equidistant. If the vol- ume to be treated is large, then it is necessary to per- RD = 85%BD form implants in more than one plane, the separation between planes being such that the principle of equidis- The reference isodose should encompass the target volume as tant line sources is observed. closely as possible. The treated volume is the volume encom- passed by the reference isodose surface. The high dose volume, This implies, for volume implants, that the intersections of line ‘the hyper-dose sleeve’, is the region around each wire receiv- sources with the central plane form either the corners of squares ing more than twice the reference dose. It is to be noted, that or the apices of equilateral triangles. the Ir-wires extend outside the treated volume for a Paris system Brachytherapy generally produces a very heterogeneous dose implant, and the ratio of treated length to active wire length is 0.7. distribution, with high doses close to the sources, and lower doses The Paris system has been used for many years with good between them. A careful arrangement of the sources, following clinical results. Originally developed for 192Ir-wires, the system the rules in the Paris system, results in a relatively uniform dose can be used also for 192Ir seed ribbons and for afterloading units between sources, and these dose values are used as reference val- with a stepping source when a standard step size and equal dwell ues for the evaluation of the dose delivered to the tumour. Thus, times are used. Modifications based on the Paris system have the dose specification is based on dose values inside the target been introduced for stepping source systems, taking advantage of volume in the central plane. the possibility to optimise the dwell times. The basal dose rate, BD, is defined in the central plane as the Abbreviations: ADC = Analogue digital converter and EPI = mean value of the minimum dose rates, the elementary basal dose Echo planar imaging. rates, BDi: Related Articles: Dosimetry systems, Escargot curves, Cross line curves, Reference isodose volume BD1 + BD1 + + BD Further Reading: Pierquin, B. and G. Marinello. 1997. A BD = n n Practical Manual of Brachytherapy, Medical Physics Publishing, Madison, WI. For a single plane implant with four wires (Figure P.9), the BDi is calculated at the midpoint of segments joining the intersections of Parking position the line sources with the central plane. (Diagnostic Radiology) X-ray equipment where the x-ray tube For volume implants with five wires (Figure P.10), the elemen- assembly moves during the exposure require specific starting tary basal dose rates are calculated ‘in the triangle midpoints’; point of the movement known as Parking position. For example linear (classical) tomography, ortho pan tomography (OPG), some single slice CT scanners, etc. have specific parking position of the x-ray tube assembly. Normally the x-ray exposure begins when the x-ray tube assembly starts its movement from this position. Partial Fourier imaging (PFI) (Magnetic Resonance) Partial Fourier imaging (PFI) is a par- tial acquisition technique in MRI that exploits the fundamental BD = (BD1 + BD2 + BD3)/3 Hermitian symmetry in k-space when magnitude images are of interest. Contrary to partial parallel imaging, it does not require FIGURE P.9 Single plane implant. the use of phased array receive coils. Partial parallel imaging (PPI) 685 Partial volume effect A certain fraction of phase encoding lines is not acquired (starting from high-spatial frequencies, but clearly above 50%). The missing data points are extrapolated from the central k-space signals using specific algorithms (for example Margosian, POCS). Under-representation of low-SNR spatial frequencies will result in deterioration of the point spread function and increased SNR, especially with zero-filling. Partial parallel imaging (PPI) (Magnetic Resonance) Partial parallel imaging (PPI) is the gen- eral term for MRI acquisition techniques using multiple receiver (a) coils to allow k-space undersampling. The receive profiles of the coils must be spatially dependent and complement each other in at least one direction (as expressed by the g-factor). If these are known the origin of the signals can be reconstructed from the (otherwise aliased) MR images (SENSE) or the undersampled k-space signal (SMASH). PPI cannot be performed in a direction along which the coils’ sensitivity changes in the same manner, e.g. axial for head phased P array and transverse for spine phased array. PPI is performed in one of the phase-encoding directions to allow for shorter measurement times (spin-warp) or echo-trains (EPI). It comes at a penalty of reduced signal-to-noise ratio (SNR). Depending on the design of the coil array and direction, (b) the reconstruction problem becomes ill-conditioned at some degree of undersampling, resulting in image artefacts. FIGURE P.12 (a) Partial volume effect results in reduced contrast. (b) Partial volume effect eliminated by use of narrow imaged slice thickness. (Courtesy of ImPACT, UK, www .impactscan .org) Partial volume effect (Diagnostic Radiology) In CT, the partial volume effect (artefact) occurs when a high contrast structure extends only partly into the imaged slice. The CT numbers of the pixels should represent the average attenuation of the voxel within the slice. However, due to the logarithmic relationship between intensity and attenuation val- ues, the calculated average CT number of the pixels will not be a true representation of the materials within the voxels (Figure P.11). The partial volume effect results in a reduced contrast (Figure P.12a). It can be eliminated, and contrast optimised, by using a narrower imaged slice thickness (Figure P.12b). The partial volume effect can also result in artefacts which appear as streaking in the image. This will be the case for off- centre objects projecting slightly into the scan plane. Due to beam FIGURE P.13 Partial volume artefact resulting from inconsistent pro- I jection data. (Courtesy of ImPACT, UK, www .impactscan .org) 1 I2 divergence, inconsistencies will occur between views obtained I1 + I from opposing directions, which give rise to streaking artefacts. 2 When the x-ray tube is at 0° the high density object is in the path ln(I1 + I2) ≠ ln I1 + ln I2 of the beam but with the tube at 180°, the object is not in the path of the beam (Figure P.13). The partial volume artefact will also be FIGURE P.11 Partial volume effect resulting in incorrect CT number reduced by use of a narrower imaged slice thickness. representation. (Courtesy of ImPACT, UK, www .impactscan .org) Related Articles: Streaking artefact, Slice thickness Particle radiation 686 Passive device Particle radiation distribution volume V1 will be smaller than the actual distribu- (Radiation Protection) The term ‘particle radiation’ is normally tion volume Vt and vice versa if the tracer concentration is higher. used to differentiate from electromagnetic radiation, although The distribution volume of the tissue equals the ratio between the photons can be considered as particles. The particles involved can activity and the concentration which is given by be charged or uncharged. The term encompasses not only par- ticles such as protons, neutrons and electrons but also ions. A C t t = (P.16) Related Articles: Beta radiation, Beta+ radiation, Photons, Vt Radiation, Alpha particle Combining these two relationships (Equations P.15 and P.16) and Particle therapy Equation P.14 yields (Radiotherapy) Particle therapy is an alternative name for Hadron V l = 1 (P.17) therapy. See article Hadron therapy for more information. Vt Related Article: Hadron therapy Therefore the partition coefficient can also be interpreted as the Particle velocity distribution volume per unit mass of tissue for a tracer, which is (Ultrasound) When an ultrasound wave propagates in a medium, convenient when estimating blood flow and perfusion. the particles within the medium will start oscillating. The dis- Related Articles: Tracers, Analogue tracers, Distribution placement is the particle’s distance, at a certain time, from its rest volume P position. A transducer surface moving in a sine-shaped movement Further Reading: Cherry, S. R., J. A. Sorenson and M. E. will give rise to sinusoidal motions of the particles with the veloc- Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, ity v. The relations between displacement, particle velocity and Philadelphia, PA, pp. 380–381. pressure for such a wave (expressed in 1D) are Passive beam scattering Displacement: s = s0 cos(wt - kx) (Radiotherapy) In passive scattering, spreading in energy and range of a mono-energetic proton beam is achieved by a range Particle velocity: v = ¶s/¶t = -ws0 sin(wt - kx) modulator wheel (also known as a range-shifter wheel or propel- ler), which rotates to insert incrementally thicker segments of Pressure: P = Zv = -Zws0 sin(wt - kx) plastic into the beam. A spread-out Bragg peak can be obtained by controlling the thickness of each plastic segment and its dura- If the pressure amplitude p and the acoustic impedance are tion in the beam (Khan and Gibbons, 2014). known, the particle velocity amplitude (ωs0) can be calculated as Lateral spreading is achieved by placing two scattering foils in the beam. High atomic number materials scatter the beam with minimum energy loss, while low atomic number materials ws = p/Z 106 / . 1 6 0 = Pa 1 6 ´ 0 Rayls » 0.6 m/s decrease beam energy with minimum scattering. Therefore, scat- tering and beam energy can be controlled through a combination Related Articles: Displacement, Pressure of low and high atomic number scatterers. This is usually referred to as double scattering or the principle of scattering foils (Khan Partition coefficient and Gibbons, 2014). (Nuclear Medicine) In tracer kinetics a compartment may be open Patient-specific hardware such as collimators and compensa- or closed to a tracer. For example when tracers are able to move tors are used to improve the dose distribution by adjusting the between the two compartments of blood and tissue, these com- range to coincide with the distal edge of the target. Compensators, partments are open. After a given period of time the activity ratio usually made of Perspex, shape the beam in the depth direction. in the blood and tissue reaches a steady state. The ratio of tissue Collimators, usually made of brass, shape the beam in the lateral activity concentration Ct (Bq/g) to blood concentration Cb (Bq/ direction. One disadvantage of passive beam scattering is that mL) at this point is known as the partition coefficient l (mL/g). collimators and compensators are an undesirable source of neu- The partition coefficient l (in

a steady state system ) is defined tron dose to the patient (Horton and Eaton, 2017). by The reader is referred to IPEM Report 75 for a schematic dia- gram showing a passive scattering delivery system. C l = t (P.14) Further Readings: Horton, P. and D. Eaton. 2017. Design and Cb Shielding of Radiotherapy Treatment Facilities, IPEM Report 75, 2nd edn., IOP Publishing; Khan, F. M. and J. P. Gibbons. Cb can be directly measured by using blood samples. 2014. Khan’s the Physics of Radiation Therapy, 5th edn., Wolters If one assumes that the tracer concentration in the blood is Kluwer Health. equal to the concentration in tissue, the apparent distribution vol- ume is given by Passive device (Magnetic Resonance) The term passive device refers to any med- A V t 1 = ical device that serves its function without the supply of power. C b (P.15) MRI is contraindicated in the presence of metallic objects inter- A C t b = nal to the patient such as surgical clips, pins, plates, screws, metal V1 sutures or wire mesh because of possible movement of the devices due to the strong magnetic static field and because of the heating where At is the activity in the tissue. If the concentration in the effect determined by the RF exposure. The type of metal used in blood is lower than the concentration in the tissue the apparent such small or large devices is one factor which determines the Passive implant 687 Patient position force exerted on them in the static magnetic field. While non- ferrous metallic implants may show little or no deflection to the field anyway they could cause significant heating due to the RF absorption. The presence of a ferromagnetic passive device in the fringe field of the MR equipment represents the risk of a projectile effect with potential for patient damage or death. Passive implant (Magnetic Resonance) The term passive implant refers to any medical device that serves its function without the supply of power. Examples of passive implants include but are not limited to dental implants, ocular implants, orthopaedic implants and penile Magnetic field strength (H) implants. Some metallic implants have shown considerable torque when placed in the presence of the strong static magnetic field. The force or torque exerted on small and large metallic implants FIGURE P.14 Schematic magnetisation curve. can cause serious effects as unanchored implants can potentially move unpredictably within the body. The mass and the type of metal used in such implants is one factor which determines the PAT (parallel acquisition technique) force exerted on them in the static magnetic fields. While non- (Magnetic Resonance) See Parallel acquisition technique (PAT) P ferrous metallic implants may show little or no deflection to the field they could cause patient burning because of the significant Patch-field technique heating due to their inability to dissipate the heat caused by RF (Radiotherapy) The patch-field technique is an irradiation tech- absorption. nique used in charge particle therapy to improve tumour dose cov- erage and/or decrease critical organ dose. Two irradiation fields Passive shielding are combined so that the first field treats only a segment of the (Magnetic Resonance) Access to the static magnetic fringe field target volume, avoiding adjacent critical organs with the lateral is restricted within the so-called pacemaker safety limit (for mag- beam penumbra. The second field treats the remaining segment, netic fields above 0.5 mT), extends 5–10 m (depending on design) also avoiding the critical organ with the lateral beam penumbra. from a completely unshielded high field strength MRI magnet. The second particle beam matches its distal fall-off 50% dose Therefore most clinical installations require a degree of shielding. with the other field’s lateral penumbra’s 50% dose. Tissue hetero- Usually the magnet is equipped with active shielding (see Related geneities make it difficult to obtain a uniform dose along the patch Article). However, passive shielding is possible and is obtained by junction. The overshoot of the patch field aims to minimise the use of ferromagnetic materials, most commonly iron. low dose region and a combination of patch fields with different The iron can be integral to the magnet housing (self shield- junctions can be used to ensure that target coverage is sufficiently ing) but is usually mounted in the walls/roof/floor of the scan- uniform. ner room. If passive shielding is used in combination with active shielding, partial shielding of the room is often sufficient. This Path length may be the case when the 0.5 mT limit otherwise would extend (Radiation Protection) Path length is the term given to the dis- outside the scanner room (requiring access to be restricted) or if tance that a charged particle travels within an absorbing medium there is equipment in an adjoining room that is very sensitive to before interacting with the electrostatic forces created by the static magnetic fields, such as a gamma camera or a PET scanner. atoms of the medium. It is more usually referred to as the mean The iron shielding works due to its magnetic permeability: free path. Related Article: Mean free path m = B/H [H/m] where Patient position B is the magnetic flux density [T] (General) There are a series of terms used to describe the H is the magnetic field strength [A/m] position of an individual when undertaking different imaging examination. This can be described as the material’s ability to ‘concentrate’ magnetic fields. The passive shielding thus provides a more pref- Erect: Standing or sitting up. For example, an erect chest erable return path for the magnetic flux around the magnet than x-ray. the air would and limits the extent of the static magnetic field. Supine: Lying on the back. For example, the position for The so-called magnetisation curve is shown schematically in supine abdominal imaging. Figure P.14, showing B as a function of H. Eventually, for a suf- Prone: Lying on the front. ficiently high B, the material saturates and cannot further ‘con- Oblique: Turned usually to a specific angle. For example, a centrate’ (or shield) the static magnetic field. 45° angle for an oblique chest projection. Other materials with very high permeability, such as permal- Lateral: Standing, sitting or lying down with one side in loy, could be an alternative but they are expensive and tend to contact with the equipment couch or stand. For exam- saturate at lower magnetic flux densities than iron. ple, erect lateral chest x-ray. Related Articles: Active shielding, Fringe field Decubitus: Lying on the side. For example, imaging the Further Reading: Krestel, E. 1990. Imaging Systems for Medical right kidney in the left lateral decubitus position. Diagnostics, Siemens Aktiengesellschaft, Berlin, Germany. Trendelenburg: Tipping the individual head down. Magnetic flux density (B) P-RDSR 688 Peak assignment P-RDSR scaling, but in general values are still relaxation-weighted and (Radiation Protection) P-RDSR is the acronym for Patient- hence only comparable with values collected using the same pulse Radiation Dose Structured Report (RDSR). sequence timings, unless correction factors are applied. Use of To define a model for the recording and reporting of organ short TE and long TR reduces this problem. doses from medical procedures, the DICOM Standard has speci- There are several approaches to the measurement of peak fied a P-RDSR data object. areas. In early work, it was common to print out spectra and The RDSR incorporates an information set required to quantify the individual peaks by means of manual triangula- estimate the radiation dose to the patient. This information relates tion, or by the crude but effective process of cutting them out and to the modality, i.e. the geometric and technique details and the weighting them. Naturally this approach has now been replaced dose indicators adopted as the summary metric for a specific by computer-based methods, but many of the same problems modality (kerma area product [KAP], CT dose index [CTDI] and remain, such as how to allocate areas between overlapping peaks, dose length product [DLP], etc.). and how to handle broad baseline features. Sophisticated model- The P-RSDR represents the evolution of RDSR because it also based fitting packages, operating in either the time domain or the includes organ dose estimations. frequency domain, allow the user to tackle these issues and to To generate a P-RSDR, modality details alone are not suffi- assess the quality of the resulting fit (Figure P.15). cient since organ doses can only be roughly estimated from this information. Peak assignment To improve the accuracy of the estimation, patient information (Magnetic Resonance) Peak assignment refers to the process by P recording is mandatory (patient dimension, actual scan range, which the resonance peaks in an NMR spectrum are attributed position of the patient on the table, etc.). to specific chemical groups and compounds. For in vivo MRS, it relies heavily on prior knowledge. Patient-reported outcome measures (PROMs) Historically, peak assignment in in vivo spectroscopy has (Radiotherapy) Historically, clinical trials have focused on been a difficult task. It is far from trivial in NMR spectroscopy physician-reported outcome measures. Yet in some cases, physi- in general, since there is no one-to-one relationship between cians and patients will have different priorities regarding disease resonance peaks and specific chemical compounds. A peak may outcomes and quality of life. The recognition that the patient’s contain contributions from a number of compounds containing perspective should be central to the evaluation of healthcare the nucleus of interest in the same chemical environment, while led researchers to develop patient-reported outcome measures compounds that contain the nucleus in more than one chemical (PROMs): standardised, validated questionnaires for completion group give rise to multiple peaks in the spectrum. by patients. They are used to assess patients’ perceptions of their A high field spectrum may contain resolvable signals from health status (impairment), their functional status (disability) and dozens, if not hundreds, of different compounds. At the field their health-related quality of life (well-being). Typically, they strengths used for in vivo work, there is a further complica- are designed to compare patient satisfaction before and after a tion in that these peaks collapse into a small number of com- course of treatment. PROMS have become important within both posite, often overlapping features. Establishing the composition clinical trials and routine patient care. They may be generic or of these features can be a formidable task; particularly as the disease-specific. proportions of different compounds contributing to a peak may Related Articles: Clinical trial endpoints, Quality-adjusted life vary with tissue type, age and pathology. Peak assignment in 31P years (QALYs) spectroscopy is arguably somewhat more straightforward than in 1H spectroscopy, although complicated by such phenomena PBS (Pencil beam scanning) (Radiotherapy) See Pencil beam scanning (PBS) pCT (Proton CT) (Radiotherapy) See Proton CT (pCT) Peak areas (Magnetic Resonance) Fundamentally, the area under a peak in an NMR spectrum reflects the quantity of spins in the corresponding chemical group within the region from which the signal has been 1 2 3 4 acquired. However, the peak area peak may also be dependent upon a number of other factors: The NMR visibility of nuclei may be affected by physicochemical properties such as binding. Peaks due to a specific chemical group may be split due to coupling. Importantly, in a conventional NMR experiment, peak area may be weighted according to T1 (dependent on the sequence relax- ation time, TR) and/or T2 (dependent on the echo time, TE, in a spin-echo experiment). The signal amplitude also depends on hardware factors such as coil sensitivity and signal amplification. Peak area measurement is often used as a crude form of quan- tification, albeit subject to the above caveats. In such an approach, peak areas may be expressed in arbitrary (or ‘institutional’) units, FIGURE P.15 Evaluating peak areas in a proton NMR spectrum. From or in the form of ratios of the area of one peak to another. The bottom to top: raw spectrum, individual fitted peaks, sum of fitted peaks, latter approach provides values that are independent of signal residual noise. Peak kilovoltage (kVp) 689 Peaking as pH-dependent chemical shifts and the broad phospholipid Source Source baseline. Peak assignment has relied heavily on a combination of prior knowledge of tissue biochemistry (i.e. what compounds are d = SSD expected to be present in a specific tissue) and analysis of tis- sue extracts. For example, perchloric acid extracts can be used for analysis of low molecular weight,

water soluble metabolites P P dmax and chloroform-methanol extraction for lipids. Comparison of high-resolution NMR spectra of these different extracts can pro- vide clues as to the identity of specific components. Comparison A A of the more detailed spectra with the results of in vivo studies gives insight into the make-up of composite peaks, and finally assignment can be confirmed by comparison with spectra from FIGURE P.17 Setup for measuring the peak scatter factor (PSF). pure metabolites in solution. Alternatively the extracts can be subjected to other forms of chemical analysis to identify the components. The measurement set-up is shown in Figure P.17. At low energies, dmax = 0 and PSF is known as the back-scatter factor. Peak kilovoltage (kVp) There are two components to the behaviour of PSF with pho- (Diagnostic Radiology) The peak kilovoltage is the maximum, or ton energy – the amount of back-scatter and the energy of the peak, voltage that is applied to an x-ray tube during the duration scattered photons. At low energy, there is a large amount of P of the exposure, or during the electrical power supply cycle as back-scatter but the energy is low, causing rapid absorption in illustrated in Figure P.16. the medium. As the energy increases the amount of scattering The electrical potential (voltage) applied to an x-ray tube might decreases, yet the photons have a higher energy and larger pen- change during the exposure interval, with the waveform depend- etrating power. Hence the PSF increases with energy, reaching a ing on the type of high-voltage power supply or generator being maximum at a particular beam quality, and then decreases with used. The technique factor value that is set and displayed for a further increases in beam energy. The PSF increases with field specific clinical procedure is the peak value, the kVp. This is a size, and may approach a saturation value (see Podgorsak 2003). quantity that can be measured when calibrating x-ray equipment. PSF is a special case of TAR (i.e. PSF(Ad) = TAR(dmax, Ad). The significance of the instantaneous voltage is that it deter- Related Articles: Percentage depth dose (PDD), Scatter air mines the efficiency of x-ray production that varies throughout ratio (SAR), Tissue air ratio (TAR) the cycle. Further Reading: Podgorsak, E. B. 2003. Review of Radiation The significance of the effective voltage is that it determines Oncology Physics: A Handbook for Teachers and Students, the rate of heat production. International Atomic Energy Agency (IAEA), Vienna, Austria. With constant potential generators, where the voltage does not vary during the exposure, the peak, effective and instantaneous Peak systolic velocity values are the same. With sinusoidal power supply the peak value (Ultrasound) Systole describes the contraction of the chambers is equal to the effective value multiplied by √2. For example, of the heart. The peak systolic velocity is the maximum velocity with an effective voltage of 220 V the peak value is 311 V. in the systemic arterial circulation produced by the contraction of the left ventricle (Figure P.18). Peak scatter factor (PSF) Abbreviations: PSV = Peak systolic velocity and EDV = End (Radiotherapy) The peak scatter factor is the ratio of Dp, the diastolic velocity. absorbed dose in tissue at the depth of dose maximum, to D* p , the Related Article: End diastolic velocity absorbed dose due to primary radiation only. The absorbed dose due to primary radiation is measured with just enough material Peak voltage to provide electronic equilibrium (i.e. ionisation chamber with (General) The maximum value of an AC voltage, either positive build-up cap). It is defined as follows: or negative, measured from the point of reference. Peak voltage is only a moderately useful way of measuring AC voltage. The ( ) Dp (dmax , A,d,E ) PSF A,E = most common way to measure/quantify AC voltage is the effec- D* p ( A,E ) tive voltage or RMS voltage (see the eponymous article). Related Articles: Effective energy, RMS voltage, Effective voltage value kVp (peak) Peaking KV kVe (effective) (Nuclear Medicine) The process of setting the energy window centre for a gamma camera or other system that uses energy kVi (instantaneous) discrimination. In gamma camera imaging, the magnitude of the signal gener- 1/120 1/60 1/40 ated by an interaction within the detector crystal is proportional Time (s) to the energy deposited by the incoming photon. Due to this pro- portionality, an upper and lower bound, or window, for the magni- FIGURE P.16 Illustration of the relationship of the peak, effective, and tude of the signal can be defined that will correspond to an energy instantaneous voltage for a full-wave rectified 60 Hz power source for window. The values of the bounds are chosen such that they are x-ray production. (Courtesy of Sprawls Foundation, www .sprawls .org) centred upon the photo-peak of the isotope to be imaged. Pencil beam 690 Pencil beam scanning X e– Y Z Beam broadening device Primary collimator θx Δx Secondary Δy collimator P Z = 0 Bone X΄ P(x, y, z) FIGURE P.18 Peak systolic velocity (PSV) in a monophasic and tripha- sic arterial waveform. A pulse height analyser (PHA) is used to exclude signals that Lung are not within the energy window, therefore, improving image contrast by reducing the number of scattered photons contribut- ing to the image. Peaking is the process of aligning the centre of the energy θx, θy window used by the PHA with the photo-peak of the isotope to be imaged. This can be done manually by choosing the keV of the FIGURE P.19 Schematic representation of the electron pencil-beam window centre having observed the isotope spectrum produced dose calculation. (From Hogstrom, K.R. et al., Phys. Med. Biol., 26(3), by the camera. On modern systems, however, an auto-peaking 445, 1981.) function is usually available, which adjusts the position of the energy window such that an equal number of counts are in each half of the window. about the mean direction (σθ), an energy (Ep,0) and a planar fluence For gamma camera systems, the position of the PHA energy (electrons cm−2). A pencil beam is defined as all electrons passing window can vary day to day by as much as a few keV due to fluc- through a pixel that is typically 2 mm2 at 100 SSD. Only the anat- tuations in the high voltage supply to the camera head and other omy along the central axis of the pencil beam is used to calculate factors. The peaking should therefore be checked and adjusted the dose distribution to the patient from that pencil beam. Four if necessary, as part of the daily quality assurance protocol. It physical properties of the algorithm limit its accuracy: the slab is important to use a scatter free source when peaking, i.e. not a geometry assumption, the planar fluence-base model, the Eyges patient, to ensure correct placement of the energy window. Gaussian scatter distribution function and the inability to model Related Articles: Gamma camera, Photo-peak, Energy win- the secondary electrons. dow, Pulse height analyser (PHA) Further Reading: Hogstrom, K. R., M. D. Mills and P. R. Further Reading: Prekeges, J. Nuclear Medicine Almond. 1981. Electron beam dose calculation, Phys. Med. Biol., Instrumentation, 2nd edn. 26(3):445–459. Pencil beam Pencil beam scanning (Radiotherapy) A number of algorithms for computing the dose (Radiotherapy) Pencil beam scanning (PBS), also known as spot distribution similar to the ones used for photon beams have been scanning or scanned beam therapy, is a method of delivering applied for electron beams. Accurate planning with electron accelerated protons to uniformly cover a target volume. beams is complicated when the beams are incident on anatomical Either a cyclotron or a synchrotron can deliver a narrow beam regions containing heterogeneous tissue. A common algorithm is of protons via the beamline to the treatment room. Without any based on the decomposition of a broad clinical beam into pencil modulation, this beam can deliver a single Bragg peak or ‘spot’ beams each affected by the presence of small tissue inhomogene- at an energy dependant depth. This spot has a finite size that will ities in their direct path. The conventional pencil beam algorithm be smaller than the intended treatment volume, so to treat a larger was introduced by Hogstrom (1981). The concept of the Hogstrom volume the spot must be moved in three dimensions. pencil beam model is shown in Figure P.19. Each pencil beam has To change the lateral location of the spot, two sets of orthog- a size (Δx, Δy), a mean direction (θx, θy), an RMS angular spread onal magnets are used. Because protons are positively charged, SSD Lo SCD Penetration depth 691 Pentetreotide their path will be deflected by the magnetic field. The stronger the magnetic field, the more the beam will be deflected from the iso- centre. Spots are delivered with varying magnetic field strengths to ‘paint’ a single energy layer. The energy is then changed, and the process repeated to cover the whole tumour target volume, layer by layer. The depth of ultrasonic penetration is influenced by both tech- P nological and anatomical factors. The penetration depth decreases with an increase in the transducers fundamental output frequency, while operator adjustments to the focus position can also influ- ence penetration depth. Highly reflected anatomical structures will also significantly reduce beam penetration posterior to them. Ultrasound transducers for cardiac applications are designed so the beam avoids traversing the rib cage by using a footprint suit- able for travelling between them. This avoids the large amount of attenuation from the bone which would prevent adequate imaging of the heart. Pentetreotide (Nuclear Medicine) Pentetreotideis a somatostatin analogue used for tumour imaging for neuroendocrine malignancies. Somatostatin is a small peptide produced by the neuroendocrine system and somatostatin receptors can be found on many cells of the neuroendocrine system, but they are also found on other cells not affiliated with the system. Pentetreotide attached to 111In is used for SPECT or planar imaging of tumours that express soma- tostatin receptors. The latter are overexpressed on many tumours originating from the neuroendocrine system but also on non- Upper diagram: path of proton spot (deflected magnetically) to ‘paint’ a neuroendocrine tumours. It can be used for initial diagnosis, re- single energy layer. Lower diagram: combination of spot maps for mul- staging, selection of patients for treatments with peptide receptor tiple energy layers. radionuclide therapy (PRRT) and assessment of therapy success. 111In-Pentetreotide (111In-DTPA-octreotide) is commercially Abbreviations: PBS = Pencil beam scanning. available as Octreoscan and is delivered as two vials. Vial A con- Related Articles: Pristine Bragg peak, Spot size, Scanned tains 111In, while vial B has a combination of 10 µg of pentet- beam reotide and other substances for long-term stabilisation. For the preparation of the kit, vial A is added to vial B. Routine quality Penetration depth assurance tests often encompass the measurement of activity con- (Ultrasound) Penetration depth refers to the furthest distance centration by assessing the activity in the vial using a radionuclide from which transmitted ultrasound echoes are received. The calibrator and the assessment of radiochemical purity using thin- depth is measured from the point most proximal to the probe layer chromatography (TLC). surface to the furthest point of visible speckle, or if in reference After injection of usually around 120–220 MBq of to Doppler imaging, the further depth either a pulse wave signal is 111In-Pentetreotide, physiological uptake is seen in the liver, obtainable or colour Doppler is visible. spleen, pituitary gland, thyroid and kidneys. 111In-Pentetreotide B-mode penetration can be sub-categorised into low contrast clears quickly from the blood with less than 1% remaining after penetration depth and high contrast penetration depth. This refers 24 hours and the main route of excretion is through the kidneys to the depth of reflections from objects that are minimally and with urinary excretion being approximately 85% at 24 hours. highly reflective, respectively. As shown, the low contrast speckle Related Articles: Receptor targeting, Indium-111 ends at a depth of about 50mm, whereas echoes are received from Further Readings: Bombardieri et al. 2010. 111In-pentetreotide the high contrast pins at 70 mm. scintigraphy: Procedure guidelines for tumour imaging. EJNMMI Penetrating radiation 692 Percentage depth dose 37:1441–1448; Sharp, Gemmell and Murray. 2005. Practical Penumbra effect Nuclear Medicine, 3rd edn., Springer. (Diagnostic Radiology) A penumbra is a partial shadow between regions of full shadow (the umbra) and full

illumination. This Penetrating radiation produces a blurring effect in images that are in the form of shad- (Radiation Protection) This is radiation which has sufficient ows like radiographs. In the past, the term ‘penumbra’ was used energy to penetrate a material which it is incident on. X-rays, to describe this effect, but more appropriate and accurate terms gamma rays and beta particles are types of penetrating ionising are blurring, visibility of detail and spatial resolution. The pen- radiation. An obvious example of penetrating non-ionising radia- umbra (partial shadow) around or at the edge of objects leads to tion is light when incident on materials such as glass or water. increased blurring of image details and reduced spatial resolution. The properties of the radiation (especially the wavelength/energy) The penumbra occurs because the effective focal spot of an and of the material itself determine how penetrating the radiation x-ray tube is not a point source of radiation. If it was a point will be. source, the x-ray projection of an object (the object ‘shadow’) to the detector would have very sharp boundaries (Figure P.22a). Penumbra In reality, the focal spot has some finite dimensions (often of (Radiotherapy) The penumbra of a beam is normally defined as the order of 0.3 to 1 mm). The x-rays originate from all the areas the distance between the 20% and 80% points of the dose profile, of the focal spot, creating a blurred region around the boundar- measured at 10-cm depth in a water phantom (Figure P.20). ies of the object ‘shadow’ (Figure P.22b). This blurred region is There are three components to the penumbra: transmission called penumbra (see Figure P.22b, c and d, with small, large and P through the collimators, scatter penumbra and the geometric pen- extra-large penumbra). umbra from the physical size of the source (i.e. physical penum- Obviously, the larger the focal spot size, the larger the pen- bra). The geometric penumbra is the most significant contribution umbra effect (Figure P.22c) – i.e. more prominent blur exists, overall and is illustrated in Figure P.21. creating radiographic images with reduced spatial resolution. For cobalt units, penumbras are generally 15-mm wide (cf. The penumbra also increases with increased magnification of the ∼ 3−5 mm for LINACs). Reducing the distance b (as shown in radiograph (by manipulating the position of the object between Figure P.20) will reduce the penumbra width, but at the expense of the focal spot and the detector) (Figure P.22d). increased scatter dose from the head components. The smaller the Related Articles: Magnification; Focal spot effective penumbra the more localised the delivered dose, enabling more accurate matching of fields. Percentage depth dose (Radiotherapy) A percentage depth dose curve, PDD, is com- monly used in radiotherapy to graphically describe the dosi- metric characteristics of a radiation beam. It shows the dose deposition as a function of depth, d, in a particular medium, nor- Head malised to maximum dose at dmax. The percentage depth dose calculation, for depth d, field size As and source–surface dis- tance (SSD), is given in Equation P.18. The field size parameter Penumbra Penumbra As refers to the value at SSD or at dmax, not necessarily at the point of measurement: DD ( D P d, A SSD) = 100 d s , Ddmax Tails Tails (P.18) D (d, As ,SSD) = 100 FIGURE P.20 Schematic of the beam profile showing the three regions. D (dmax, As ,SSD) Photon PDD: A typical PDD for a MV photon beam in a tis- s sue-equivalent material is shown in Figure P.23. The beam enters giving a surface dose Ds, and as it penetrates the dose rises rap- idly, reaching a maximum value at zmax, after which it decreases almost exponentially. If the beam then exited the phantom, its a value Dex would be slightly lower than the extrapolated line due to the absence of scatter from points beyond the exit dose point. The low surface dose (10%–30%) is known as the skin spar- SSD P = s (b + d) ing effect, which is absent in orthovoltage or superficial beams. a The subsequent build-up of dose results from the long range of b secondary electrons released from the photon interactions, which carry energy deeper into the phantom. Charged particle equilib- rium (CPE) is said to exist at z = zmax where z is approximately equal to the range of secondary charged particles, and is where d the dose reaches its maximum. The dose then falls off due to pho- ton attenuation, resulting in a transient CPE. Electron PDD: A typical PDD for a MeV electron beam is a FIGURE P.21 Geometric penumbra, s – size of source. tissue-equivalent material shown in Figure P.24. The build-up is Percentage depth dose 693 P ercentage depth dose P FIGURE P.22 Indicative example for penumbra effect (the diagram shows only the distal x-ray beams from both ends of the focal spot of the x-ray tube): (a) ideal case of point source (microscopic focal spot) – no penumbra: (b) small focal spot (e.g. 0.5 mm) of the x-ray tube – small penumbra (i.e. blur) exists on both sides of the object shadow; (c) large focal spot (e.g.1 mm) of the x-ray tube – large penumbra exists on both sides of the object shadow; (d) large focal spot with additional magnification – extra-large penumbra exists, which presents a very blurred image. q Source Patient 100 R 90 Dmax = 100 50 0 zmax zex Dex 0 R90 R50 Rp Rmax DS Depth in water (cm) 0 zmax Depth z zex FIGURE P.24 Typical electron beam percentage depth dose curve illus- trating the definition of Rq, Rp, Rmax, R50 and R90. (From Podgorsak, E.B., Review of Radiation Oncology Physics: A Handbook for Teachers and FIGURE P.23 Typical percentage depth dose curve, showing dose depo- Students, International Atomic Energy Agency, Vienna, Austria, 2003.) sition in a patient from a megavoltage photon beam caption title. (From Podgorsak, E.B., Review of Radiation Oncology Physics: A Handbook for Teachers and Students, International Atomic Energy Agency, Vienna, Austria, 2003.) maximum, due to the electrons’ continuous energy loss and scat- ter. An additional consideration is the bremsstrahlung tail pro- duced from the electrons interacting in the head, air and patient. less pronounced for electron beams than for photon beams, due to The magnitude of this contribution depends on electron energy; the larger percentage surface dose for electrons (∼80%). A single typical contributions are less than 1% for 4 MeV beams, and less electron deposits the majority of its energy at the end of its track than 4% for a 20-MeV beam. The strong scattering of electrons in (a Bragg peak); however, the energy deposited along the central air necessitates the use of applicators placed near the skin surface axis of a beam of electrons is more diffuse, a ‘smeared out’ Bragg of patients, to help collimate the beam and reduce unnecessary peak. This is due to the oblique scattering of the electrons into the dose outside the field. central axis. The surface dose increases with increasing energy Abbreviation: PDD = Percentage depth dose. (opposite to that of photons), due to the greater and larger angle Further Reading: Podgorsak, E. B. 2003. Review of Radiation electron scatter at low energies. At high energies, scatter is pre- Oncology Physics: A Handbook for Teachers and Students, dominately forward. The dose undergoes a sharp fall off after the International Atomic Energy Agency, Vienna, Austria. Percent depth dose (%) Perception 694 Periodic motion Perception CBF quantification using the Kety–Schmidt equation. The most (General) In its general sense, perception is the process by common MRI methods for potential absolute quantification of which information about one’s environment is sensed, organ- CBF are arterial spin labelling (ASL) and dynamic susceptibil- ised and interpreted. Perception in medical imaging involves ity-contrast MRI (DSC–MRI). In ASL the tracer is created by the understanding of images and sounds and has implications inversion of spins in a brain-feeding artery. The inversed spin for image acquisition, processing, display and operator training population is transported to the tissue with the blood and sub- in order to optimise the understanding of features for diagnosis sequently reduces the longitudinal tissue magnetisation in pro- and therapy. portion to the regional CBF. DSC−MRI is an intravascular bolus-tracking technique based on a temporary contrast-agent Perfusion imaging induced signal loss during the bolus passage. First-passage moni- (Magnetic Resonance) In physiology, the term perfusion refers to toring of the intravenously injected gadolinium tracer concentra- the flow of blood through the capillary system. Perfusion or tis- tion at a rate of approximately one image per second is required sue blood flow (TBF) is traditionally quantified in terms of the for subsequent calculation of cerebral blood volume, MTT and volume of blood transported to a given mass of tissue per unit CBF. Deconvolution of the measured tissue concentration func- time, and a commonly used unit is mL/(min 100 g). Other haemo- tion with the arterial input function (AIF) yields the tissue residue dynamic parameters such as blood volume and mean transit time function scaled by the CBF. Absolute quantification is at pres- (MTT) are sometimes loosely included in the term ‘perfusion ent hampered by difficulties in retrieving a reliable AIF. Finally, parameters’. Additionally, a number of non- or semi-quantitative dynamic CT perfusion imaging is an emerging technique, based P perfusion indices have been proposed in various applications. on basically the same bolus-tracking concept, using intravenous Perfusion imaging refers to the use of a medical-imaging modality injection of an iodinated contrast material. Figure P.25 displays for measurements of tissue blood flow. In quantitative approaches, examples of CBF maps obtained by different modalities. The the regional perfusion is normally visualised as a parametric map top row of Figure P.25 shows (from left to right) Tc-99m-labelled calculated on the basis of an appropriate tracer-kinetic model. ECD SPECT (patient case), Xe-133 SPECT (normal volunteer) 1 Perfusion imaging is of particular interest for the assessment of and H 5 2 O PET (normal volunteer). The bottom row (from left to cerebral blood flow (CBF) and myocardial blood flow (MBF), right) illustrates pulsed ASL MRI (normal volunteer), DSC-MRI although investigations of numerous other organs or tissue types (normal volunteer) and dynamic CT (patient case). exist. A number of medical imaging modalities and techniques Related Articles: Mean transit time (MTT), Arterial spin are used for the assessment of tissue perfusion, for example, labelling (ASL) SPECT, PET, MRI and CT. Further Readings: Schwitter, J. 2006. Myocardial perfu- Myocardial Blood Flow: Qualitative assessment of myocar- sion imaging by cardiac magnetic resonance. J. Nucl. Cardiol. dial perfusion has traditionally been carried out by myocardial 13:841–854; Wintermark, M., M. Sesay, E. Barbier, K. Borbély, scintigraphy, that is by SPECT imaging of radiolabelled tracers W. P. Dillon, J. D. Eastwood, T. C. Glenn, C. B. Grandin, S. such as Tl-201 (thallium chloride) and Tc-99m-labelled sestamibi Pedraza, J.-F. Soustiel, T. Nariai, G. Zaharchuk, J.-M. Caillé, V. and tetrofosmin. Using these tracers, absolute quantification is Dousset and H. Yonas. 2005. Comparative overview of brain per- prohibited by the fact that the myocardial uptake is not propor- fusion imaging techniques. Stroke 36:83–99; Detre, J. A. and D. tional to the myocardial blood flow (MBF) at high flow rates. C. Alsop. 1999 May. Perfusion magnetic resonance imaging with Absolute quantification of MBF is feasible by use of PET in com- continuous arterial spin labeling: Methods and clinical applica- bination with H 15 2 O as a freely diffusible tracer. MRI has emerged tions in the central nervous system. Eur. J. Radiol. 30(2):115–124. into an interesting alternative for myocardial perfusion imaging with excellent spatial resolution, primarily exploiting the first- Perfusion imaging pass tracer kinetics of conventional T1-shortening extracellular (Nuclear Medicine) This refers to the imaging of the perfusion gadolinium-based contrast agents. During the first passage, the process, namely the nutritive delivery of arterial blood to the cap- contrast agent enters the microvasculature and starts to diffuse illary bed. into the interstitial space, leading to increased signal intensity. In nuclear medicine perfusion studies are performed on a rou- Various imaging strategies exist, but a magnetisation preparation tine basis, e.g. in lung perfusion studies and myocardial perfusion by saturation recovery followed by a steady-state free precession imaging (MPI). pulse sequence for readout is a common approach. Absolute quan- In lung ventilation/perfusion studies, the examination is tification

of MBF by cardiac MRI is indeed a challenging task, divided into two parts: ventilation and perfusion. The ventilation and it is fair to conclude that several problems remain. test is designed to check that air is able to reach all parts of the Cerebral Blood Flow: Relative distributions of cerebral blood lung. The perfusion test is used to identify a potential pulmonary flow (CBF) are most commonly obtained by SPECT imaging of embolism. A pulmonary embolism in a vein causes the blood flow the intracellular retention of a Tc-99m-labelled tracer, such as to decrease in the vein following the embolism, hence a lower hexamethyl propylenamino oxime (HMPAO) and ethyl cystein- relative perfusion compared to unobstructed veins. ate dimer (ECD). A historically important diffusible tracer for Abbreviation: MPI = Myocardial perfusion imaging. regional CBF quantification in absolute terms is Xe-133. The Related Articles: Technetium generator, Cyclotron Xe-133 tracer can be either inhaled or intravenously injected, the tracer kinetics is typically monitored by SPECT and CBF is Periodic motion calculated from modifications of the Kety–Schmidt tracer model. (Magnetic Resonance) In magnetic resonance, periodic motion Somewhat similarly, the xenon-enhanced CT technique relies on refers to regular patient movement such as cardiac and respiratory inhalation of stable xenon and CBF calculation using a modified motion. This type of motion can cause artefacts to occur in the Kety–Schmidt equation. PET is regarded as the gold standard for images. A ghost of the image occurs in the phase encoding direc- CBF quantification. PET measurements are performed using bolus tion, which can cause problems if the ghost overlays a part of the injection of H 15 2 O or continuous inhalation of C15O2, followed by image which is clinically relevant. Periodic table 695 Peripheral nerve stimulation P FIGURE P.25 Examples of cerebral blood flow (CBF) maps. Top row (from left to right): Tc-99m-ECD SPECT, Xe-133 SPECT, H 15 2 O PET. Bottom row (from left to right): pulsed ASL MRI, DSC-MRI, dynamic CT. There are several ways in which this type of artefact can radiation beam and the air before the beam reaches the patient. be reduced. The first is by breath-hold imaging where data is Routine checks are carried out on radiotherapy units to ensure acquired in a single breath-hold. This is only useful if the pulse that the peripheral dose is as minimal as possible. sequence is extremely quick and if the patient is well enough to hold their breath for the appropriate length of time. The next pos- Peripheral nerve stimulation sible solution is gating which can be used with both cardiac and (Magnetic Resonance) A peripheral nerve stimulation (PNS) can respiratory motion to ensure that the images are acquired at the occur in patients undergoing MRI or MRS if magnetic field gra- same point in the cycle every time. The final possibility is to swap dients are switched rapidly. A high switch rate results in a short the frequency and phasing encoding directions so that the artefact rise time for the gradient field and a subsequent rapid change of does not overlay the anatomical region of interest. the magnetic flux density in the patient. The time for the gra- Related Article: Motion artefacts dient fields to reach the maximum gradient amplitude, which is expressed in mT/m, is known as the rise time, measured in Periodic table units of microseconds. The faster the rise time the greater is the (General) This is a table with the 117 known basic elements organ- likelihood of PNS. Typical gradient systems are capable of pro- ised in a systematic manner. The table is designed to demonstrate ducing gradients from 20 to 100 mT/m. The PNS has been stud- the recurring (periodic) nature of the element properties. Due to ied in volunteers under different experimental circumstances. the smart design of the table, properties of an element can be con- PNS occurs above a threshold which is expressed in terms of cluded from its position. The elements are arranged in ascend- dB/dt given in mT/s. The PNS threshold level is proportional to ing order by atomic number. Elements with similar properties t0.5 where t is the switching time of the gradients. The modulus are placed in the same column. Each horizontal row represents of the gradient vector field |B| is closely correlated to the PNS the filling of a quantum shell in the atomic model. The rows and threshold level than Bz, the imaging component of the gradient columns are also referred to as periods and groups respectively. field. In normal MRI sequences the induced currents are of a Each element is represented by its specific element symbol and few tens of milliamperes per square metre, a value that is below the atomic number. current densities present in the normal brain and heart tissue. In single-shot techniques the rapid switching of magnetic field Peripheral blurring gradients is able to produce induced currents that exceed the (Diagnostic Radiology) See Unblanking nerve depolarisation threshold and cause peripheral nerve stim- ulation. The number of gradient switching operations per unit of Peripheral dose time determines the duration of the induced current while the (Radiation Protection) The peripheral dose is the unintentional rate of change of the magnetic field determines the maximum dose which is delivered to tissue outside the primary radiation amplitude of the induced current also referred as the maximum field. The common sources of this peripheral dose may be leakage slew rate with a given gradient system. The likelihood of PNS radiation from the radiotherapy unit treatment head, or scattered occurring is greatest during echo-planar imaging, mainly in the radiation from the collimators and from interactions between the acquisitions in oblique planes of section, since a higher slew rate Permanent implant 696 Permeability-surface area product results from the combined contributions of gradients from more keV); 103Pd (half-life 17.0 days, photon energy 20.1 and 23.0 keV) than one axis. The likelihood of PNS is also greatest when the is also used (Figure P.27). readout gradient at echo-planar imaging is in the craniocaudal Related Articles: Brachytherapy, Interstitial brachytherapy, direction. Temporary implant, Iodine-125 Permanent implant Permanent magnet (Radiotherapy, Brachytherapy) (Magnetic Resonance) A magnet where the magnetic field origi- Duration of the Treatment in Brachytherapy: There are two nates from a permanent ferromagnetic material is called a per- different types of brachytherapy treatments. These are permanent manent magnet. The permanent magnet doesn’t need any power and temporary depending in general on the time for which the supply or cooling as the electro- or superconductor magnets. The radioactive material is applied: iron-core limits the fringe-field but the weight of the magnet lim- its the magnetic field, B0, to 0.4 T. The weight and the cost of the 1. Permanent implants magnet is the main disadvantage and there are also inhomogene- a. The sources are placed in the target volume by an ities in the permanent magnetic field, limiting the field-of-view interstitial technique. (FOV) (Figure P.28). b. The sources are then left permanently in the target Related Articles: Electro-magnet, Field-of-view (FOV), to decay. Fringe-field, Magnet, Resistive magnet, Superconductive magnets c. The total dose is delivered over a long time with P decreasing dose rate. Permeability-surface area product d. There is just one implant procedure for the whole (Nuclear Medicine) The permeability-surface area (PS) product treatment. is a composite constant that describes the extractive properties e. This is a low dose rate technique. of the capillary bed. PS is the product of permeability P for the 2. Temporary implants capillary wall and the surface area S available for transfer and has a. The sources are placed in the target volume by the same unit as flow, that is mL/h. PS relates to the blood flow Q either interstitial or intracavitary techniques. and the extraction fraction E (see Related Articles) by b. The sources stay in the target volume until the cor- rect dose is delivered. PS = -Q ln (1- E ) c. Fixation of applicators can be used for short HDR treatments, with good control of both total dose and dose distribution. d. The treatment is often delivered in several frac- tions, i.e. several implant procedures are required. e. Temporary implants are suitable for all dose rates, high dose rate, low dose rate and pulsed dose rate. Permanent implants are often used today for the treatment of ‘early small prostate tumours’ (low risk patients). Small radioac- tive seeds are placed in the prostate according to a planned pattern under transrectal ultrasound guidance (see Figure P.26), using a template to position the interstitial needles. The most common source used is 125I (half-life 59.4 days, average photon energy 28.4 FIGURE P.27 Seeds inserted into implant needles: sterile conditions. Usual seed size: diameter 0.8 mm and length 4.5–5.0 mm. B0 FIGURE P.26 Permanent interactive prostate implant showing the seeds placed in the prostate, the ultrasound transducer and the Foley catheter in FIGURE P.28 Indicative diagram of a magnet (resembles a horse-shoe the bladder with contrast in the balloon. magnet) inside an MRI-scanner with a permanent magnet. Personalised medicine 697 Personnel dosimetry or, radiation doses to staff, patients and the public to be as low as reasonably practicable (ALARP), start with engineering E = 1 - e-PS /Q control, followed by the use of written safety instructions, and finally the use of PPE. These measures are described in more Related Articles: Extraction Fraction, Personal dosimetry detail here. Further Reading: Peters, A. M. 1998. Fundamentals of tracer In an ideal radiation facility, exposure of persons is com- kinetics for radiologists. Br. J. Radiol. 71:1116–1129. pletely avoided by containing the radiation source within a shielded room without allowing any access by persons during exposure. The boundaries of the room have appropriate struc- Personalised medicine tural protection (lead, concrete, etc.) such that no radiation dose (General) Personalised medicine is a medical model focused on from the facility is measurable outside. Furthermore, if any per- the individual characteristics of each patient. The patient’s risk son attempts unauthorised or uncontrolled access to the room factors and treatment response are considered, which are specific during an exposure, there will be locks to prevent such access, to each individual or group based on their specific genotypes and or interlocks that automatically make the room safe to enter, phenotypes. Personalised medicine relies on a large set of IT tools either by switching of the electrical supply to x-ray generating for prediction and analysis of treatment response at various levels, equipment, or by withdrawing a radioactive source to a shielded primarily molecular and genome. container. Personalised medicine does not accept the classical one-size- However, in many circumstances it is necessary for persons fits-all approach commonly used in today’s healthcare. In contrast to be present within the room during radiation exposures – for P to this, personalised medicine relies on tailor-made solutions for example during medical procedures such as x-ray examinations. diagnostics, treatment and prevention to satisfy people’s individ- In these situations it is necessary to have written safety instruc- ual health needs. tions (sometimes called local rules or systems of work) together Personalised medicine has been a subject of discussion with operating protocols to tell those present what they should between expert bodies, the WHO and other related organisa- and shouldn’t do during the work procedure to reduce their radia- tions for people-centred health care, which lies at the heart of the tion exposure to a minimum. European health policy framework Health 2020. It is only if the radiation risk assessment determines that nei- ther engineering controls and structural shielding, nor written instructions, can prevent the possibility of persons receiving a sig- nificant radiation dose that the use of PPE should have to be con- sidered. If it is considered necessary, then the radiation employer has a duty in law to provide appropriate and suitable PPE, and to ensure that it is maintained. Employees also have a duty in law to wear or use any such PPE that is provided. Related Articles: Risk assessment, As low as reasonably prac- ticable (ALARP), Lead apron Further Reading: IAEA. 1996. International Basic Safety Standards for Protection against Radiation and for the Safety of Radiation Sources, Safety Series No. 155, International Atomic Energy Agency (IAEA), Vienna, Austria. Personnel dosimetry (Radiation Protection) Radiation workers must be monitored to ensure that the dose they receive from occupational exposure does not exceed dose limits. Dosimeters

used for personnel dosime- try are most often passive devices (such as film badge, OSL and Personalised vs. classical medicine. others), but also may be active devices (such as the electronic semiconductor dosemeter). Dosimeters have also been designed Further Reading: NHS. Personalized Medicine, www .e nglan to monitor specific parts of the body. All personnel dosimeters d .nhs .uk /h ealth care- scien ce /pe rsona lised medic ine; World Health require specific calibration. Organization. www .who .int. The specific use of the terms dosimeter and dosemeter in English refers to the following common meanings: Personal protective equipment (PPE) (Radiation Protection) Exposure to external radiation may be • Dosimeter is usually a passive device used for patient restricted or reduced by the use of appropriate personal protec- or staff dosimetry. Examples of dosimeters are film tive equipment (PPE), either worn by an individual (such as lead badges, thermoluminescent dosimeters, OSL, etc. aprons), or placed between the individual and the source of radia- • Dosemeter is an active device which measures dose tion (e.g. bench shields). PPE is assumed not to include structural (or dose rate) in real time (e.g. ionisation chamber, radiation shielding such as lead-lined walls and doors, or lead- semiconductor dosemeter, etc.) (Figures P.29 and glass windows. P.30). The use of PPE is regarded as the last resort in the hierarchy of control measures identified by a radiation risk assessment Related Articles: Extremity dosimeters, Film badge, Personal when planning a facility designed for carrying out work with protective dosimetry, Pocket dosimeters, Whole body dosimeters, ionising radiation. Such control measures, designed to reduce Dose limits Perspex characteristics 698 Perspex characteristics Perspex characteristics imaging and dosimetry phantoms for quality control purposes. (Radiation Protection; General) Perspex is a thermoplastic – a An example of a Perspex dosimetry phantom for use in CT scan- polymer that turns to liquid when heated, but is solid at normal ners is given in Figure P.31. ambient room temperatures (melting point approximately 140°C). In radiation protection, Perspex is frequently used as a shield- It is also transparent at optical wavelengths (it is transparent from ing material against beta radiation, for example from the follow- approximately 300 to 2800 nm); however, it is less than half the ing radionuclides commonly used in medicine – 32Phosphorus density of normal plate glass. As such it can be used as a light- (32P), 35Sulphur (35S), 14Carbon (14C), and 90Strontium (90Sr) and weight alternative, although it does have disadvantages such as 90Yttrium (90Y). Perspex is used because its low effective atomic being easily scratched. It is also known and commercialised number (∼6.5) means that the absorption of beta radiation by under different names: Plexiglas, Lucite, etc. Chemically it is a the Perspex does not give rise to significant bremsstrahlung as poly(methyl methacrylate), sometimes referred to as PMMA. would be the case with higher atomic number materials such as Perspex has a range of uses in medicine (implants, orthopaedic lead which is used for shielding from photonic radiations – for surgery, etc.), as well as proving a useful material for use in imag- example x-ray or gamma radiation. ing phantoms, and in radiation protection. The following is a summary of some of the characteristics of In radiation dosimetry, small strips of Perspex are used during Perspex that prove useful in medical physics: the gamma irradiation process. The optical density of the Perspex changes with the gamma dose absorbed by the material, and a Effective atomic number: 6.48 (cf. water 7.42) spectrophotometer can be used to measure the change and thus Density: approximately 1.15–1.19 g/cm3 (water 1.0) P determine the absorbed dose. Linear attenuation coefficient (at 70 keV): 0.2184 cm−1 In medical imaging (e.g. x-ray including CT, and nuclear med- (water 0.1945) icine), Perspex is often used as a material for the construction of Mass attenuation coefficient (at 70 keV): 0.1836 cm2/g (water 0.1945) CT number: approximately +120 (water +0) Range of 32P beta particles in Perspex: 6.1 mm FIGURE P.29 Passive personnel dosimeters: top – TLD-based ring and FIGURE P.31 CT Dosimetry Phantoms made from Perspex. (Courtesy finger stall; bottom – film badge (open and closed). of ImPACT, UK, www .impactscan .org) FIGURE P.30 Passive personnel dosimeters with OSL (closed and open). Pertechnetate 699 Phase angle Related Articles: Beta radiation, Bremsstrahlung, Acrylic Further Reading: Wernick, M. N. and J. N. Aarsvold. 2004. Hyperlinks: IAEA, www .IAEA .org; CTUG: www .c tug . Emission Tomography. The Fundamentals of PET and SPECT, o rg .uk /meet 04 -01 -13 /c tdi _p erspe x _tis sue _e quiva lent_ phant oms Elsevier, London, UK, pp. 195–197. .p df Phagocytosis Pertechnetate (Nuclear Medicine) A mechanism of human physiology that (Nuclear Medicine) A pertechnetate is a compound containing allows for the concentration (localisation) of a radiopharmaceu- the TcO- 4 ion. tical in a particular organ of interest for nuclear medicine imag- In the example of the technetium generator, the pertechnetate ing. The reticuloendothelial cells of organs such as liver, spleen is extracted from the technetium generator using a liquid solution and bone marrow can recognise 99mTc labelled sulphur colloid of sodium chloride. The extracted compound is sodium pertech- particles that are ~ 100 nm and can rapidly remove them from netate (Na TcO- 4) and it is the basis for a wide diversity of radio- circulation and sequestering them in the organ. About 80% to pharmaceuticals used for diagnostic purposes. 85% of the colloidal particles accumulate in the liver, 5% to 10% in the spleen and the remainder in the bone marrow. The 99mTc PET (positron emission tomography) labelled sulphur colloid particles are permanently trapped. This (Nuclear Medicine) See Positron emission tomography (PET) allows for the organ to be imaged easily with a gamma cam- era, aiding the nuclear medicine physician to arrive at a medical PET clinical applications diagnosis. (Nuclear Medicine) The clinical use of positron emission tomog- Phagocytosis of colloids is governed by several factors such as P raphy (PET). The most common clinical use for PET is tumour blood flow to the organ, reticuloendothelial cell integrity and the localisation using 18F-FDG which targets cells with high glu- size and number of the administered particles. The size of the col- cose uptake, i.e. high metabolism. PET is more sensitive and loidal particles is important in the imaging of the reticuloendothe- has a higher resolution compared to SPECT. Table P.3 gives the lial system. Larger particles (>100 nm) preferentially get trapped most common radionuclides used for PET imaging and their in the liver and spleen; smaller particles (<20 nm) accumulate in applications. higher concentrations in the bone marrow. Abbreviation: PET = Positron emission tomography. Related Articles: Radiopharmaceuticals, Radionuclide Further Reading: Miller, J. and J. Thrall. 2003. Clinical imaging molecular imaging. J. Am. Coll. Radiol. 1(1):4–23. Further Readings: Bushberg, Seibert, Leidholdt and Boone. 2012. The Essential Physics of Medical Imaging, 3rd edn., Lippincott Williams and Wilkins; Saha, Gopal B. 2004. PET/CT Fundamentals of Nuclear Pharmacy, 5th edn., Springer. (Nuclear Medicine) The combination of a PET scanner and a CT scanner in a single imaging system. The PET scanner will acquire functional information from the in vivo tracer bio-distribution Phantom which is complemented by the morphological information given (Ultrasound) A phantom is an artificial object that mimics the by the CT. The CT data can also be used to estimate a low-noise ultrasonic interaction in the human body. It consists of a tissue attenuation correction for PET. With such an accurate attenua- mimicking material, targets and a container (Figure P.32). The tion correction the PET scanning time (15–30 min) is shorter than acoustic properties set to imitate human tissue characteristics for a comparable stand-alone PET. Using the morphological CT are usually speed of sound, attenuation coefficient and scatter- information it is possible to separate specific uptake (e.g. tumour- ing characteristics, but sometimes properties such as elasticity specific) from non-specific and to more accurately localise malig- and thermal properties are important. A common phantom mate- nancies. The CT also provides additional diagnostic information, rial is an aqueous gel mixed with graphite particles to produce for example, tumour size or other lesions not evident in the PET ultrasound images with a similar speckle pattern and attenuation image. to human tissue. Nylon wires are often used as ‘point targets’ in the imaging plane to test axial and lateral resolution and calliper accuracy (Figure P.33). Phantoms are used predominantly for calibration and qual- ity assurance of ultrasound equipment, training of personnel, TABLE P.3 research and development. Phantoms are designed to allow tests Common Radionuclides in PET Imaging and Their of important imaging parameters such as geometrical accuracy, spatial and contrast resolution, calliper accuracy, penetration and Applications image consistency. Radionuclide Half-Life Molecular Imaging Applications Related Articles: Quality assurance, Spatial resolution, Contrast resolution 18F 1.8 h Metabolic activity (tumour, inflammation, infection), receptor binding, transporter Phase angle function, enzyme activity (Magnetic Resonance) The acquired signal from an MR mea- 11C 20 min Metabolic activity (myocardium, cardiac surement is complex, i.e. it has both a real and an imaginary part. infarction detection), receptor binding The net magnetisation vector M (or the net magnetic moment) of 15O 2 min Metabolic activity (cognitive function) an undisturbed spin population is initially parallel to the main 13N 10 min Protein synthesis, cell proliferation magnetic field B0. Immediately after excitation, M is rotated (mitotic rate) away from the direction of B0, which produces a vector com- 124I 4 days Antibody binding ponent Mxy of M in the transverse xy plane, perpendicular to B0 (Figure P.34). Phase coherence 700 P hase coherence What’s in a phantom – target groups and their use Dead zone Penetration/contrast/spatial resolution 1.0 cm 2.0 cm 2.0 cm 5.0 5.0 cm mm Axial/lateral Horizontal group –1.5 dB resolution for linear arrays 1.0 cm 4.5 cm 2.0 cm +5 dB Contrast resolution 1.0 cm +3 dB 2.0 cm 2.0 cm 2.0 –3 dB cm Lateral and 1.5 cm axial calliper Horizontal group –5 dB accuracy for sector array 5.0 4.0 cm cm –15 dB P 8 6 4 3 2 FIGURE P.32 Description of different target groups in a commercial ultrasound phantom for calibration and quality assurance use. (Courtesy of EMIT project, www .emerald2 .eu) Z B0 M y MXY x FIGURE P.33 Ultrasound image of a phantom is illustrated here. The white dots are echoes from nylon wires; the wires have a consistent diam- eter throughout the phantom but appear larger at depth due to decreased FIGURE P.34 The magnetisation vector M after an excitation pulse. lateral resolution. The phantom also contains circular ‘cysts’ of varying The angle between the y-axis and the vector component of M in the trans- diameters (to the right of the image). (Courtesy of EMIT project, www verse plane (Mxy) is called phase angle (φ). .emerald2 .eu) Related Articles: Diffusion, Larmor frequency/precession, The vector M is precessing at the Larmor frequency around Phase contrast, Phase-mapping, Velocity mapping the B0 vector. To simplify interpretation, a rotating coordinate sys- Further Reading: Dixon, R. L. and K. E. Ekstrand. 1982. The tem is often used, which also rotates around B0 at the Larmor fre- physics of proton NMR, Med. Phys. 9(6):807–818. quency. In this coordinate system, the vector M is static, and thus also the vector component of Mxy in the xy plane is not rotating. Phase coherence The phase angle can then be defined as the angle between an appro- (Magnetic Resonance) The phase φ is defined as the angle that priate coordinate axis in the rotating reference system and Mxy. Mxy in the transverse plane describes with one of the axes (x,y), The phase angle information is used in a variety of MR meth- where Mxy is the projection of the total magnetisation M onto ods, for example diffusion MRI, susceptibility weighted MRI this plane (Figure P.34). Mxy is the resultant of all spins that are (SWI), MR phase-contrast flow measurements and in phase- polarised in the xy plane. contrast angiography. Additionally, phase images can be used in Mxy precesses around the z-axis with the Larmor frequency. mapping field inhomogeneities. The RF-excitation that polarises the spins in the xy-plane makes Phase contrast 701 Phase-contrast imaging all excited spins start their precession at the same angle φ. If all velocity in three orthogonal directions (see Velocity encoding, spins experience exactly the same external magnetic field, they Phase contrast). will have exactly the

same frequency and will rotate with the Two image sets are recorded, with opposite velocity encoding same angular velocity, corresponding to ω = γB0. The spins will gradients. In these two image sets, static spins will acquire phase thus rotate synchronously, all having the same phase angle at any shifts that are equal in magnitude but opposite in sign, and a mag- given point in time. In MRI, this is called phase coherence. nitude subtraction cancels this signal. This property is especially Related Articles: Larmor frequency, Phase contrast MRI, attractive when performing MRA, as good background suppres- Phase dispersion sion is crucial for getting diagnostic images. Moving spins, on the other hand, will accumulate phase shifts Phase contrast of different magnitude as well as phase. In the magnitude subtrac- (Magnetic Resonance) The acquired signal from a MR measure- tion, a signal will remain (Figure P.35). ment is complex. That is, it has both a real and imaginary part, A PC-MRA sequence is most often a 3D gradient echo corresponding to the vector components of the magnetisation sequence with a short TR and short TE. Velocity encoding is vector in the x- and y-directions in the rotating reference system, performed in three orientations, in order to capture motion in all respectively (Figure P.34). directions. The data is post processed with maximum intensity Most MR images are presented as magnitude images, which projection (MIP) to get a 3D overview of the vessel tree. is a vector addition of the real and imaginary parts of the signal: PC-MRA is clinically used in brain, abdomen and the extrem- ities. For an example, see Figure P.36. M (S ) = (Re(S ))2 + ( 2 Im (S )) Related Articles: Contrast-enhanced angiography, Maximum P (minimum) intensity projection, Phase contrast, Time of flight (TOF), Velocity encoding (VENC) where M(S) represents the magnitude of the acquired signal Re(S) and Im(S) are the real and imaginary part of the signal, Phase-contrast CT respectively (Diagnostic Radiology) See Phase-contrast tomography The phase of the signal is calculated as Phase-contrast imaging (Diagnostic Radiology) Phase-contrast imaging (PCI) refers to æ Im (S ) ö the plethora of x-ray imaging techniques that enable the conver- j = arctan çç è Re( ) ÷ S ÷ sion of the phase distortion (or phase shift) of x-ray waves travel- ø ling through a sample into detectable intensity modulations. PCI Ideally, the phase signal should be zero in a normal MR image. However, due to, e.g. eddy currents and field inhomogeneities, the phase value is often not zero. Likewise, if the spin is moving dur- ing imaging, it acquires a net phase offset. Generating an image of the phase distribution gives what is known as phase contrast images. This can be utilised in mapping the velocity of moving spins in the image, as the phase is related to the velocity. For fur- ther information regarding phase-contrast, see Velocity encoding. Related Articles: Velocity encoding, Phase contrast angiography Further Reading: Dixon, R. L. and K. E. Ekstrand. 1982. The physics of proton NMR, Med. Phys. 9(6):807–818. Phase contrast angiography (Magnetic Resonance) MR can be used to visualise blood ves- sels by several methods, commonly called magnetic resonance angiography (MRA). The method described next uses the phase- contrast, or velocity encoding (Phase contrast MRA [PC-MRA]) technique. With this method, bipolar gradients are used to encode FIGURE P.36 PC-MRA examination of the brain. M M 2 1 φ φ2 – = M1–M1 FIGURE P.35 Moving spins acquire signals with different amount of phase under the two different velocity encoding gradients. Subtraction of the magnitude of the signal results in a non-zero remainder. For static spins, the subtraction cancels the signal. Phase-contrast imaging 702 Phase-contrast imaging is mainly used for imaging light materials (e.g. soft tissues) which all the conventional diagnostic techniques, while δ is proportional are highly transparent or exhibit a poor attenuation contrast. to the phase shift, that is, the source of contrast in PCI. All PCI techniques stem from the same physical description For any low-atomic number material and x-ray energies of x-rays as waves which, propagating through a sample, suffer a of interest in the radiological practice (10–100 keV), δ is two change in their phase (phase shift) in addition to the reduction of or three orders of magnitude larger than β (see Figure P.38), their amplitude (attenuation). The phase shift causes a distortion implying that phase shift is much greater than attenuation. in the wavefront meaning that the local direction of the x-ray wave For this reason, enabling to detect phase shift related effects, is modified by the sample. A sample can be entirely described PCI is more sensitive to density variations of low-Z samples by the spatial distribution of its complex refractive index n (see than conventional attenuation-based x-ray imaging. In this Figure P.37): context, PCI has been applied to many biomedical applica- tions encompassing breast, lung, musculoskeletal, vascular n = 1 - d + ib and brain imaging. In the last several years, a variety of PCI techniques have been where the imaginary term β is proportional to the attenuation developed. With the only exception of Bonse–Hart interferom- coefficient of the material, that is, the source of contrast for nearly etry, which is seldom used nowadays, all of them are sensitive to P FIGURE P.37 An x-ray beam, represented by the plane wave ψ, impinges on a sample described by the refractive index n(x,y,z). The wave ψout emerg- ing from the object is affected both in amplitude and in phase. Arrows represent the local direction of the x-ray wave. FIGURE P.38 Tabulated values of δ and β for polymethyl methacrylate (PMMA) plastic, often used as tissue equivalent material in phantoms. Phase-contrast imaging 703 Phase-contrast imaging FIGURE P.39 Comparison between conventional attenuation (a) and PCI (b) (obtained in propagation-based configuration) images of a foraminifer shell. (Reproduced with permission from Spanne, P. et al., Phys. Med. Biol. 44(3):741, 1999.) the first or second derivative of the phase shift. This means that Non-Interferometric: phase contrast typically arises at the interfaces between the detail P and the embedding background (see Figure P.39), where the phase • Propagation-based imaging (PBI), or free-space changes abruptly, in the form of single or double pairs of dark/ propagation or in-line holography, is the simplest bright fringes. Optionally, ‘pure phase’ maps can be obtained by PCI to implement. If a source of sufficient coher- adequately integrating the first or second derivatives of the phase ence is available, phase effects can be observed by shift: the pool of algorithms allowing this integration process is simply distancing sample and detector without using referred to as phase retrieval. In addition to absorption and phase any optical element. This technique is sensitive to the information, some PCI techniques (e.g. analyser-based imaging) second derivative of the phase shift and the produced are sensitive to the ultra-small-angle scattering (USAXS) signal, phase-contrast signal is strongly dependent on the which is the result of a very large number of coherent scatter- (propagation) distance between sample and detector. ing events happening at the nanoscale. USAXS is related to the At present, it is the only PCI technique that has been presence of sub-micrometric features in the sample (e.g. nanopar- applied for in vivo clinical studies in the field of mam- ticles) that cannot be resolved individually, but whose presence mographic imaging. can be detected from the overall scattering signal. Techniques • Analyser-based imaging (ABI) is based on the use of enabling the detection of USAXS are often referred to as ‘scatter- a single crystal, referred to as ‘analyzer’, positioned ing’ or ‘dark field’ imaging. downstream with respect to the sample, whose narrow Most PCI techniques have been pioneered and developed at reflectivity (or ‘rocking’) curve acts as an angular band- synchrotron radiation facilities and are now widely used in many pass filter. Depending on the angular deviation gener- synchrotron beamlines. Some of them have already been success- ated from the phase shift due to the sample, the x-rays fully translated to conventional x-ray sources (e.g. x-ray tubes), are reflected or absorbed by the crystal. This technique while others could be widely used with the advent of compact requires the beam to be monochromatic and collimated, synchrotron-like sources. which is usually achieved through one or more addi- PCI approaches can be divided into two classes: interferomet- tional crystal systems placed upstream with respect to ric and non-interferometric techniques. the sample (e.g. double crystal in Bragg–Bragg geom- etry). This technique is sensitive to the first derivative of Interferometric: the phase shift and can be adapted to detect the USAXS signal. • Bonse–Hart interferometry was the first PCI technique • Edge-illumination (EI) imaging is conceptually similar to be implemented in the 1960s. Its layout requires a to the analyser-based imaging technique. In its simplest monolithic crystal interferometer consisting of three version, it consists of a couple of apertures engraved in blades used in Laue–Laue geometry, while the sample an absorbing material, positioned upstream and down- is positioned between the second and the third blade. stream with respect to the sample, respectively. The This technique is directly sensitive to the phase shift. beam is shaped down to a narrow blade by the first aper- • Talbot/grating interferometry (GI) makes use, in its ture, then it traverses the sample, and it impinges on the simplest set-up, of two gratings, referred to as ‘phase’ edge of the second aperture that is placed in front to and ‘analyzer’ gratings, usually positioned downstream the image detector. This arrangement acts as an angular with respect to the sample at a Talbot distance one to filter since, as a function of the relative displacement of the other, and it is sensitive to the first derivative of the the two apertures, x-rays deviated by the sample will be phase shift and USAXS signal. This technique has been absorbed or transmitted through the second aperture. EI successfully translated from synchrotron to conven- is sensitive both to the first derivative of the phase shift tional sources by positioning a third ‘source’ grating, and to USAXS. This technique has been successfully producing an array of mutually incoherent, but indi- translated from synchrotron to conventional sources. vidually coherent ‘sourcelets’. The latter arrangement is • Tracking-based imaging refers to a broad cate- referred to as Talbot–Lau configuration. gory of PCI techniques where a known structure is Phase-contrast tomography 704 Phase-contrast tomography superimposed to the incoming radiation field and the PCT makes use of the entire formalism developed for conven- distortions caused by the sample to this structure can tional attenuation computed tomography. Conceptually, the only be directly tracked. This structuring can be imposed difference between conventional and phase-contrast CT lies in the through many different arrangements, e.g. making use physical content of the projection images that are fed to the recon- of a lenslet array, a microprobe, an absorption grid, a struction algorithm. In fact, while projections in conventional CT phase grating or a speckle pattern, and the distortion rely on the sample attenuation, projection images in PCCT con- can be tracked with either sub-pixel resolution analysis tain a signal related to the phase-shift imparted by the sample to or by using Fourier-based analysis. the incoming x-rays. In the context of x-ray imaging, the sample can be described Related Articles: Propagation-based imaging, Analyser-based by its refractive index spatial distribution n(x,y,z): imaging, Edge illumination, Grating interferometry Further Readings: Bravin, A., P. Coan and P. Suortti. 2012. n(x, y,z) = 1- d(x, y,z) + ib(x, y,z) X-ray phase-contrast imaging: From pre-clinical applications towards clinics. Phys. Med. Biol. 58(1):R1; Endrizzi, M. 2018. l = 1 - d(x, y,z) + i m(x, y,z) X-ray phase-contrast imaging. Nucl. Instrum. Methods Phys. 4p Res. Sect. A 878:88–98; Olivo, A. and E. Castelli. 2014. X-ray phase contrast imaging: From synchrotrons to conventional where β is proportional to the linear attenuation coefficient µ, λ is sources. Rivista del nuovo cimento 37(9):467–508; Pelliccia, D., the x-ray wavelength and δ is responsible for the sample-induced P M. J. Kitchen and K. S. Morgan. 2017. Theory of X-ray phase- phase-shift and it is proportional to the electron density, viz. the contrast imaging. In Russo P. (ed.) Handbook of X-ray Imaging: number of electrons per unit volume. Physics and Technology, CRC Press, pp. 971–998; Rigon, L. Since multiple images of the sample at different rotation 2014. X-ray imaging

with coherent sources. In Brahme A. (ed.) angles must be recorded to perform tomographic reconstruction, a Comprehensive Biomedical Physics 2, Elsevier, Amsterdam, pp. rotated reference frame is usually defined, where θ is the rotation 193–220; Spanne, P. et al. 1999. In-line holography and phase- angle about the vertical axis y, while the rotated transverse coor- contrast microtomography with high energy x-rays. Phys. Med. dinates are defined as ξ and η (see Figure P.40). Mathematically, Biol. 44(3):741. each projection will be the line integral of a fundamental property of the object (e.g. its attenuation) along the propagation direction Phase-contrast tomography of the beam η. In case of conventional tomography, after a suitable logarith- (Diagnostic Radiology) Phase-contrast computed tomography mic transformation, the collected attenuation projection image is (PCCT) is the method to recover the 3D distribution of the phase- written as: shifting properties of a given sample from a series of 2D projec- tions acquired at different relative orientations of the sample. In PCCT projection, images are acquired through a phase-contrast pa (x,q, y) = òm(x,h, y)dh imaging (PCI) technique that enables transform phase distortion of the x-ray wave introduced by the sample into detectable inten- Starting from a set of projections acquired over an angular sity modulations on the imaging detector. span of (at least) 180°, the 3D distribution of the sample’s linear FIGURE P.40 Sketch of the coordinate system used for the description of the tomographic acquisition and reconstruction. Phase-contrast tomosynthesis 705 Phase-contrast tomosynthesis attenuation coefficient (i.e. µ(x,y,z)) is commonly reconstructed in-focus planes through the imaged object at various depths. The via the filtered back-projection (FBP) algorithm. aim of tomosynthesis is to (partially) overcome structural over- The same formalism can be extended to PCT when the projec- lap, which is inherent in planar imaging, and arguably its most tion image is given by the object-induced phase-shift, which is widespread application is in the field of breast imaging (i.e. digital proportional to the line integral of the real decrement (δ) of the breast tomosynthesis). refractive index As with most x-ray radiological techniques, the image contrast observed in tomosynthesis relies on the attenuation properties on 2p pph(x,q, y) = - òd(x,h, y)dh the investigated sample. Conversely, in addition to attenuation, l the signal recorded through PCTI is related to the phase-shift hence the spatial distribution of δ(x,y,z) can be similarly obtained imparted by the sample to the impinging x-ray wavefield, which is through FBP. ultimately linked to the sample’s refractive properties. Several PCI Of note, with the exception of Bonse–Hart interferometry, techniques have been implemented in the context of tomosynthesis projections acquired through any of PCI the techniques do not imaging, including analyser-based imaging (ABI), grating inter- directly provide the phase-shift, instead they are proportional to ferometry (GI) and propagation-based imaging (PBI). All of these its first or second spatial derivative. For this reason, prior to tomo- techniques are not directly sensitive to the sample-induced phase graphic reconstruction, projections have to be integrated through shift but, instead, the recorded phase-contrast signal is related to adequate phase-retrieval algorithms. Alternatively, in case of the phase shift’s first or second derivative, mainly arising across phase-contrast techniques sensitive to the phase-shift gradient, as sharp interfaces where the phase-shift changes abruptly. For this grating interferometry or edge illumination, the phase integration reason, as in the case of phase-contrast computed tomography, P can be performed within the tomographic reconstruction process projection images should be adequately integrated (or phase- by substituting the usual ramp (Ram–Lak) reconstruction filter retrieved) either prior or within the reconstruction process to with the so-called Hilbert filter, thus significantly speeding up yield a map of the real decrement (δ) of the refractive index (n). the process of 3D reconstruction: this approach is referred to as Alternatively, projections containing differential (e.g. in ABI and refractive tomography. GI) or double differential (in PBI) phase-contrast, can be directly With the advent of PCI techniques (e.g. analyser-based imag- reconstructed producing maps where phase-contrast effects arise ing, grating interferometry, edge illumination) and algorithms across interfaces of details embedded in the sample, thus favour- capable of extracting ultra-small-angle scattering (USAXS) infor- ing the identification of margins and/or spiculations. mation, PCCT has been extended to USAXS tomographic appli- In addition to synchrotron radiation experiments, both GI- cations. In fact, as for attenuation and phase-shift, the USAXS and PBI-based tomosynthesis set-ups have been implemented signal can be expressed as a line integral along the x-ray propa- with compact x-ray sources. In particular, GI operated in the gation direction, thus the tomographic reconstruction formalism Talbot–Lau configuration seems to be the most promising tech- can be easily translated to yield 3D maps of the local scattering nique towards a potential clinical translation of PCTI in the field distribution. Since USAXS is related to the presence of sub- of breast imaging. In fact, the main advantage offered by GI is micrometric features in the sample (e.g. nanoparticles), scattering the possibility of maintaining the design and hardware of conven- maps provide a valuable insight on the sample composition on a tional digital breast tomosynthesis devices, while adding source scale much smaller than the imaging system spatial resolution or grating within the rotary gantry, and integrating phase and analy- reconstructed voxel spacing. ser grating within the detector housing (see Figure P.41). Related Articles: Phase-contrast imaging, Computed tomography Further Readings: Diemoz, P. C. et al. 2017. Non- interferometric techniques for X-ray phase-contrast biomedical imaging. In Russo P. (ed.) Handbook of X-ray Imaging: Physics and Technology, CRC Press, Boca Raton, FL, pp. 999–1023; Pelliccia, D., M. J. Kitchen and K. S. Morgan. 2017. Theory of X-ray phase-contrast imaging. In Russo P. (ed.) Handbook of X-ray Imaging: Physics and Technology, CRC Press, Boca Raton, FL, pp. 971–998; Rigon, L. et al. 2008. Generalized diffraction enhanced imaging: Application to tomography. Eur. J. Radiol. 68(3):S3–S7; Vittoria, F. A. et al. 2017. Multimodal phase-based X-ray microtomography with nonmicrofocal laboratory sources. Phys. Rev. Appl. 8(6):064009. Phase-contrast tomosynthesis (Diagnostic Radiology) Phase-contrast tomosynthesis imaging (PCTI) refers, in general, to any implementation of x-ray tomo- synthesis where the projection images are acquired through a phase-contrast imaging (PCI) technique. X-ray tomosynthesis is a pseudotomographic imaging tech- nique consisting of the sequential acquisition of 2D images (i.e. projections) recorded at different view angles within a limited angular span (usually within an angular range of 60° or lower). FIGURE P.41 Sketch of grating interferometry phase-contrast tomo- The set of projection images is further processed with a limited- synthesis imaging system. G0, G1 and G2 are the source, phase and angle tomographic reconstruction algorithm, yielding a stack of analyser gratings, respectively. Phase dispersion 706 Phase encoding As in the case of phase-contrast tomography, GI PCTI has also Even if all spins start their rotation in the xy plane with the been extended to the reconstruction of ultra-small-angle scatter- same phase angle, their slightly differing angular velocities will, ing (USAXS) maps, potentially providing a valuable insight on over time, cause them to accumulate different phases. the sample composition on a scale much smaller than the imaging In MRI, this is called phase dispersion, and causes signal loss system spatial resolution or reconstructed voxel spacing. as the length of the resulting Mxy becomes shorter. Related Articles: Phase-contrast imaging, Phase-contrast Related Articles: Phase coherence, Phase contrast MRI, tomography, Analyser-based imaging, Propagation-based imag- Shimming, Susceptibility ing, Grating interferometry Further Readings: Li, K. and G. H. Chen. 2017. X-ray phase- Phase encoding contrast tomosynthesis imaging. In Russo P. (ed.) Handbook of (Magnetic Resonance) Phase encoding is one of the principal X-ray Imaging: Physics and Technology, CRC Press, Boca Raton, methods used for spatial localisation in MRI. It is normally used FL, pp. 1049–1062; Li, Ke et al. 2014. Grating‐based phase con- in conjunction with frequency encoding to localise signal within trast tomosynthesis imaging: Proof‐of‐concept experimental a selected slice. studies. Med. Phys. 41(1):011903; Schleede, S. et al. 2014. X-ray Frequency encoding allows position along a chosen axis to be phase-contrast tomosynthesis for improved breast tissue discrim- encoded into the NMR signal and recovered using Fourier trans- ination. Eur. J. Radiol. 83(3):531–536; Sechopoulos, I. 2013. A formation. This yields a projection through the object, but if the review of breast tomosynthesis. Part I. The image acquisition pro- same thing could be done along the perpendicular axis as well, 2D cess. Med. Phys. 40(1):014301. Fourier transformation could be used to recover the entire image. P However, application of two perpendicular frequency encoding Phase dispersion gradients simultaneously would not achieve this: instead, it would (Magnetic Resonance) The phase φ is defined as the angle that result in a single projection along a direction defined by the sum Mxy in the transverse plane describes with one of the axes (x,y), of the two gradients. Phase encoding is a clever means of getting where Mxy is the projection of the total magnetisation M onto around this problem, encoding positional information into signal this plane (Figure P.42). Mxy is the resultant of all spins that are frequency by indirect means. polarised in the xy plane. To perform phase encoding, a gradient is applied for a short time Mxy precesses around the z-axis with the Larmor frequency. during the interval between excitation and signal acquisition. The The RF-excitation that polarises the spins in the xy plane makes gradient is oriented perpendicular to the frequency encoding axis all excited spins start their precession at the same angle φ. If all in the plane of the selected slice. While the gradient is on, elements spins experience exactly the same external magnetic field, they of transverse magnetisation at different locations along the gradi- will have exactly the same frequency and will rotate with the ent direction precess at different frequencies. When the gradient is same angular velocity, corresponding to ω = γB0. switched off the magnetisation returns to a common precessional However, in a realistic environment, the external field is nei- frequency, but the phase dispersion accumulated in the presence ther static nor homogenous. On a microscopic scale, each spin of the gradient remains. This phase is given by the equation ϕ(y) is affected by the time-varying field from its neighbours. On a = γGyyty, where y is the displacement along the gradient direction, macroscopic scale, the external field B0 is modulated by inhomo- Gy is the gradient amplitude (in mT m−1), ty is the gradient duration, geneities that can result either from imperfect shimming or from γ is the gyromagnetic ratio of the nucleus (normally 1H) and B0 is susceptibility effects within the body. Within each voxel, spins the strength of the static magnetic field. Frequency encoding is then in different subvolumes may therefore experience different effec- applied along the perpendicular direction as normal. The acquired tive magnetic fields. Each subvolume will thus contain spins that echo contains positional information along the phase encoding gra- rotate with an angular velocity ω corresponding to the sum of the dient axis, encoded in form of phase, as well as along the frequency magnetic fields that it experiences, ω = γB0 + ΔB0. encoding axis, encoded in the form of frequency. This entire data acquisition process is repeated a number of times, corresponding to the desired image resolution (typically 256, 512 or, increasingly, 1024). On each repetition a different B Z 0 phase encoding gradient amplitude is used, so that the phase of M the signal from magnetisation at a given position along the phase encoding gradient direction is incremented. The phase evolution across the set of acquired echoes mimics the effect of frequency encoding: if the echoes are lined up next to each other in order of phase encoding gradient amplitude, they form a ‘pseudoecho’ along the phase encoding axis that resembles the echo that forms during frequency encoding, and this ‘pseudoecho’ appears to be composed of signals with different frequencies originating from different distances along the phase encoding axis. (Both positive y and negative phase encoding gradient values are used, mimick- ω0 + Δω ing the rephasing and dephasing of the frequency encoded echo.) ω Two-dimensional Fourier transformation can thus be applied to 0 ω0 – Δω Mxy recover the 2D image from the 2D phase and frequency encoded dataset (Figure P.43). x Figure P.44 shows the symbol used to represent the phase encoding gradient in a conventional pulse sequence diagram. The FIGURE P.42 Dephasing of transverse magnetisation, caused by locally ladder-like structure is intended to

convey the idea of repetition varying magnetic fields. with different gradient amplitudes. Phase image 707 Phased array coil sin(ωt) ‘ωhigh’ Time ‘ωlow’ FIGURE P.43 Phase encoding simulates frequency encoding. cos(ωt) FIGURE P.45 Sinusoids in phase quadrature. is in quadrature to the sinusoidal signal (Figure P.45): ss(t) = A0 sin wt P FIGURE P.44 Diagrammatic representation of phase encoding gradient. Phase quadrature (Ultrasound) In general, phase quadrature means 90° out of phase. Alternatively we can say that when two signals differ in Related Articles: B0 gradients, Frequency encoding, Multi- phase by −90° or +90°, they are said to be in phase quadrature. slice, Slice selection, Image reconstruction, Pseudo echo An important quantity of two signals in phase quadrature is that Further Reading: McRobbie, D. W. et al. 2007. MRI from the expected value of their product is zero, which can be useful Picture to Proton, Cambridge University Press, Cambridge, UK. to check whether two signals are in phase quadrature. Signals in phase quadrature are used in quadrature detection for separat- Phase image ing forward and reverse flow components in Doppler systems for (Magnetic Resonance) A phase image is composed of the phase example. values, transformed into a greyscale, of the acquired MR signal. In conventional MRI, magnitude images are used. These Phased array coil images represent the signal amplitude of the transverse magneti- (Magnetic Resonance) A phased array coil is an MR receiver coil sation. However, the signal also has a phase value, which can be consisting of an array of individual receiver coils. A phased array used for imaging as well. These maps are called phase images. yields the high signal to noise ratios seen with small surface coils In an ideal MR experiment, the spins are encoded to have a while simultaneously providing a large field of view. specific phase value depending on their position in space. But In a phased array adjacent coils are overlapped to eliminate because of, for example field inhomogeneities and motion, the mutual inductance. Mutual inductance is a measure of coupling phase value can be different. As the phase value is used in the between coils. This coupling would cause an unwanted split- Fourier transform for positioning data in the Fourier domain, an ting of the tuned resonant peaks of the coils and a reduction of erroneous phase value can cause image artefacts like geometric coil sensitivity. The degree of overlap required to set the mutual distortions. inductance to zero is determined by the geometry of the individ- Phase images are used in various MR experiments. For ual coils. example, in phase-contrast MR, phase images are used to depict Each coil in the phased array is connected to a low imped- moving spins. Phase images can also be used in mapping field ance pre amplifier. Use of low impedance amplifiers helps reduce inhomogeneities or in MR-venography or SWI. coupling between distant coils. The outputs of the preamplifiers Related Articles: Diffusion, Phase contrast, Phase-mapping, in the array are sampled simultaneously and combined electroni- Velocity mapping cally (Figure P.46). Further Reading: Dixon, R. L. and K. E. Ekstrand. 1982. The physics of proton NMR, Med. Phys. 9(6):807–818. Phase mapping (Magnetic Resonance) A term identifying the technique also known as velocity mapping (see Velocity mapping). 0.75 Related Article: Velocity mapping diameter Phase quadrature (Magnetic Resonance) Two signals are said to be in phase quadra- Preamplifiers ture if they are +90° or −90° out of phase. The cosinusoidal signal FIGURE P.46 Phased array coil showing overlapping of circular coils sc(t) = A0 cos wt to eliminate mutual inductance. Phased array transducer 708 Phosphor A single receiver coil may then consist of many individual OH CH3 O coil ‘elements’. Parallel imaging techniques depend on the use of HO P N C N CH2 C OH multi-element array coils. Differences in spatial sensitivities of each element are exploited in parallel imaging to deliver scan time O NH2 savings (see Parallel imaging). Further Reading: Roemer, P. B., Edelstein, W. A., Hayes, C. FIGURE P.49 Molecular structure of PCr. E., Souza, S. P. and Mueller, O. M. 1990. The NMR phased array. Magn. Reson. Med. 16:192–225. Phosphocreatine (Magnetic Resonance) Phosphocreatine is a chemical compound Phased array transducer that features in in vivo phosphorus (31P) NMR spectra of a number (Ultrasound) The phased array transducer produces a sector- of organs. In muscle and brain spectra, the PCr resonance is usu- shaped scan by simultaneously firing all of its piezoelectric ally the most prominent peak in the spectrum, and may be used as transducer elements (typically 128 or 256) for each scan line an internal chemical shift reference (Figure P.49). (Figure P.47). This is contrary to the linear array, which only uses The importance of phosphocreatine in 31P NMR arises from parts of its total elements per line. The spacing between the ele- its role in the body’s energy metabolism. PCr acts as a source of ments is usually below λ/2, where λ denotes the centre frequency, ATP, and is depleted to maintain ATP levels during ischaemia to reduce grating lobe effects. and hypoxia. Thus PCr levels can be used to monitor the effect The other difference compared to linear arrays is that to pro- P of fatiguing exercise in skeletal muscle (Figure P.50). It also pro- duce the sector-shaped image the ultrasound beam is steered in vides a prognostic indicator in birth asphyxia that is well corre- all directions within the sector. This beam steering is accom- lated with neurodevelopmental outcome (Figure P.51). plished by introducing individual time delays to the transducer elements in a similar way as with electronic focusing in a linear Phosphodiesters (PDE) array (Figure P.48). (Magnetic Resonance) Phosphodiesters are chemical compounds In normal use the operator will angle the transducer in such in which a phosphate group is joined to two other molecules via a way that the field of interest is in the straight-ahead direction. ester bonds. These compounds are found in vivo, and feature in This is because the beam width at the focal point will widen with phosphorus (31P) NMR spectra of a number of organs. the steering angle, resulting in a worse lateral resolution in the The PDE region of an in vivo 31P spectrum occupies the outer parts of the scan sector. This can be understood by the fact chemical shift range around 3 ppm (relative to phosphocreatine). that the effective aperture (i.e. the aperture seen from the point of It consists of a broad peak due primarily to glycerol 3-phosphor- view of the propagating beam) of the transducer gets smaller with ylethanolamine and glycerol 3-phosphorylcholine. The individual the steering angle (Figure P.48). compounds cannot normally be resolved (Figure P.52). The phased array transducer is generally used for cardiac Biochemically, these compounds are catabolites of phosphati- imaging as the small aperture fits in between the ribs, and the dylethanolamine and phosphatidylcholine, which are major com- large sector-shaped scanning area covers the heart. ponents of membrane phospholipids and of myelin in the brain. Related Articles: Linear array, Grating lobes Increases in PDE levels are often observed in malignant tumours, due to increased membrane metabolism (Figure P.53). Phosphomonoesters (PME) (Magnetic Resonance) Phosphomonoesters are chemical com- pounds in which a phosphate group is joined to another molecule via ester bonds. These compounds are found in vivo, and feature in phosphorus (31P) NMR spectra of a number of organs. The PME region of an in vivo 31P spectrum occupies the chemical shift range between about 5 and 7.5 ppm (relative to phosphocreatine). It consists of a broad peak due primarily to phos- Linear array Curved linear array Phased array phorylcholine, phosphorylethanolamine and sugar phosphates. The individual compounds cannot normally be resolved (Figure P.54). FIGURE P.47 Sector-shaped scanning area from a phased array trans- Biochemically, these compounds are precursors of phospha- ducer compared to the scanning area from the linear and curved linear tidylethanolamine and phosphatidylcholine, which are major arrays. (Courtesy of EMIT project, www .emerald2 .eu) components of membrane phospholipids and of myelin in the brain. Increases in PME levels are usually observed in malignant tumours, due to cell proliferation and associated increased mem- brane metabolism (Figure P.55). Phosphor (General) Molar mass 20–300 g mol−1 Density at STP 2000–6000 kg m−3 Melting point 500–3000 K FIGURE P.48 Time delays for beam steering combined with focusing. Boiling point 1000–5000 K (Courtesy of EMIT project, www .emerald2 .eu) Transmitters Time delay Timer Phosphor 709 Phosphor Phosphocreatine P 25.000 20.000 15.000 10.000 5.0000 .00000 –5.0000 –10.000 –15.000 –20.000 –25.000 (ppn) FIGURE P.50 31P NMR spectrum of the human brain showing PCr resonance. A phosphor is a substance that exhibits phosphorescence, the prolonged emission of light following exposure to radiation. Phosphorescence is a type of photoluminescence, similar to fluo- rescence, which immediately re-emits the absorbed radiation. The study of phosphorescence contributed to the discovery of radioactivity in 1896. Recovering Phosphors are various transition metal or rare earth com- pounds. Phosphors are usually made by adding an ‘activator’ to a ‘host material’. The host materials are often oxides, sulphides or halides of zinc, manganese or aluminium. The activators sustain the emission time or ‘afterglow’. Other materials, such as nickel, may be used to shorten the afterglow. In general, the persistence of the afterglow increases as the wavelength of the emitted light increases. Common phosphors include copper-activated and sil- ver-activated zinc sulphide, which emit green light. PC Fatiguing Phosphors are commonly used to produce ‘glow-in-the-dark’ objects, cathode ray tube displays and fluorescent lights. They are added to paints and cosmetic creams for ‘glow-in-the-dark’ face ATP paints. Phosphors, such as zinc sulphide, were once used to paint γ α β dials of watches by mixing them with radioactive substances, P1 such as radium, to excite the phosphor. Materials which exhibit Resting electroluminescence produce light sources from a large area by excitation with an electric field. These substances are suitable for FIGURE P.51 Depletion and recovery of PCr in skeletal muscle during backlights, such as those used in a liquid crystal display (LCD). and after fatiguing exercise. Medical Applications: One of the most common phosphors is sodium iodide doped with thallium, NaI(Tl), used in scintillation detectors due to its high light output. NaI(Tl) crystals are often coupled with photomultiplier tubes, used in gamma cameras for O CH3 + nuclear medicine. Control of crystal growth is used to adjust its HO CH2 CH CH2 O P O CH2 CH2 N CH3 parameters, such as radiation hardness, afterglow and transpar- OH O– CH3 ency. NaI(Tl) is also used in x-ray detectors. Caesium iodide (CsI) is regularly used as the input phosphor of FIGURE P.52 Molecular structure of glycerol 3-phosphorylcholine, a x-ray image intensifiers for fluoroscopy. ZnS is used as a scintil- constituent of the PDE peak. lation detector for x-ray screens and cathode ray tubes. Lithium Phosphor layer 710 P hosphorescence Phosphodiesters (PDE) P 25.000 20.000 15.000 10.000 5.0000 .00000 –5.0000 –10.000 –15.000 –20.000 –25.000 (ppm) FIGURE P.53 31P NMR spectrum of the human brain showing PDE region. O CH3 Phosphorescence + (Nuclear Medicine) Phosphorescence is a slow process that briefly HO P O CH2 CH2 N CH3 prevails beyond the exciting source being removed (afterglow), as opposed to fluorescence which is a prompt (fast) process that O– CH3 will generally not exist beyond the exciting source being removed. Phosphorescence occurs when electrons are trapped in an FIGURE P.54 Molecular structure of phosphorylcholine, a constituent impurity-induced energy state where further de-excitation is of the PME peak. forbidden. The electrons will eventually de-excite to the ground state after being excited to an adjacent energy state. This delay in de-excitation will cause some single-photon registrations after fluoride (LiF), known as TLD-100, is used to record gamma and the original pulse has died out. This effect is also referred to as neutron exposure in thermoluminescent dosimeters. Many other afterglow. TL dosimeters are also known such as calcium sulphate (CaSO4), The scintillation property of a material is dependent on the calcium fluoride (CaF2), lithium borate (Li2B4O7), and aluminium energy states determined by the lattice structure. The electrons oxide (Al2O3). in such materials are only allowed in certain discrete bands, the Related Articles: Gamma ray, Gamma camera, Image inten- upper two bands are called valence band (the lower of the two) sifier, Fluoroscopy, Neutron, Nuclear medicine, Radioactivity, and conduction band (the upper of the two). In a pure crystal these Radium,

Thermoluminescent dosimeter (TLD), X-ray two bands are separated by a gap of ‘forbidden’ energies. The emitting of scintillation light is a part of a two-step process: (1) Phosphor layer incident radiation excites electrons in the lower valence band up (Diagnostic Radiology) In conventional diagnostic x-ray using to the conduction band, (2) electrons release energy when they are film, a fluorescent intensifying screen is used. The screen con- de-excited, thus sending out a scintillation photon. Scintillation tains phosphors which fluoresce on exposure to x-ray radiation. photons are not in an optimal energy range in terms of photo- The visible light emitted from the phosphors blackens the film multiplier tube (PM tube) efficiency. In many inorganic scintil- and a latent image is formed. The phosphor layer is the layer of lators, like NaI(Tl), an impurity is induced to create multiple the screen containing phosphor particles suspended in a support- energy states in the band gap. When electrons de-excite from one ing substrate. The thickness of the phosphor layer is in the range of the impurity-induced energy states to another, photons with the 100–300 μm. Thinner phosphor layers will support higher resolu- desired energy are emitted. tion as there is less dispersion of light in the phosphor thickness. The electrons do not all follow the same de-excitation pattern, A thicker phosphor provides a higher speed, as more x-ray inter- and some electrons will be caught in energy states where further actions will occur in the thicker material. de-excitation is forbidden. Further de-excitation is only possible Phosphorus-32 [32P] 711 Phosphorus-32 [32P] Phosphomonoesters (PME) P 25.000 20.000 15.000 10.000 5.0000 .00000 –5.0000 –10.000 –15.000 –20.000 –25.000 (ppm) FIGURE P.55 31P NMR spectrum of the human brain showing PME region. if the electron is first excited (thermal excitation) to an adjacent Melting point 317.3 K energy state from which the electron is allowed to de-excite. Such Boiling point 550 K a process will cause a delay in excitation, thus creating a slow Density near room temperature 1823 kg/m3 component, i.e. phosphorescence. In NaI(Tl) some of the com- ponents have a decay time of 0.15 s and contribute some 9% of the total light yield. The main scintillation signal component in Phosphorus is highly reactive element that was discovered in 1669 NaI(Tl) has a decay time of 230 ns. Since the time resolution of the by Hennig Brand and has several allotropes; white, red and black. PM tube is below 0.15 s, delayed events can easily be registered Phosphorous is an essential component of living systems, since and discriminated. However for certain applications phosphores- it is a major component of adenosine triphosphate (ATP), a mol- cence can cause problems, namely high count rate applications. ecule used to store the energy produced as cells respire. It is also In such applications, delayed pulses will overlap which prevents present in bones and nervous tissue. energy discrimination and therefore the delayed signals will cre- Medical Applications: In magnetic resonance imaging, phos- ate a background signal which ultimately leads to worse activity phorus spectroscopy is performed to investigate muscle metabo- quantification and image quality. lism. Spectra show changes during and after exercise, which result Related Articles: NaI(Tl) detector crystal, Scintillators, from a change in the amount of ATP in the muscle. Phosphorus Inorganic scintillators, Light yield in scintillation detectors, spectroscopy is also used to investigate changes in liver metabo- Bismuth germanate (BGO) lism as a result of hepatitis and cirrhosis. Further Reading: Knoll, G. F. 2000. Radiation Detection and 32P is a radioactive isotope of phosphorus used in the form of Measurement, 3rd edn., John Wiley & Sons, Inc., New York, pp. 32P-BioSilicon to treat advanced pancreatic cancers, since it is a 231–234. beta-emitter with a half-life of 14.3 days. Related Articles: Beta particles, Half-life, Magnetic reso- Phosphorus nance imaging, Phosphorus-32, Phosphorus-33, Radioactive (General) Phosphorus-32 [32P] (Nuclear Medicine) Symbol P Element category Non-metals Element: Phosphorus Mass number A 31 Isotopes: 11 < N < 45 Atomic number Z 15 Atomic number (Z): 15 Atomic weight 30.974 g/mol Neutron number (N): 17 Electronic configuration 1s2 2s2 2p6 3s2 3p3 Symbol: 32P Phosphorus-33 [33P] 712 P hoto peak - Production: Reactor, e.g. 31P(n,g)32 P ®b 32S Total absorption (electron range): 0.5–1 mm lucite 14,6d Daughter: 32S Biological half-life: Bone >3 years, WB 257 days Half-life: 14.3 days Critical organ: Bone surfaces, red bone marrow Decay mode: β− - decay ALImin (50 mSv): 3000 MBq Radiation: β− 1710.7 keV (max) 570 keV (mean) Absorbed dose: 2.0 mSv MBq−1 to bone surfaces, 1.5 ms Gamma energy: None MBq−1 to red bone marrow Effective dose: 0.33 mSv MBq−1 (oral) 0.1 mSv MBq−1 Skin dose rate from 1 MBq: 120 μSv h−1 at 30 cm (point source); 0.0054 μSv h−1 at 1 m (10 mL glass vial) (inhalation) Total absorption (electron range): 6–10 mm lucite Biological half-life: bone >3 years, WB 257 days 15 4S°3/2 Critical organ: red bone marrow, bone surfaces 1/2+ 25.34 d ALImin (50 mSv): 10 MBq P 32 Absorbed dose: 0.88 mGy MBq−1 red bone marrow Phosphorus P 15 β– Effective dose: 2.3 mSv MBq−1 (oral) 0.8–3.2 mSv MBq−1 3/2+ 30.973761 (inhalation) [Ne]3s23p3 33 16S 10.4867 Qβ–248.5 15 4S° P 3/2 1+ 14.262d Clinical Applications: 33P may be used as an alternative to P 32P biomedical research as a tracer substitute for phosphor in, e.g. 32 Phosphorus P 15 β– nucleic acid sequences (DNA-hybridism). 30.973761 Related Article: P-32-Sodium orthophosphate Further Readings: Annals of the ICRP. 1987. Radiation Dose [Ne]3s23p3 to Patients from Radiopharmaceuticals, Biokinetic models and 10.4867 Qβ–1710.66 Data, ICRP Publication 53, Vol. 18. Pergamon Press, Oxford, UK; Chu, S. Y. F., L. P. Ekström and R. B. Firestone. 1999. The Lund/LBNL Nuclear Data Search. [http: / /nuc leard ata .n uclea r Clinical Applications: In nuclear medicine 32P is mainly .lu. se /nu clear data/ toi/]; Firestone, R. B. 1999. Table of Isotopes, used for radionuclide therapy of polycythemia vera (a rare blood 8th edn., Update with CD-ROM. [http://ie .lbl .gov /toi .html]; disease characterised by an elevation of the immature red blood Kowalsky R.J. and S. W. Falen. 2004. Radiopharmaceuticals in cells) and other neoplastic diseases, for example treatment of Nuclear Pharmacy and Nuclear Medicine, 2nd edn., American peritoneal or pleural effusions. It is frequently used in biomedical Pharmacists Association, Washington, DC. research as a tracer substitute for phosphor in, e.g. nucleic acid sequences (DNA-hybridism). Photic stimulation Related Article: Phosphorus-32 (Magnetic Resonance) During a fMRI study, the subject under- Further Readings: Annals of the ICRP, 1987. Radiation Dose takes a series of tasks known as a paradigm. One of the original to Patients from Radiopharmaceuticals, Biokinetic models and proof-of-concept fMRI paradigms was that of photic stimulation Data, ICRP Publication 53, Vol. 18. Pergamon Press, Oxford, UK; (Kwong et al., 1992). Chu, S. Y. F., L. P. Ekström and R. B. Firestone. 1999. The Lund/ In the simplest type of photic stimulation paradigms, the visual LBNL Nuclear Data Search. [http: / /nuc leard ata .n uclea r .lu. se /nu stimulus is switched on and off alternately (in a block design, per- clear data/ toi/] (accessed on 9 July 2012); Firestone, R. B. 1999. haps with 30 s blocks). In the brain areas which are responsive to Table of Isotopes, 8th edn., Update with CD-ROM. [http://ie .lbl the photic stimulation (such as the visual cortex), the fMRI signal .gov /toi .html] (accessed on 9 July 2012); Kowalsky R.J. and S. W. rises and falls as the stimulus is turned on and off (with a slight Falen. 2004. Radiopharmaceuticals in Nuclear Pharmacy and delay due to the delay in the blood flow response or haemody- Nuclear Medicine, 2nd edn., American Pharmacists Association, namic response function). Washington, DC. During early implementations of photic stimulation the sub- ject wore a mask which covered their eyes and blocked out all Phosphorus-33 [33P] ambient light. Mounted within the mask were a series of LEDs (Nuclear Medicine) which flickered with adjustable frequency/intensity (Figure P.56). Related Articles: fMRI (functional magnetic resonance imag- Element: Phosphorus ing), Block design, Haemodynamic response function Isotopes: 11 < N < 45 Further Reading: Kwong, K. et al. 1992. Dynamic magnetic Atomic number (Z): 15 resonance imaging of human brain activity during primary sen- Neutron number (N): 18 sory stimulation. Proc. Natl. Acad. Sci. USA 89:5675–5679. Symbol: 33P Production: Reactor, e.g. Photo peak Daughter: 33S (Radiation Protection) The process where a photon (x- or γ-ray) Half-life: 25.3 days is absorbed and an electron is ejected from the atom is called the Decay mode: β− - decay photoelectric effect. The energy of the photon Eph should be equal Radiation: β− 248.5 keV (max) 76.4 keV (mean) to or greater than the electron binding energy Eb. The energy Ee of Gamma energy: None the ejected electron (photoelectron) is thus equal to Skin dose rate from 1 MBq: 0 μSv h−1 at 30 cm (point source) Ee = Eph - Eb Photoablative effects 713 Photobiological lamp safety Photic stimulation–IR images Photoablative effects (Non-Ionising Radiation) Photoablative effects are observed 2860 off on off on when using lasers. For photoablative effects to occur, the energy of the laser beam must be great enough to break molecular bonds within the target material. Following the breaking of bonds, the 2800 atoms dissociate or vaporise with no damage to adjacent material. Photoablation may be used to precisely remove tissue or bone, 2740 depending on the absorption of the laser wavelength in the mate- rial to be ablated. Due to the precision of the laser ablation pro- cess, this effect is applied in laser eye surgery. 2680 Related Articles: Photothermal effects, Photochemical 0 70 140 210 280 (s) effects, Photomechanical effects Further Readings: Bertolotti, M. 1983. Masers and Lasers, Photic stimulation–GE images An Historical Approach, Adam Hilger Ltd, Bristol, ISBN 6050 off on off on 0-85274-536-2; Brown, R. 1968. Lasers, Tools of Modern Technology, Aldus Books, London; Carruth, J. A. S. and A. L. McKenzie. 1986. Medical Lasers Science and Clinical Practice, 5900 Adam Hilger Ltd; Henderson, A. Roy. 1997. A Guide to Laser Safety, Chapman & Hall. P 5750 Photo-activation therapy (Radiotherapy) Photo-activation therapy (PAT) is a novel thera- 5600 peutic technique that aims to improve the effectiveness of targeted 0 60 120 180 240 radiotherapy treatment. Increased tumour cell kill is achieved by (s) the introduction of a high-Z atom into the DNA of the tumour cell followed by irradiation with x-rays whose energy is chosen to FIGURE P.56 Signal-intensity changes for a region of interest (≈60 enhance the photoelectric effect on the particular atom. mm2) within the visual cortex during darkness and during 8-Hz photic It is thought that this technique may find application in the stimulation. Results using both inversion recovery (IR) and gradient echo treatment of brain tumours such as gliomas where conventional (GE) techniques are shown. Brain signal change for this particular subject is ≈3% for both IR and GE acquisitions. (Image taken from Kwong, K. et radiotherapy treatment can at best be described as palliative. In al., Proc. Natl. Acad. Sci. USA, 89, 5675, 1992.) these cases, although radiotherapy enhances tumour control, the radiosensitivity of the surrounding healthy tissues does not per- mit the delivery of a sufficiently high dose required to treat such radio-resistant tumours. In vitro experiments on a rat glioma cell line have demonstrated that irradiation of cells treated with cis- platin, a widely used chemotherapy drug, with x-rays of energy Photopeak corresponding to the platinum absorption K-edge increases local Noise Compton toxicity as would be expected if there was an enhanced photoelec- scattering tric effect on the platinum atoms. FWHM Abbreviation: PAT = Photo-activation therapy. Related Articles: Radiosensitivity, Therapeutic effect Further Reading: Biston, M., A. Joubert, J. Adam. et al. 2004. Cure of fisher rats bearing radioresistant F98 Glioma treated with cis-platinum and irradiated with monochromatic synchrotron Height of pulse x-rays. Cancer Res. 64:2317–2323. FIGURE P.57 Example of gamma radiation spectrum registered with Photobiological lamp safety NaI(Tl). (Non-Ionising Radiation) Under IEC 62471:2006, photobiologi- cal safety of lamps and lamp systems, non-coherent sources of artificial optical radiation, including IPLs, are classified into The photo peak in the measured energy spectrum (Figure four risk groups, see Table P.4. These classifications indicate the P.57) represents those events where the energy of the photon potential risk of adverse health effects and allow the user to apply is fully absorbed and thus indicates the energy of

the incident appropriate control measures to minimise the risks. photons. Under the Control of Artificial Optical Radiation at Work Detectors (e.g. scintillation detectors, proportional counters, Regulations 2010, an in-depth risk assessment must be carried out or germanium detectors) that have output pulse amplitudes depen- for Risk Group 3 equipment such as IPLs. dent on the energy of the incident radiation are used for x- and Related Articles: Hazard value, IPL γ-radiation energy measurement. Further Readings: A Non-Binding Guide to the Artificial Related Articles: Germanium detectors, Photoelectric effect, Optical Radiation Directive 2006/25/EC, Radiation Protection Proportional counter, Scintillation detector Division, Health Protection Agency; Medicines and Healthcare Further Reading: Knoll, G. F. 2000. Radiation Detection Products Regulatory Agency. 2015. Lasers, intense light source and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, systems and LEDs – guidance for safe use in medical, surgical, p. 428. dental and aesthetic practices, Crown copyright.; International Number of pulses Signal intensity Signal intensity Photocathode 714 Photoelectric effect TABLE P.4 Classification of Non-coherent Artificial Optical Radiation Exempt No photobiological hazard under foreseeable conditions. e.g. domestic and office lighting, computer monitors Risk group 1 Low-risk group, the risk is limited by normal behavioural limitations e.g. household torch, desk lamp, photocopiers on exposure Risk group 2 Moderate-risk group, the risk is limited by aversion response, e.g. desktop projectors, vehicle headlights, entertainment although these responses are not universal floodlights, UV insect traps Risk group 3 High-risk group, may pose a risk even for momentary or brief e.g. IPLs, UV sterilisation lamps, welding lamps, UV exposure – safety control measures are essential curing devices Electrotechnical Commission: IEC 62471:2006 Photobiological treat medical conditions, for example, the diagnosis of skin safety of lamps and lamp systems. IEC 2006. conditions by identifying bacteria that fluoresce, applying blue light to oxidise bilirubin and treat jaundice and applying vis- Photocathode ible light to activate cancer treatment drugs in photodynamic P (Nuclear Medicine) The cathode that transforms incident scintil- therapy. lation photons to photoelectrons in a photomultiplier tube (PM Exposure to optical radiation may also cause damage by pho- tube). The transfer efficiency or sensitivity of the cathode is tochemical effects, for example, sunburn. This damage may be referred to as the photocathode quantum efficiency. cumulative, for example, the formation of cataracts from UV The transfer from scintillation light to photoelectrons is a exposure over time or skin ageing. three-step process: first, the scintillation photon is absorbed and Related Articles: Photothermal effects, Photomechanical releases an electron in the process, the electron migrates to the effects, Photoablative effects cathode surface where it has to have enough remaining kinetic Further Readings: Bertolotti, M. 1983. Masers and Lasers, energy to escape the potential barrier. An Historical Approach, Adam Hilger Ltd, Bristol; Brown, R. The photocathode sensitivity or quantum efficiency (QE) is 1968. Lasers, Tools of Modern Technology. Aldus Books, London; defined as Carruth, J. A. S. and A. L. McKenzie. 1986. Medical Lasers Science and Clinical Practice, Adam Hilger Ltd; Henderson, A. Number of photoelectrons emitted Roy. 1997. A Guide to Laser Safety, Chapman & Hall. QE = Number of incident photons The maximum QE registered in most photocathodes is typically Photoconductor 20%–30%. (Nuclear Medicine) A photoconductor is a device whose con- ductivity increases in response to electromagnetic radiation. In nuclear medicine, the photocathode of a photomultiplier (PM) Photocathode of photomultiplier tubes tube is such a device. The photocathode emits electrons when (Nuclear Medicine) A material or substance that releases elec- struck by photons of visible light. trons when it is irradiated by light photons (∼400 nm). The most Related Article: PM Tube common use of photocathodes in nuclear medicine is in photo- multiplier tubes (PM tubes). In the PM tube construction the pho- tocathode follows directly after the glass entrance window. The Photoelectric absorption back surface of the entrance window is coated with a photoemis- (Radiation Protection) See Photoelectric effect sive substance. The substance typically consists of CsSb (caesium antimony) and other bi-alkali compounds. The electrons emitted Photoelectric effect from the photocathode are called photoelectrons. The number (General) The photoelectric effect describes the process in which of photoelectrons produced per incident photon is referred to in an incident photon interacts with a bound atomic electron. If the terms of the quantum efficiency (QE). Typically the QE is 1–3 photon has sufficient energy to overcome the binding energy of photoelectrons per 10 visible light photons. Quantum efficiency the electron, it is displaced from its orbit. The photon is com- depends on the wavelength of the incident photons and it peaks pletely absorbed. Any energy in excess of the electron binding at ∼400 nm. energy is transferred to kinetic energy of the ejected orbital elec- Related Articles: Photomultiplier (PM) tube, Dynode tron. This results in a vacancy in the electron shell. Other elec- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. trons cascading down from outer orbital shells fill the vacancy Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, left by this electron. During the cascade process, characteristic Philadelphia, PA, pp. 101–102. x-rays are emitted. The energy of these x-ray photons may also be transferred to an atomic electron, which may be ejected as an Photochemical effects Auger electron. (Non-Ionising Radiation) The body relies upon photochemical The photoelectric effect dominates interaction processes at interactions to carry out several of its important natural functions, low photon energies (up to ∼50 keV in water or human tissues). these include vision, formation of vitamin D and melanin produc- The interaction cross-section for the photoelectric effect varies tion in tanning, for example. with Z3 and 1/E3, where Z is the atomic number of the medium and The chemical changes and interactions caused by the E is the energy of the incident photon. action of optical radiation may also be used to diagnose and Related Articles: Auger electron, Characteristic x-rays Photoelectric interaction 715 P hotomultiplier (PM) tubes Photoelectric interaction 180 nm and 400 nm (incoherent optical radiation). Health Phys. (Radiation Protection) See Photoelectric effect 87(2):171–186; ICNIRP. 2000. Revision of the guidelines on lim- its of exposure to laser radiation of wavelengths between 400 nm Photoelectric relay and 1.4 µm. Health Phys. 79(4):431–440; Sihota, Ramanjit and (General) An electric relay which can be controlled using light. A Radhika Tandon. 2011. Parsons' Diseases of the Eye. Elsevier range of devices exists incorporating many forms of light sensi- India. tive detectors to trigger the switching of electrical contacts when a certain level of light is detected. Photomechanical effects Photoelectric relays now normally use semiconductor light (Non-Ionising Radiation) Photomechanical effects are caused sensors with inbuilt amplifiers and comparators to drive either by short (<10 µs), high-power exposure to optical radiation. This electromechanical relays or semiconductor switches. type of exposure may result in two outcomes: They can commonly be found in situations where switching needs to be performed in high voltage environments (x-ray gener- a. The fast, local rate of change of temperature causes ators) or high insulation conditions (for patient safety), as the light a high-pressure gradient and associated mechanical may be provided from an electronic circuit via LED or incandes- waves or ‘shock waves’, which cause damage to cells. cent bulb, and fed to the photoelectric relay through air or other b. High power electric fields are induced, causing ionisa- insulating medium such as fibre-optic cable. tion and leading to the production of localised plasma. Photoelectric relays may also be used directly where it is This plasma expands, forming a shock wave, which necessary to provide an alarm or automatic action when an area exerts mechanical forces on surrounding tissue, causing P becomes too dark or light for safe working. ablation and fracturing (photodisruption). Related Article: Relay Related Articles: Photothermal effects, Photochemical Photoelectron effects, Photoablative effects (General) An electron ejected from atomic orbit due to interac- Further Readings: Bertolotti, M. 1983. Masers and Lasers, tion with a photon or other incident ionising radiation. The energy An Historical Approach, Adam Hilger Ltd, Bristol, ISBN of the incident photon is totally absorbed and transferred to the 0-85274-536-2; Brown, R. 1968. Lasers, Tools of Modern electron. Technology, Aldus Books, London; Carruth, J. A. S. and A. L. In a photomultiplier (PM) tube photoelectrons refer to the McKenzie. 1986. Medical Lasers Science and Clinical Practice, electrons generated by incident scintillation light in the photo- Adam Hilger Ltd; Henderson, A. Roy. 1997. A Guide to Laser cathode. The photocathode is typically situated after the scintil- Safety, Chapman & Hall. lation crystal in a PM tube. The photoelectrons (i.e. the signal) are accelerated and focused on a dynode where the incident pho- toelectron produces several secondary photoelectrons which are Photomultiplier (PM) tubes then focused onto another dynode. This process is repeated until (Nuclear Medicine) A signal amplifier used in nuclear medicine the signal is considered strong enough to measure. to magnify the signal from the scintillator. When ionising radia- The number of photoelectrons produced per energy of inci- tion interacts within a scintillator a number of light photons are dent scintillation photons is called the photocathode quantum created. When coupling the scintillator crystal to a photomulti- efficiency. The maximum quantum efficiency of a photocathode plier tube (from here on abbreviated as PM tube(s)) the signal is is typically 20%–30%. magnified up to 108 times depending on PM tube design. Related Articles: Photoelectric effect, Photocathode Using a NaI (Tl) scintillator an average of 38 light photons is created for each keV deposited in the scintillator. If these photons were to be directly transferred to an electric signal (i.e. using a Photographic detail photocathode) the signal would be almost undetectable among all (Diagnostic Radiology) See Detail resolution background noise. PM tubes can produce a detectable electrical signal from even such a weak light signal. Photokeratitis (Eye) The electric design of a PM tube is displayed in Figure P.58. (Non-Ionising Radiation) The term is used to indicate a painful Low energy photons from interactions in the crystal are led condition of the eye caused by excessive exposure of the cornea to the entrance window. The entrance window is coated with and conjunctiva of the eye to artificial or solar ultraviolet light. photo-emissive substance. This substance is called the photo- Related Articles: AORD, Eye, Retina, Blue light hazard, cathode and it will eject electrons when radiated by low energy Thermal light hazard photons. The efficiency (number of electrons per incident photon) Further Readings: Coleman, A., F. Fedele, M. Khazova, is referred to as quantum efficiency. The electrons emitted from P. Freeman and R. Sarkany. 2010. A survey of the optical haz- the photo-emissive substance are referred to as photoelectrons. ards associated with hospital light sources with reference to the The quantum efficiency is typically 0.1–0.3 photoelectrons per Control of Artificial Optical Radiation at Work Regulations 2010. photon. The quantum efficiency is dependent on wavelength of J. Radiol. Prot. 30(3):469; ICNIRP. A closer look at the thresholds the incident photons and for most photocathodes it peaks at ∼400 of thermal damage: workshop report by an ICNIRP task group. nm which is visible blue light. Health Phys. 111(3):300–306; ICNIRP. 2013. Guidelines on limits The photoelectrons are released into a long tube where they of exposure to incoherent visible and infrared radiation. Health are accelerated by a potential difference between the photo- Phys. 105(1):74–91; ICNIRP. 2013. Guidelines on limits of expo- cathode and a metal plate called a dynode. The photoelectrons sure to laser radiation of wavelengths between 180 nm and 1,000 are directed by a focusing grid so that they strike the dynode. µm. Health Phys. 105(3):271–295; ICNIRP. 2004. Guidelines on When the dynode is hit by the photoelectron a number of new limits of exposure to ultraviolet radiation of wavelengths between electrons are released, these are called secondary electrons. The Photoretinitis (Eye) 716 Photothermal effects Detector Amplifier Photocathode Dynodes PHA Anode Scaler P FIGURE P.58 PM tube connected to amplifier, pulse height analyser and scaler as part of a scintillation detector. secondary electrons are then accelerated by a potential difference between the first dynode and a second one where new electrons are released. This procedure is repeated typically 9–12 times and the typical electron multiplication factor is 3–6 (depends on the potential difference between the two dynodes). If the PM tube contains 10 dynodes and the multiplication factor is 6, the total number of electrons registered at the end of the PM tube

is 610 (∼6 × 107). As a result, the PM tube has enhanced an undetected signal into a larger detectable signal. In PET, and some cases SPECT, the PM tubes are placed in a ring around the patient. In most SPECT and gamma cameras the PM tubes are placed in a square, creating a detector head. In SPECT two or more of these detector heads are rotated around the patient to attain projections from all angles (Figure P.59). Abbreviations: PM tubes = Photomultiplier tubes and SPECT FIGURE P.59 PM tubes arranged over the scintillating crystal of a = Single photon emission computed tomography. gamma camera. Related Articles: PET, SPECT Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, 87(2):171–186; ICNIRP. 2000. Revision of the guidelines on limits Philadelphia, PA, pp. 101–103. of exposure to laser radiation of wavelengths between 400 nm and 1.4 µm. Health Phys. 79(4):431–440; Sihota, R. and R. Tandon. Photoretinitis (Eye) 2011. Parsons' Diseases of the Eye. Elsevier India. (Non-Ionising Radiation) The term is used to indicate damage to the retina of the eye due to exposure to either solar or intense Photothermal effects artificial light. (Non-Ionising Radiation) The thermal effects of artificial opti- Related Articles: AORD, Eye, Retina, Blue light hazard, cal radiation are the effects that are most likely to cause damage. Thermal light hazard These effects may be considered with regard to two processes: Further Readings: Coleman, A., F. Fedele, M. Khazova, coagulation and vaporisation. Either of these effects may occur, P. Freeman and R. Sarkany. 2010. A survey of the optical haz- dependant on the temperature to which the tissue is heated. ards associated with hospital light sources with reference to the Coagulation occurs at temperatures between 60 and 80°C, Control of Artificial Optical Radiation at Work Regulations 2010. at this temperature, collagen fibres within tissue shrink and J. Radiol. Prot. 30(3):469; ICNIRP. A closer look at the thresholds coagulation occurs. Vaporisation occurs at temperatures greater of thermal damage: workshop report by an ICNIRP task group. than 100°C when water within tissue cells boils and vaporises Health Phys. 111(3):300–306; ICNIRP. 2013. Guidelines on limits to become steam. The volume expansion associated with steam of exposure to incoherent visible and infrared radiation. Health production leads to the explosive rupture of the cell walls, and Phys. 105(1):74–91; ICNIRP. 2013. Guidelines on limits of expo- the removal of tissue in this way is called vaporisation. Any fur- sure to laser radiation of wavelengths between 180 nm and 1,000 ther increase in temperature leads to further vaporisation until all µm. Health Phys. 105(3):271–295; ICNIRP. 2004. Guidelines on cellular water is removed. Following from this carbonisation and limits of exposure to ultraviolet radiation of wavelengths between burning of tissue will follow at 250–300°C and 500°C, respec- 180 nm and 400 nm (incoherent optical radiation). Health Phys. tively (Table P.5). Photon 717 P hoton flux Photon TABLE P.5 Frequency energy Wave- (Hz) (eV) length Effect of Localised Temperature Rise on Body Tissue Temp. (°C) Cell Effect 1020 106 1 MeV γ rays 60–80 Collagen is denatured causing tissue contraction and X-rays coagulation 1018 104 1 keV —1 nm > 100 Cellular water boils and vaporises to steam, expansion 1016 Ultra violet 102 causes explosive rupture of cell walls 250–400 Tissue burns and becomes carbonised VL 1 —1 μm 500 Carbonised tissue burns and evaporates 1014 Infra red 1012 10–2 Related Articles: Photochemical effects, Photomechanical 10–4 —1 cm 1010 Short wave radio effects, Photoablative effects Further Readings: Bertolotti, M. 1983. Masers and Lasers, TV 10–6 —1 m 108 An Historical Approach, Adam Hilger Ltd, Bristol; Brown, R. 1968. Lasers, Tools of Modern Technology, Aldus Books, London; P Carruth, J. A. S. and A. L. McKenzie. 1986. Medical Lasers 106 10–8 —1 km Science and Clinical Practice, Adam Hilger Ltd; Henderson, A. Roy. 1997. A Guide to Laser Safety, Chapman & Hall. 10–10 104 Long wave radio Photon VL = visible light (General) A photon is a quantum of electromagnetic radiation. It may be thought of as a ‘packet of energy’ or as a ‘particle’. FIGURE P.60 Electromagnetic spectrum. Photons have no charge and, in a vacuum, travel at 3 × 108 m s−1 (the speed of light). Electromagnetic radiation can be described in terms of waves photon beams produced by linear accelerators have superseded or as photons – the two approaches are entirely complementary. the kilovoltage x-ray beams in many applications. In kilovoltage The wave approach is most applicable at the lower energy end x-ray beams the description of the beam quality is given by the of the electromagnetic spectrum where one is normally dealing beam half-value layer (HVL) of aluminium or copper while the with large numbers of photons – a typical electric light bulb emits specification of a photon beam produced by a linac is indicated about 1020 visible light photons per second – whilst a nuclear med- by the value in megavolt (MeV) unit of the nominal accelerat- icine test may involve a million photons of gamma radiation being ing potential of electrons that produce the photon beam. In this emitted per second from the radioactive material used and each case the linac beam quality is given by the tissue phantom ratio photon can be counted separately. (TPR). Linear accelerators are used to produce photon beams Plank’s constant (h) links frequency (v) and energy: E = hv, usually in the 4–25 MeV range. Advantages of increasing the and the relationship between energy and frequency is given by: energy of the bremsstrahlung radiation consist of a deeper pen- ν = c/λ, where c is the velocity of light; hence the relationship etration in patients and also of the reduction of the relative (dif- between energy and frequency is e = hc/λ. ferential) absorption of photons in various tissues such as fat, The energy of photons is widely described in terms of the elec- bone or muscle as the Compton effect becomes the main interac- tron volt (eV), although the standard unit for measuring energy is tion process. Other advantages are in patient skin sparing, small the Joule (J): One eV is approximately 1.6 × 10−19 J (Figure P.60). penumbra and reduced side scatter. Related Articles: Gamma radiation, Photons, interaction in Related Articles: Tissue phantom ratio (TPR), Half value matter layer (HVL) Photon beam Photon fluence (Radiotherapy) Photon beams used in external beam radio- (Radiation Protection) Photon fluence (φ) is a measure of the therapy can be produced as gamma radiation emitted by a radiation field. If we consider a sphere of cross section, a, then radioactive source such as 60Co or as x-ray radiation created by accelerated electron hitting a target (bremsstrahlung radiation). dN The 60Co source emits two gamma rays of 1.17 and 1.33 MeV j = (P.19) da but mean photon energy of 0.9 MeV is generally accepted for the clinical 60Co beam because of the presence in the beam of where N is the number of particles incident on the sphere. low energy photons resulting from Compton interaction of the primary photons with the collimating system. The bremsstrah- Photon flux lung x-ray beams present a wide continuum energy spectrum (Radiation Protection) Flux (Φ) is defined as the average number whose maximum energy is approximately equal to the energy of photons (N) per unit time (t) that pass through a unit area (A) of the electron beam and the mean energy is between 1/3 and perpendicular to the direction of propagation of energy: 1/2 of the maximum energy. Kilovoltage x-ray beams formed at an accelerating potential ranging from <50 kV up to 300 kV N f = (P.20) have been used in radiotherapy for many years but nowadays At Photon scattering 718 Physical phantoms Photon scattering Related Articles: Photoelectric effect, Compton scatter, (Radiation Protection) Photons are scattered by interactions with Rayleigh scatter, Pair production, Photo disintegration, Thomson orbital electrons of absorber atoms. During scattering the photon scatter, Attenuation loses energy and momentum to the electron(s) it interacts with and moves off at an angle. As energy is lost in the process, there is a Photon(s): mean free path change in the frequency of the photon according to the formula (Radiation Protection) This term is described in the article Mean Energy (E) = Planck constant (h) × Frequency (υ). free path. Related Article: Mean free path Planck constant, h = 6.6 ´10-34 Js Photon(s): secondary Several different types of photon scattering are described: Secondary photons are those that have been removed from the primary photon beam as a result of scattering interactions. • Compton scattering Related Articles: Scattered radiation, Secondary radiation, • Rayleigh scattering Secondary ionising radiation • Thomson scattering Photosensitivity These interactions are described in more detail under the corre- (Radiation Protection) The sensitivity of skin to the effects of sponding headings. optical and ultraviolet (UV) radiation. Exposure of skin to opti- P Related Articles: Compton scattering, Rayleigh scattering, cal/UV radiation will invoke erythemas of varying severity Thomson scattering dependent on radiation dose and the individual’s photosensitivity. Abnormalities of photo-sensitivity (extreme reactions to expo- sure) are broadly categorised into idiopathic (presumed immuno- Photon(s): annihilation in positron decay logical origin), genetically inherited and biochemical aetiologies. (General) Positron decay is a radioactive process in which a pro- The trigger in each of a variety of dermatological conditions is ton in the nucleus changes into a neutron and a positively charged either optical or UV radiation. electron (positron or beta plus particle, β+) is emitted, together Overly photo-sensitive conditions have an abnormally low with a neutrino (ν): p+ → n + β+ + ν. Gamma rays may also be deterministic threshold for effect. Furthermore, normal humans emitted. of different skin-types exhibit different photo-sensitivities for the After emission, the positron (β+-particle) is slowed (loses severity of erythema. Caucasians are generally more photo-sensi- kinetic energy) via interactions with surrounding matter. When tive than Afro-Caribbeans, for example. brought to a near or total halt, the positron forms a particle, a Further Reading: NRPB (National Radiation Protection so-called positronium (Ps), with an electron. The positronium Board). 2002. Health effects from ultraviolet radiation, Report of (with lifetime measured in picoseconds) is consumed in an an Advisory Group on Non-ionising Radiation, Documents of the annihilation process that emits two photons with a back to back NRPB 2002, Chilton, UK. direction: Photostimulable phosphor plate b+ + e- ® Ps ® g + g (Diagnostic Radiology) See Storage phosphor Related Articles: Computed radiography, Storage phosphor, The annihilation is an example of the conversion of mass into Fluorescence, Thermoluminescence energy and the energy of the two photons is given by applying Einstein’s equation (E = mc2) with the consequence that each pho- Photostimulated luminescence ton has an energy of 0.51 MeV. However, the particles are sel- (Diagnostic Radiology) This is a process by which energy is dom brought to a full stop and the residual momentum results in a trapped in a phosphor and then released from the phosphor by the small deviation from a 180° emission angle. use of a laser. It is the basis of computed radiography (Figure P.61). Related Articles: Beta decay, Beta+ radiation, Electron cap- The drawing shows the basis of a storage phosphor. X-ray ture, Photons, Positron decay, Radioactive decay energy causes the promotion of electrons which are captured in f-trapping centres. A laser is used to release the energy which is Photon(s): interaction in matter captured to form an image. See the article on Storage phosphor. (Radiation Protection) There are several different ways in which Related Articles: Computed radiography, Storage phosphor photons may interact with matter in the diagnostic energy range. These processes are described in detail in separate articles: Physical penumbra (Radiotherapy) See Penumbra • Photoelectric effect • Compton scatter Physical phantoms • Rayleigh scatter (Nuclear Medicine) Physical phantoms are used both to simulate • Pair production the conditions being investigated during a patient examination • Photodisintegration and to measure the emission camera parameters (spatial resolu- • Thomson scatter tion, sensitivity, linearity and uniformity). Phantoms used for flood measurements are one example of phantom use in a patient These interactions comprise absorption and scattering of pho- situation. tons from an incident photon beam. The sum of these processes There are a number of phantoms used to measure the imag- is equal to the total attenuation of a beam

of photons passing ing parameters in an emission camera system. For example, through matter. a common set-up to measure spatial resolution is to use a bar Physicochemical adsorption 719 Pinhole collimator Trap modalities, including various possibilities to add extended fea- Conduction band tures such as HIS and/or RIS integration, data sharing, remote access, etc. to facilitate patients and the medical and administra- τtunnelling τ Phonon Relax recombination tive staff of the institution. PACS systems provide reliable, flexible and cost-effective 4f 6 tools for storage and management of imaging data. PACS uses Laser DICOM-based client-server architecture, where imaging modali- F/F+ stimulation PSL 2.0 eV 8.3 eV ties are the main sources of data while users may vary in type, 3.0 eV τE location and modality (Figure P.62). 47 Piezoelectric crystal E3/E2 (Ultrasound) The piezoelectric crystal is the active actuator and e sensor in a conventional ultrasonic transducer. It converts electri- Valence cal energy to acoustical and acoustical energy to electrical. The band Incident most common material used is PZT–lead zirconate titanate. x-rays Related Articles: Transducer, Backing material, Matching layer, PZT, Lens, Linear array FIGURE P.61 Principle of photostimulated luminescence. (Courtesy of J.A. Seibert.) Pincushion distortion P (Diagnostic Radiology) Pincushion distortion is an image aber- ration where an image appears ‘pinched’ towards its centre. It is phantom. The bar phantom consists of a number of line sources mainly associated with image intensifiers, where it is caused by a with a decreasing distance between the line source when mov- non-uniform magnification across the image, with magnification ing along the bar phantom. A subjective approach to estimate increasing away from the image centre (Figure P.63). the spatial resolution is to estimate where one can see the last visible line pair, i.e. before the lines overlap. A more quantita- Pincushion distortion tive approach is to use a point-spread function or a line-spread (Nuclear Medicine) The pincushion image distortion is an effect function where one uses a point and a line source phantom of camera non-linearity. Non-linearity effects occur when the sig- respectively. nal in the photomultiplier tubes (PM tubes) does not change lin- Related Article: Software phantoms early with the displacement of a source across the detector face. Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Consider a source that moves from the edge of a PM-tube to the Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, centre. If the light collection efficiency increases more rapidly Philadelphia, PA, pp. 253–259. rather than linearly as the source approaches the PM tube centre, the x- and y-position signals will also change in a non-linear way, Physicochemical adsorption thus creating the characteristic pincushion distortion. The typi- (Nuclear Medicine) Physicochemical adsorption is a type of cal pincushion distortion of a straight-line object is apparent in chemical reaction of a compound and an adsorbant surface that Figure P.64. The opposite distortion is called a barrel distortion involves the generation of chemical bonds between the compound and is discussed in a separate article. and the surface. Related Articles: Photomultiplier (PM) tubes, Barrel Physicochemical adsorption is the process by which bone distortion scanning agents are taken up. Examples of bone scanning agents Further Reading: Cherry, S. R., J. A. Sorenson and M. E. are MDP and HDP, labelled to Technetium-99m, which are phos- Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, phonate compounds. The exact uptake/accumulation mecha- Philadelphia, PA, pp. 234–235. nism of MDP or HDP in bone is not well understood, but it is assumed that the phosphonate compounds are incorporated into Pinhole collimator the hydroxyapatite or amorphous calcium phosphate by physico- (Nuclear Medicine) A collimator design which typically consists chemical adsorption. of a metal cone made from high absorbing material and a pinhole Related Articles: Tc-99m-disphophonates, Bone, Tc-99m- insert. The main clinical use of a pinhole collimator is to attain labelled bone imaging agents magnified images of small organs like the thyroid. Pinhole col- Further Reading: Zolle. 2007. Technetium-99m limators play an important role in pre-clinical studies with small Pharmaceuticals: Preparation and Quality Control in Nuclear animal SPECT. The collimator insert is placed at the end of the Medicine, Springer, Berlin Heidelberg. high absorbing metal cone. The length of the metal cone is typi- cally 20–25 cm. Picture archiving and communications system (PACS) Gamma rays that pass through the collimator will project an (General) PACS refers to a computerised system used for picture inverted image on the detector. The projected image will be: mag- archiving and communication in medical environments, includ- nified if the distance b between the collimator and the object is ing hospitals, primary care units and/or other institutions related smaller than the collimator to detector distance f; and: minified if to healthcare, research and education. the conditions are the opposite. The relationship between image PACS size, complexity and features may vary significantly size I, and object size O is depending on the needs of the institution. PACS basic functionality includes acquisition, storing, distri- I f = (P.21) bution and management of data originating from various imaging O b Pinhole collimator 720 Pinhole collimator P FIGURE P.62 PACS architecture. FIGURE P.64 Distortion caused by non-linearity in the PM tubes response. This particular distortion is referred to as the pincushion FIGURE P.63 Effect of pincushion distortion on the appearance of a distortion. rectilinear image matrix. When the distance between the object and collimator changes, From Equation P.22 it follows that a large magnification factor for so does the imaged area. If the object is further away from the an image acquired from an object close to the collimator results collimator a larger object can be imaged with a lower magnifica- in a small imaged area. tion; if the object is closer to the collimator a small object can be There are some limitations when imaging 3D objects since magnified to a greater extent. The relationship between the detec- the magnification depends on the source to collimator distance. tor diameter D and the image area projected onto the detector D′ Differences in object depth, i.e. different magnifications may is determined by the magnification factor I/O: result in image distortions and the user must be well aware of this effect or else it may ultimately lead to misdiagnosis and/or inef- D D¢ = (P.22) fective/harmful treatment. I / O The pinhole collimators contribution to the system resolution is Pion therapy 721 P ixel value (l + b) published by these institutions suggests the main sites treated Rcoll » deff (P.23) were astrocytomas and prostate cancers. l Eventually, these centres began to discontinue treatment with treatment ending in the early 1990s. This was due to several fac- where deff is the effective pinhole diameter, accounting for photon tors including ageing and difficult-to-use equipment, advances penetration at the edges of the pinhole opening. A pinhole col- in photon therapy and a lack of positive clinical trials. Currently limator provides very high resolution but also a limited field of (2020), there are no centres treating cancers with pion therapy. view. A way to get a larger field of view is to move the object fur- Further Readings: Goodman, G. 1990. Pion Therapy for ther away from the collimator, but this leads to a loss in efficiency. Cancer – What are the Prospects? pp. 242–243; Pickles, T. The efficiency of a pinhole collimator is 1997. Pion radiation for high grade astrocytoma: Results of a randomized study. Int. J. Radiat. Oncol. Biol. Phys. 37(3); d 3 g » eff cos q (P.24) Wisser, L. 2004. Pion treatment of prostate carcinoma at Paul 16b2 Scherrer Institute (formerly Swiss Institute for Nuclear Research (SIN)) from 1983 to 1992. Cancer/Radiothérapie. 8(2):88–94. where θ is the angle between the pinhole, the central line and the doi:10.1016/j.canrad.2003.12.004. off-centre source. The efficiency will drop quickly with distance as seen in Equation P.24. Pinhole collimators are therefore typi- Piston cally used to image small organs like the thyroid and heart. (Ultrasound) The analysis of ultrasound fields is based on the The effective pinhole diameter is assumption that a simple circular transducer element behaves as P a piston. é a Related Articles: Diffraction, Near zone, Far zone, Transducer deff = d + -1 æ öù êd 2m tanç (P.25) 2 ÷ú ë è øû Pitch where (Diagnostic Radiology) See Helical pitch α is the angle between the walls of the metal cone μ is the linear attenuation coefficient of the cone material Pixel (Nuclear Medicine) The smallest possible element in an image Another pinhole application is the multi-pinhole application is referred to as a pixel. Images typically consist of a 2D array with several hexagonal holes in the pinhole device. This appli- of pixels. In nuclear medicine imaging each pixel contains a cation is not often used in clinical imaging but rather for small value proportional to the intensity of the incoming radiation to animal imaging and research. the corresponding detector element. The number of colours (or Related Articles: SPECT Collimator, Parallel-hole collima- greyscale) that each pixel can represent depends on the number tor, Diverging collimator, Converging collimator, Collimator, of bits per pixel, bpp. A pixel with 1 bbp can have two different Collimator design, Collimator parameters colours (21 = 2) and such images are called monochrome pictures. Further Reading: Cherry, S. R., J. A. Sorenson and M. E. For each extra bit per pixel the number of colours available is Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, doubled (2 bit: 22 = 4, 3 bit: 23 = 8, … 28 = 256). In some systems Philadelphia, PA, pp. 245–246. as much as 24 bpp is used (∼16.7 million colours per pixel) with 8 pixels for green, red and blue. Pion therapy (Radiotherapy) In particle physics terms, a pion is a type of Pixel value meson; a particle composed from a quark and anti-quark pair. (Diagnostic Radiology) Pixel value is a fundamental characteris- The negatively charged pion, π−, composed of a down and anti-up tic of a digital image. The signal in an analogue x-ray film is the quark pair, was first proposed as a candidate for cancer treatment optical density of the image features, which correspond to various in the 1960s. level of x-ray absorption by the imaged tissues. The digitising of π− has several properties that are of interest for cancer treat- an analogue film (e.g. through an x-ray film scanner) transverses ment. Due to its relatively low mass (compared to a proton or car- the optical densities into picture cells (pixels), the content of each bon ion) the accelerator energies required for clinical treatment cell being a number corresponding to the respective optical den- are attainable. The particle’s single negative charge also permits sity of the film (hence to the x-ray absorption of the imaged tis- beam steering with relative ease. sue). This value of the picture cell is the pixel value. The particle also has a unique dosimetric feature. At high In CT scanning (or other imaging methods using back-pro- energies, the particle travels through tissue with relatively low jection reconstruction), the result of the calculation process is a linear energy transfer (LET) (0.3 keV/um). However, at lower number representing the value of each pixel from the image. In energies at the end of range, the particle can be captured by a posi- this particular case, the pixel value is called a CT number - e.g. tively charged nucleus, which in turn leads to an unstable nucleus. water has 0 CT number (or Hounsfield units), which means that The result is a ‘pion star’, an explosion of short-range high LET the content of the picture cell is equal to zero (in reality, this con- nuclear fragments. It was hoped this would attain a high relative tent of various pixels is not equal, but fluctuates around zero, due biological effectiveness and low oxygen enhancement ratio (OER) to the noise). within the target. OER has been suggested to reach around 2.2, The basis of quantitative imaging is the analysis of pixel val- which is comparable to the OER seen in carbon therapy. ues in a digital image – i.e. analysis of the signal received in this Small numbers of patients were treated at research institu- particular part of the tissue as a response to a specific reaction of tions over a period spanning the 1970s to the 1990s. These cen- the tissue to

external energy (e.g. ionising radiation, ultrasound, tres included Los Alamos National Laboratory (USA), TRIUMF electromagnetic). An example of such analysis is bone densitom- (Canada) and Paul Scherrer Institute (Switzerland). The literature etry. Usually, pixel values are measured in a region of interest Plain film radiography 722 Planning target volume (PTV) P FIGURE P.65 Measurement of pixel value (PV) from a digital radio- FIGURE P.66 Measurement of pixel value (PV) in a square ROI of 100 graph using ImageJ software – here PV = 183 for the pixel under the pixels, showing mean pixel value of 187.35, while individual pixel values ‘cross’ cursor (individual pixel with coordinates X = 1184 and Y = 398). in the ROI vary from 180 to 193. Note that the pixel value does not change The same radiograph is shown in Figure P.66 with different contrast (by (compared with Figure P.65), despite the visualisation of the image with using different window parameters). better contrast (resulting from the different window parameters – see the histogram). (ROI) with a certain area – the average value of all pixels in the compiled to a 3D tracer distribution using image reconstruction ROI represents the signal, while the standard deviation of the programs on computers. This technique is called tomographic measurement represents the fluctuation of the signal – i.e. the imaging. noise (see Figure P.65 and Figure P.66). Related Article: Tomographic imaging The process of digital windowing or contrast enhancement of a digital image is related to visualising the pixel values with dif- Planning target volume (PTV) ferent optical densities (different brightness of the pixels in active (Radiotherapy) The planning target volume (PTV) is used in displays). However, the pixel values will have to continue to show radiotherapy in treatment planning. It consists of the clinical the actual signal – i.e. the reaction of the tissue to the respective target volume (CTV) plus margins to account for geometrical energy (see Figure P.65 and Figure P.66). variations and inaccuracies to ensure that the prescribed dose is In analogue x-ray films, the characteristic curve shows the actually absorbed in the CTV. It is purely a geometrical concept, relation between the optical density and the logarithm of the x-ray and may sometimes even lie outside the patient’s body. For a given exposure, while in computed radiography, the characteristic curve CTV, the PTV may vary with different beam arrangements. shows the relation between the pixel values and the logarithm of The use of PTV as a planning volume was proposed by the the x-ray exposure. ICRU in Report 50 (with addendum ICRU Report 62). This Related Articles: Pixel, Bone densitometry, Characteristic report provides a common framework on prescribing, recording curve, Window, Grey levels and reporting therapies, with the aim to improve the consistency and inter-site comparability. It details the minimum set of data Plain film radiography required to be able to adequately assess treatments without having (Diagnostic Radiology) Plain film radiography is another term to return to the original centre for extra information. used for normal x-ray radiography (producing a static x-ray image ICRU 62 added two new concepts – the internal margin (IM) over film). and the set-up margin (SM), which together form the margin between the CTV and the PTV. The IM allows for error from Planar imaging physiological variation of the size and shape of the volume, and (Nuclear Medicine) Images acquired from a specific projection the SM allows for patient positioning and variation in the align- angle are referred to as planar images. Planar images produce a ment of the treatment beams. The margins may vary individually 2D representation of the tracer distribution and planar imaging for the different directions from the CTV (Figure P.67). has played an important part in the history of nuclear medicine. Abbreviation: PTV = Planning target volume. Bone scintigraphy can be used to evaluate the skeletal system and Related Articles: ICRU, Clinical target volume (CTV), Gross to locate fractures invisible to x-ray examinations. Today, images tumour volume (GTV), Treated volume, Irradiated volume, are acquired from a number of different projection angles and Internal margin, Setup margin Plaque radiotherapy 723 Platinum H H H CH3 H Cl C C C C C C H H H H H H n n n (a) (b) (c) FIGURE P.68 Chemical structures of common plastics: (a) polyethyl- Gross tumour volume ene, (b) polypropylene and (c) polyvinylchloride (PVC), respectively from Clinical target volume left to right. Planning target volume Treated volume have a melting point as well as a glass transition. Semi-crystalline Irradiated volume plastics include polyethylene, poly (vinyl chloride) and polyesters. Amorphous plastics include polystyrene, poly (methyl methacry- late) and thermosets. FIGURE P.67 Definition of target volumes as in ICRU 50. Plastics are extremely diverse in their uses due to their versa- tility, ease of manufacture, relatively low cost and impervious- ness to water. Their uses are limited by their density, hardness Further Readings: ICRU. 1993. Prescribing, reporting and and ability to resist heat, organic solvents and ionising radiation. P recording photon beam therapy. International Commission on Plastics generally have low toxicity due to their insolubility in Radiation Units and Measurements, ICRU Report 50, Washington, water and chemical inertness. However, they may contain toxic DC; ICRU. 1999. Prescribing, recording and reporting photon additives, such as plasticisers to make them pliable for products beam therapy (Supplement to ICRU Report 50), International such as food packaging and tubing. Plastics are still too expensive Commission on Radiation Units and Measurements, ICRU Report for large scale uses such as buildings and bridges. 62, Washington, DC. Medical Applications: The medical uses of plastics vary sub- stantially from syringes to prosthetic limbs to medical equipment Plaque radiotherapy casing. Plastics have a range of applications in medical physics, (Radiotherapy) Plaque radiotherapy is a type of radiation ther- including their use as phantom materials and tissue substitutes. apy used to treat eye tumours. A thin piece of metal (usually They are suitable for such uses due to their low toxicity, chemical gold) with radioactive seeds placed on one side is sewn onto inertness, ease of sterilisation and relatively low cost. the outside wall of the eye with the seeds aimed at the tumour. Related Article: Acrylic It is removed at the end of treatment, which usually lasts for several days. Plates, deflection (General) See Deflection plates in cathode ray tubes Plastic (General) Platinum (General) Molar mass >10 kg mol−1 Density at STP varies Symbol Pt Melting point varies Element category Transition metal Boiling point varies Mass number A 195 Atomic number Z 78 Plastic is a general term for a range of synthetic organic solid Atomic weight 195.084 kg/kg-atom materials. Plastics are normally polymers with a high molecu- Electronic Configuration 1s2 2s2 2p6 3s2 3p6 3d10 4s2 4p6 4d10 lar weight, consisting of chains of several thousand repeating 4f14 5s2 5p6 5d9 6s1 units of monomers of mostly carbon and hydrogen, commonly Melting point 2041 K with oxygen, nitrogen, chlorine or sulphur in the backbone. The Boiling point 4098 K properties of plastics are varied by altering the monomer and Density near room temperature 21460 kg/m3 various molecular groups. The adjective ‘plastic’ refers to mate- rials which experience a permanent change of shape known as a ‘plastic deformation’ when strained beyond a critical amount. History: Platinum was first known to pre-Columbian indig- Materials described as plastic in this way, such as aluminium, are enous Americans who extracted this naturally occurring metal not necessarily classified as plastics and, conversely, not all plas- from alluvial sands and alloyed it with gold to produce a white tics exhibit plastic deformation (Figure P.68). gold. In the 1740s Europeans rediscovered the element by exam- Plastics can be classified in several ways, including their ining artefacts from the earlier civilisation, resulting in a descrip- chemical structure (molecular units), such as acrylics and poly- tion of the metal being presented by William Brownrigg to a esters, or the chemical process of their synthesis, such as con- meeting of the Royal Society in 1750. densation and polyaddition. Other classifications or gradings are Medical Applications: Needle electrodes – platinum is often based on their properties such as thermoplastic/thermoset, biode- used as a conducting component for needle electrodes. Needle gradability, density or tensile strength. The molecular structure electrodes pass through the skin and record potentials from small of plastics may be partially crystalline, meaning that they often areas (such as motor units within muscles). Platinum is a good Plumbicon tube 724 Point dose kernel choice for this purpose as it tolerates sterilisation very well and the film to facilitate identification of the type and energy of the tissue reactance to platinum is small. radiation that caused the exposure. Endovascular coiling – as a brain aneurysm treatment, tiny Thermoluminescent materials are also commonly used for platinum coils can be threaded through a catheter and deployed whole-body monitoring, although these materials are less reliable into the aneurysm, blocking blood flow and preventing rupture. at determining the source of the radiation exposure. The coils are made of Platinum so that they are visible on x-ray More recently, some personnel dosimetry companies have scans and are flexible enough to conform to the aneurysm shape. introduced optically stimulated materials (OSL). These have Related Articles: Electrode, Potential difference, Muscle, much greater sensitivity than either film or thermoluminescent Aneurysm (TLD) materials. In situations where there is a significant risk of exposure or Plumbicon tube where exposures are expected to be high, active (direct reading) (Diagnostic Radiology) The Plumbicon TV camera tube has been dosimeters are used. The original active devices were quartz-fibre developed by Philips. The camera has a light sensitive layer (tar- electrometers. These were notoriously unreliable and have now get) of lead oxide (PbO). The operation of the camera is explained been replaced by electronic devices using solid state detectors. in the article Video camera tube. According to the light intensity There are a variety of such solid state devices, commonly referred at the target of the camera an electrical charge pattern is formed to as ‘pocket dosimeters’. In addition to displaying the cumula- over the light sensitive layer. This light sensitive layer (usually tive dose at any instant, they also have the ability to set alarms with 1 in. diameter) is sometimes called retina. When the target to warn the wearer when a particular exposure has been reached P is scanned by an electron beam the variously charged micro areas (Figures P.69, P.70 and P.71). are discharged, respectively the varying discharging current (pro- Related Articles: Thermoluminescent dosimeter, Optically portional to the charge of the layer, hence to the intensity of the stimulated luminescence dosimeter, Personnel dosimeter. incoming light) forms the video signal. The change of the discharge current is related to change of Point dose kernel the conductivity of the micro areas on the target. This means that (Radiotherapy) For photon radiotherapy, a point dose kernel rep- changing the voltage of the layer will not alter the signal. This resents the dose distribution in water resulting from both scat- way the Plumbicon tubes have fixed sensitivity (fixed gain). As tered photons and secondary electrons set in motion by primary this camera cannot apply automatic gain, it is usually used in con- photon interactions at one specific point. The dose distribution junction with an automatic brightness control (ABC) system. Plumbicon TV tubes have small dark current and are less inert than Vidicon tubes (no ghosting effect, their time constant is 2–4 times smaller than Vidicon). This has two effects. From one side the camera is fast and can be used to image rapid movement (i.e. heart movement), as it will not blur the image. From the other point of view this small inertia does not allow for the integration (averaging) of the image, which leads to increased noise in the image. The Plumbicon camera has a characteristic curve with gamma close to 1.0. This means that the camera will have a smaller con- trast range (latitude) than Vidicon, but the picture will be with higher contrast than TV tubes with gamma 0.7 (as is Vidicon). Plumbicon is more expensive than Vidicon. Newer lead oxide TV tubes include Newvicon and Saticon. Plumbicon is better suited for fluoroscopic examination of dynamic objects (for example heart) and for cine/video/DVD FIGURE P.69 Quartz fibre dosimeter. recording of the image. CCD type cameras and flat panel

detectors gradually replace the fluoroscopic x-ray systems with TV camera tubes. Related Articles: Video camera tube, Superorthicon, Vidicon Further Readings: Coulam, C., J. Erickson, D. Rollo and A. James, eds. 1981. The Physical Basis of Medical Imaging, Appleton-Century-Crofts, New York; Oppelt, A., ed. 2005. Imaging Systems for Medical Diagnostics, Siemens, Erlangen, Germany. Hyperlink: The cathode ray tube site, http://members .chello .nl/∼h.dijkstra19/page4 .ht ml PMMA (Perspex, Plexiglass, Lucite) (Radiation Protection) See Perspex characteristics Pocket dosimeter (Radiation Protection) The original monitors developed for whole-body monitoring were small films placed in a holder that provided different amounts of attenuation in different regions of FIGURE P.70 Solid state dosimeter. Point resolved surface coil spectroscopy (PRESS) 725 P oint shear wave elastography (pSWE) 90° 180° 180° RF Gz Gy Gx Echo TE P FIGURE P.72 PRESS pulse sequence. FIGURE P.71 OSL dosimeter. of interest at the intersection of all three slices for a second of the kernel is usually obtained from Monte Carlo simulations time, and the resulting volume-selected spin echo is acquired. where the energy deposited in concentric spherical shells around The combination of pulses also generates a variety of unwanted a point source is tallied. The energy scored in a spherical shell signals, which are dephased using spoiler gradients (hatched in divided by the shell mass is considered as the delivered dose for Figure P.72). some effective radius R(eff) of the shell. The effective radius Use of a spin echo, rather than a stimulated echo as in STEAM, R(eff), defined as the distance of a hypothetical zero-thickness results in a factor of two signal advantage at a given echo time. scoring region from the point, can be evaluated in different ways However, VOI definition can be poorer, particularly at short echo for a finite thickness scoring region and this can influence the times. Nevertheless, contamination with extraneous signal in accuracy of determination of the dose distribution. For a homoge- PRESS is generally minimal. neous medium, the point dose kernel is spatially invariant. If the Related Articles: ISIS, Magnetic resonance spectroscopy, incident direction of primary photons is also invariant, then the Single voxel spectroscopy, STEAM absorbed dose may be computed by a convolution operation using the point dose kernel. This is the case for a parallel-ray beam of Point shear wave elastography (pSWE) mono-energetic photons. However, the incident particle direction (Ultrasound) In pSWE, acoustic radiation force impulse (ARFI) is not spatially invariant for a beam emanating from a finite-sized is applied to induce tissue displacement in a single point location. source as in a linac MVXR beam. The energy deposited by inter- A real-time short-duration acoustic push pulse generates shear actions of primary photons is distributed correctly in an absorb- waves travelling perpendicular to the main ultrasound beam, away ing medium only when the kernel is tilted to align its axis with the from the original region of excitation (Figure P.73). The speed of direction of incident primary photons and this requires additional propagation of the shear waves of a homogeneous and isotropic computation to obtain an accurate dose distribution in the irradi- target is directly proportional to the density and elasticity of the ated material. tissue. Thus, for a given density, shear waves passing through softer tissues moves farther than stiffer tissues, presenting a lower Point resolved surface coil spectroscopy (PRESS) elasticity value. While the push pulse is produced, low-intensity (Magnetic Resonance) PRESS is one of the most common spatial tracking ultrasound beams are continuously emitted parallel to localisation techniques used for single voxel spectroscopy (SVS). the main beam in order to monitor tissue displacement. In order to Because it involves acquisition of a spin echo signal some time obtain a series of data such as the time-to-peak displacement and after excitation, it is particularly suited to nuclear species with recovery time, the tracking beams intercept the shear wave at sev- long T2 relaxation times, such as hydrogen (1H) nuclei (protons). eral fixed locations and time intervals (Bruno et al., 2016). From PRESS is a ‘single shot’ technique, in that it requires a single these data, mainly through time-of-flight algorithms (time taken acquisition of the pulse sequence to achieve localisation, and is for the wave to travel a distance through a medium), quantitative not dependent on post-acquisition signal combination (as is the estimates of the speed of propagation of the shear waves and the case with, for example ISIS). tissue stiffness are obtained (Gennisson et al., 2013) (Figure P.74). The PRESS pulse sequence uses a selective 90° pulse fol- Related Article: Shear wave elastography lowed by two selective 180° pulses. Each pulse is applied with a Further Readings: Bruno, C., S. Minniti, A. Bucci and R. gradient along a different Cartesian axis, and hence they select Pozzi Mucelli. 2016. ARFI: From basic principles to clinical orthogonal planes of spins. The first pulse and gradient combina- applications in diffuse chronic disease-a review. Insights Imaging tion is used to excite a plane of spins, and the second refocuses 7(5):735–746; Gennisson, J. L., T. Deffieux, M. Fink and M. spins at the intersection of the two selected slices to form a spin Tanter. 2013. Ultrasound elastography: Principles and techniques. echo. The third pulse refocuses magnetisation within the volume Diagn. Interv. Imaging 94(5):487–495. Point source calculation 726 Point source calculation P FIGURE P.73 Principle of ARFI imaging. On a conventional grey-scale ultrasound image, acoustic pulses are generated together with the main ultra- sound beam. The push pulses produce shear waves which propagate perpendicular to the main ultrasound beam and are sampled by detection pulses, parallel to the main ultrasound beam. (Bruno, C. et al., Insights Imaging, 7(5), 735, 2016.) (Courtesy of Sook Sam Leong, Department of Biomedical Imaging, University of Malaya Medical Centre.) FIGURE P.74 Measurement of tissue stiffness (elasticity) in a kidney using pSWE (ROI, region of interest). Point source calculation For a point source, the dose distribution is spherically symmet- (Radiotherapy, Brachytherapy) Factors that affect the dose distri- ric. For an ideal and not encapsulated point source of activity A, bution in a medium around a brachytherapy source are in vacuum, the air kerma rate in a point at distance d is given in Equation P.26 (the usual symbols are used for the ratio of the • Distance – the inverse square law mean mass energy transfer and energy absorption coefficients in • Attenuation – in the source itself and in the encapsula- medium and air): tion (not applicable for an ideal point source) • Attenuation – in the surrounding medium A * G K = d air (P.26) • Build-up of scattered photons d2 Point spread function 727 Point spread function The reference air kerma rate is thus A fundamental fact of signal theory plays a role here: convolution corresponds to pointwise multiplication in the frequency domain. A * G KR = d The shift-invariance thus translates to pointwise dependence in d2 the frequency domain. An ideal system would just transmit the image without modi- The kerma rate in vacuum at distance d to a medium m is fying it: this means that an output value at x is just a copy of the m input value at x, and does not depend on neighbouring values: the A * Gd ém ù K tr m = 2 * d ê ú PSF is 1 (or infinity for continuous systems) at x, 0 everywhere ë r ûair else. In practice, imperfect systems create a dependence on some The dose rate in vacuum at this point is (assuming electronic equi- neighbours. The distance from x at which the PSF is non-zero librium and no losses due to bremsstrahlung) explains how the input is spread over the output. This figure shows the magnitude of the PSF for the radial sam- m m pling pattern: A * G d ém ù é ù D = * en m m ê ú = K * en 1 R d2 ê ú * ë r û ë r û d2 air air Attenuation of the primary photons and the build-up of second- A good PSF should have a strong peak: ary photons in the surrounding medium can be described by a this is how much of distance dependant conversion factor/function φ(d). This function the value at the input point goes into the output describes the deviation of the radial dose distribution from the P inverse square law: e radial patterns explain long distance artefacts in m radially sampled images. ém ù 1 We would try to ‘push’ them D * e m = K n R ê ú * 2 * j(d ) (P.27) as far away as possible ë r û d air ese ripples will cause another For brachytherapy sources with ‘higher’ energies, radium, cae- This shows the ‘spread’: how much the type of ringing artefacts in images nearest neighbours of the point affect sium, iridium, etc., the dose distribution is dominated by the the output. The wider the PSF is, the inverse square law, and the loss of primary photons is to a first blurrier the output approximation compensated for by the build-up of scattered pho- tons. The conversion factor varies slowly with distance, and sev- eral approximations have been used based on polynomials, for PSF example the Meisberger polynomial, and exponential functions. For brachytherapy sources with lower energies, for example iodine and palladium used for permanent prostate seed implants, the photoelectric effect plays a more important role in the interac- tions between the photons and the medium. The effect is that the Convolution radial dose function decreases more rapidly with distance than for the higher energy sources. Actual cylindrical sources can also be characterised starting from a point source type formalism, if an anisotropy function is added as a multiplicative factor to Equation P.27. Related Articles: Treatment planning systems (brachyther- p DFT(p) apy), Source models, Meisberger polynomial Further Reading: Gerbaulet, A., R. Pötter, J.-J. Mazeron and E. van Limbergen. 2002. The GEC ESTRO Handbook of Brachytherapy, available at the ESTRO website: www .estro .be (accessed on 9 July 2012). Point spread function The ‘ideal’ image on the left is imaged so that in the frequency The image on the right shows (General) Pixel or voxel values are usually linear combinations domain, on th right, only artefacts due to the radial of the input. In shift-invariant systems, what sort of linear com- samples on the radial pattern sampling in the frequency are selected domain bination does not depend on where pixels are: if the value of the output at position x is say some multiple of the input at x plus some factors times its neighbours, then the same factors are PSF magnitude, radial sampling pattern. true for an output at another position. The system is said to be shift-invariant. This implies that the mathematical effect of the Note that if we know the PSF, algorithms exist for deconvolu- imaging system can be summarised very simply, by stating how tion which may improve the image quality, for example sharpen input pixel values contribute to an output value. The factor for all it. On the other hand, imperfect optics, as happened to the origi- input pixels forms the point-spread function (PSF) of the system. nal Hubble space telescope, lead to space-variant point-spread The operation producing the output from the input is called a functions. convolution: Abbreviation: PSF = Point-spread function. Related Articles: Modulation transfer function, Optical trans- Output = Convolution of input with PSF fer function Point spread function (PSF) 728 Polarity factor Point spread function (PSF) When modelling an imaging system by, e.g. Monte Carlo meth- (Ultrasound) The point spread function (PSF) of an ultrasound ods it is common to add Poisson noise to the simulated images to imaging system is the image produced of a point scatterer (spatial obtain realistic noise levels comparable to clinical situations. The impulse) at a given location. It is the convolution of the trans- Poisson distribution will then follow mit and receive beam responses at that location together with the x electro-acoustic conversion impulse responses. The PSF will vary p( ) m e-m x = throughout the image depending on the beam shapes, but given x! the knowledge of this distribution, the image can be predicted for any source distribution, provided that the system is linear. where μ is the expectation value. Related Articles: Poisson distribution, Monte Carlo Poiseuille’s

law (Ultrasound) The Poiseuille equation describes the relationship Polar coordinates between pressure, flow and the internal dimensions for flow of a (General) The polar coordinate system is a 2D reference system Newtonian fluid in a cylindrical tube of constant diameter with where each point is associated with an angle and a length in the steady laminar flow. Poiseuille performed his experiments using plane. The polar coordinates r and the angle θ can be used to glass capillary tubes and described how the pressure drop (P1 − express the Cartesian coordinates: P2) along the tube was directly proportional to the length (L) of the tube, the flow (Q) through the tube and a constant (K), and x = r sinq P inversely proportional to the fourth power of the diameter. K var- y = r cosq ied with temperature and was later shown to be due to viscosity. This empirically derived relationship was solved by others to give the familiar form of Poiseuille’s equation where μ is the viscosity Polarity effect of the fluid and R the radius of the tube: (Radiotherapy) The current collected from an ionisation cham- ber, exposed to a constant dose rate, changes in magnitude when pR4 (P1 - P2) the polarity of the applied collecting potential is reversed. This Q = is called the polarity effect. The measurement effect can be 8mL accounted for by applying a polarity correction factor. For more information on the origin of this effect, please see article Polarity The law is named after Jean Louis Marie Poiseuille, French physi- factor. cian and psychologist (1797–1869). Related Article: Polarity factor Related Article: Laminar flow Poisson distribution Polarity factor (Nuclear Medicine) The Poisson distribution is a discrete distribu- (Radiotherapy) The current collected from ionisation chambers tion that describes the probability of events occurring at a fixed exposed to a constant dose rate changes in magnitude when the time interval and with a known count-rate. The events are inde- polarity of the applied collecting potential is reversed. When a pendent of each other. For small mean values the shape of the new chamber is purchased the effect on its reading of using pola- distribution is not symmetrical but instead skewed. As the mean rising potentials of opposite polarity should be checked during value increases, the shape will be more symmetrical. A property the commissioning process. For most chambers, this effect may of the Poisson distribution is that the variance is equal to the mean be negligible in photon beams, except for very thin window cham- value. The Poisson distribution can be described as bers used for low energy x-rays. This polarity effect is due to the lack of charged particle equilibrium at the collecting electrode. lk - The effect is derived from the relationship of the various primary f ( e l k;l) = k! and secondary interactions occurring within the ionisation cham- ber. In a photon beam there is a net positive charge near the sur- where face in the build-up region as more electrons are removed from λ is the expected number of occurrences during a time interval a small mass of material than are deposited and it is reduced to k is the number of occurrences zero when transient electronic equilibrium is achieved. Therefore the magnitude of the polarity effect depends on the thickness of The Poisson distribution can be applied in cases where there the collecting electrode, the frontal surface area of the collecting are a large number of possible events but where each event has electrode, the depth of the collecting electrode beneath the sur- a very low probability. An example of this is radioactive decay face of the phantom, the energy of the beam and the field size. The where there are usually a large number of atoms but relatively few magnitude of the effect depends not only on the maximum energy decays within a time interval. The photon fluence that strikes a of the electrons but also on the beam contamination by low energy scintillation camera is also Poisson distributed since the photons electrons produced in the collimation system which are stopped originated from radioactive decays. This means that when per- in the first millimetre or centimetres. If the chamber is placed at forming ROI analysis on an image, the standard deviation of a the surface the photon beam initially interacts with the collect- count measurement is equal to the square root of the mean count ing electrode causing electron to be ejected in the forward direc- level in the ROI. tion. This results in a region of positive charge being established at the site of the interaction. If a negative bias is applied to the Poisson noise collecting electrode a greater positive charge is collected on the (Nuclear Medicine) In scintillation camera imaging the photon electrode than would be collected from ionisations in the cham- fluence and thereby the collected events are Poisson distributed. ber active volume alone. As the depth of the chamber approaches Poisson noise is then the effect of this Poisson distributed fluence. the range of secondary electrons the positive charge due to the Polychromatic beam 729 Polyester in film base ejected secondary electrons is balanced by the negative electrons thickness (namely the thickness of a specific material which stopping in the collector. The largest polarity effect is observed reduces the intensity of the radiation entering the material to half) at the surface of the phantom since this is the position where the varies. The homogeneity factor (HF) is the ratio between the first positive charge induced on the collecting electrode by ejection of and the second HVL and describes the polychromatic nature of secondary electrons is not compensated by those electrons scat- the beam. In the case of polychromatic beams, the HF is less than tered from above which stop in the electrode. The effect decreases one because the beam is getting harder after each HVL. with increasing depth below the phantom surface as some elec- trons from the medium above the chamber stop in the collecting Polycrystalline silicon (Si) electrode. At depths greater than dmax where a transient charged (Diagnostic Radiology) Silicon (Si) is a metalloid chemical ele- particle equilibrium exists the polarity ratio reverses but it is less ment of atomic number 14 and is a semiconductor. It is used in than 1%. For an electron beam the net deposition of charge is posi- the construction of integrated circuits which have enabled the tive at the surface from the ejection of δ rays and negative at the evolution of micro-computing and consequently digital diagnos- depth where the electrons stop in the material. Subsequently the tic imaging. Within diagnostic radiology silicon is used in flat charge deposition from electrons stopping in the medium influ- panel digital x-ray detectors and flat panel liquid crystal displays, ences the magnitude of the collected charge which depends on where it forms the principle material in thin film transistors used the polarity of the collecting electrode voltage. The average of the in active matrix pixel arrays. Silicon generally forms a fourfold, opposite polarity readings permit to obtain the true cavity ionisa- tetrahedrally bonded atomic structure; in crystalline silicon this tion reading. structure is continued over a long range, whereas, amorphous sili- The charge is the collected charge using a positive polarity is con has no long range order and may sometimes contain dangling P given by bonds as opposed to a pure tetrahedral form. Polycrystalline sili- con has a mid range order and consists of arrays of ordered silicon Qp = Q - q1 + q2 crystals. Silicon is produced from raw quartzite (silica sand) that is where purified to create ‘electronic grade’ hydrogenated amorphous q1 is the positive charge created on the collecting electrode silicon. Electronic grade silicon contains less than 0.01 ppm of due to the ejection of secondary electrons in the forward impurities. Polycrystalline Si is obtained from amorphous silicon direction by excimer laser annealing (ELA), which allows films of large q2 is the negative charge due to the electrons produced out- areas to be manufactured relatively cheaply. side the chamber sensitive volume but stopping in the Although amorphous silicon is the most widely used form of collector silicon, the use of polycrystalline silicon in large area flat panel For negative polarity the accumulated charge will be detectors is a topic of active research due to its superior electri- cal properties and recent manufacturing advances. Traditionally, Qn = Q + q1 - q2 amorphous silicon has been used in digital diagnostic imaging as it is easily deposited in thin films over large areas using PECVD Therefore the true cavity ionisation charge is given by (plasma-enhanced chemical vapour deposition). However, although the use of amorphous silicon has been favoured due to ( Qp + Qn ) its ease of manufacture the electrical properties of crystalline Si Q = are superior. This is due to a higher mobility of charge carriers 2 with the application of an external electrical field. Polycrystalline The polarity effect is particularly evident in a parallel plate ioni- silicon carrier mobilities are of the order of 100 cm2 V−1 s−1, com- sation chamber and can be reduced by the design of the ionisa- pared to amorphous silicon which has mobilities of approximately tion chamber in particular reducing the electrode thickness. If 1 cm2 V−1 s−1. the polarity effect is significant it is necessary to perform all the As digital imaging resolution is increasing, prototype flat measurements with both polarities and to evaluate the average of panel imaging systems have been developed which incorporate the two readings. in-pixel amplifiers constructed from polycrystalline silicon, When an ionisation chamber is calibrated at the standard which increase the signal to noise ratio of the final image. In dosimetry laboratory only one magnitude and sign of polarising future developments, by using polycrystalline silicon, it may also potential is normally used. The calibration coefficient therefore be possible for line drivers and readout amplifiers to be integrated refers to that magnitude and sign of the polarising potential. The on the detector array to reduce external electronics and bring standard dosimetry laboratories and the manufacturer of the down manufacture costs. dosimetry systems might refer to the polarity of the dosimeter Related Articles: Semiconductor detectors, Silicon diode polarising voltage with different conventions and terminology detector, Selenium detector, Silicon, Amorphous silicon and therefore, even if these differences do not affect the physics Further Readings: Rowlands, J. A. and J. Yorkston. 2000. of charge collection in an ionisation chamber, in case the cham- Flat panel detectors for digital radiography, Handbook of Medical ber performance with different electrometers is to be compared, Imaging, Volume 1. Physics and Psychophysics, eds., J. Beutel, special attention has to be paid by the user to the polarity of the H. L. Kundel and R. L. van Metter, SPIE Press, Washington, polarising voltage. DC, pp. 223–313; Li, Y., L. E. Antonuk, Y. El-Mohri, Q. Zhao, H. Du, A. Sawant and Y. Wang 2006. Effects of x-ray irradiation Polychromatic beam on polycrystalline silicon, thin-film transistors. J. Appl. Phys. 99: (Radiation Protection) Polychromatic beams consist of several 064501. components of radiation. Each component has a specific linear energy transfer (LET). Therefore due to the different absorp- Polyester in film base tion of each component, the progressive half value layer (HVL) (Diagnostic Radiology) See Film base Polymer gel dosimetry 730 Portal exit dosimetry Polymer gel dosimetry certain rocks and soils and most biological tissues. Synthetic (Radiotherapy) Polymer gel dosimetry is based on the free radical examples include cements, foams and ceramics. Xerogel is a type chain polymerisation of acrylic monomers dispersed in a gelatin of porous medium, which is formed from gels by drying them matrix induced by a radiation exposure. The concentration of the with unconstrained shrinkage. Such materials are used in chro- cross-linked polymers in a region is proportional to the absorbed matography and for thermal insulation. dose. The spatially localised polymerisation can then be imaged Medical Applications: Porous materials are used in medi- by MRI or optical scanning methods. When imaged by MRI the cine for their absorptive properties to prevent contamination, relaxation rate of the polymerised region is linearly proportional for example for wound dressings. In medical physics, MRI can to absorbed dose in a range relevant for radiotherapy measure- be used to study biological porous

media in order to assess flow ments. The change in the optical opacity of the irradiated polymer characteristics. gel has also led to the development of an optical scan technique Related Articles: Ceramics, Gel for measurements. The optical techniques use a methodology similar to CT scanning but with an optical laser beam instead of a Port film fan x-ray beam. Polymer gel dosimeters present many advantages (Radiotherapy) A port film (or portal film) is an x-ray film over gel dosimeters mainly since there is drastically less diffusion image taken for verification of external beam radiotherapy. Two of polymers within the gel matrix so that the radiation-induced types of film exist: those which require 1–3 cGy to give a good changes maintain their spatial integrity. However there are still exposure and those which require around 50 cGy. The low dose some disadvantages such as cost of dosimeters, duration of the films are used for a pre-treatment set-up check and the high P polymerisation reaction after irradiation and toxicity. dose films are used to image for the duration of a beam. They Related Article: Gel dosimetry are often loosely referred to as check and verification films, respectively. Advantages include high intrinsic spatial resolution. Polymer gels Disadvantages include the need for a film processing unit. (General) Polymer gels are gels that contain spatially fixed mono- Application: Port film is usually compared with a reference, mers, which polymerise due to radiation exposure. The degree of or gold standard, image showing the planned treatment set-up. polymerisation depends on the quantity of free radicals generated This may be a digitally reconstructed radiograph (DRR), a sim- by irradiation and therefore the absorbed dose. In this way poly- ulation image, or a previous treatment time image. Analysis of mer gels can be used as 3D dosimeters, which are tissue equiva- treatment accuracy is usually carried out by comparing the posi- lent due to their high water content. Examples of polymer gels tioning of bony anatomy relative to the treatment field edge in the include the so-called BANANA, BANG and MAGIC gels. two images. Polymer gels are found to have greater sensitivity compared Alternative Technology: Port film is being replaced by to conventional Fricke gel dosimeters. They have sufficient Gafchromic film in many centres, as this obviates the need for spatial resolution and enable flexible realistic phantom designs. the film processor. The properties of polymer gels depend critically on conditions, EPIDs are also taking the place of portal film for many such as temperature and exposure to oxygen and light; therefore, applications. calibration of each gel batch is required. Abbreviations: DRR = Digitally reconstructed radiograph Polymer gels can be analysed by MRI as water molecules and EPID = Electronic portal imaging device. change their binding state and exchange protons with the poly- Related Articles: Gafchromic film, Digitally reconstructed, mer. An alternative method is optical scanning of the gel’s optical Electronic portal imaging density which increases with the degree of polymerisation. Hyperlink: http://www .gafchromic .com/ Gafchromic film The term polymer gel may sometimes be used to refer to the general class of gels known as hydrogels. Hydrogels consist of Portal exit dosimetry a water dispersion medium and water-insoluble polymers with (Radiotherapy) Portal exit dosimetry involves evaluation of some hydrophilic groups that act as the gelling agent. Radiosensitive aspect of the dose distribution in the patient which is based on polymer gels are a specific type of hydrogel. measurement of the radiation passing through the patient using a Medical Applications: Polymer gels are used in medical phys- remote imaging detector. The imaging detector used is often an ics for 3D dosimetry of radiotherapy techniques such as conformal EPID or film. This process involves quantitative evaluation of the and intensity-modulated radiotherapy, brachytherapy, stereotactic pixel intensities in the image (whereas the more common set-up radiosurgery and charged particle radiotherapy. measurement with these imaging systems often involves analysis Related Articles: Fricke dosimeter, Fricke based gel, Gel of the shapes of features). dosimetry, Gel, Polymer gel dosimetry, Tissue-equivalent material Calibration: Portal exit dosimetry requires a calibration that relates pixel intensity to dose. This is complex for two main Population inversion reasons: (Non-Ionising Radiation) The process involved in the production of a laser beam, see Laser. 1. The detector will generally not be water equivalent and Related Articles: Laser hence have a different dose response as compared to tissue Porous medium 2. The goal is often to use the measurement in the detector (General) Porous media consist of a solid matrix containing a net- to infer the dose at a point, or a volume within the patient work of pores filled with a fluid (i.e. liquid or gas). They can be characterised in terms of their porosity, permeability and the indi- Hence, portal exit dosimetry often involves complex models that vidual properties of the constituent solid and fluid. They generally may be empirical or calculation based (e.g. using Monte Carlo have a low density, high porosity and a large internal surface area. methods). Such models often involve corrections for radiation Examples of naturally occurring porous media include sponge, field size, patient/phantom thickness, source to detector distance Portal film digitisation 731 P ortal image and the characteristics of the detector (including its linearity with dose). Plane Dosimetry: One application is plane dosimetry. The plane chosen may be a particular plane through the patient (often the isocentre, or middle of the tissue thickness), the entrance or exit plane, or the plane of the detector. The measurement is often compared with a TPS prediction. In the case of dosimetry in the plane of the detector, the detector is often modelled as an addition slab of material in addition to the patient data. Volumetric Transit Dosimetry: This involves taking the 2D distribution of intensities in the image and back-projecting through a representation of the patient’s anatomy (often the plan- ning CT scan) to determine a 3D dose distribution. The final pro- cess is a mixture of measurement and planning calculation. Figure P.75 presents a schematic diagram of volumetric transit dosimetry. Abbreviations: EPID = Electronic portal imaging device, TPS = Treatment planning system and CT = Computed tomography. Related Articles: Port film, Electronic portal imaging, Dosimetry P Further Readings: Hansen, V. N., P. M. Evans and W. Swindell. 1996. The application of transit dosimetry to preci- sion radiotherapy. Med. Phys. 23(5):713–721; Heijmen, B. J., K. L. Pasma, M. Kroonwijk, V. G. Althof, J. C. Deboer, A. G. Visser and H. Huizenga. 1995. Portal dose measurement in radio- FIGURE P.76 Portal film scanner with film ready to be scanned. therapy using an electronic portal imaging device (EPID). Phys. Med. Biol. 40(11):1943–1955; Huyskens, D., J. Vandam, and A. Dutreix. 1994. Midplane dose determination using in-vivo dose of films and feed them to an application via a TWAIN interface. measurements in combination with portal imaging. Phys. Med. With the development of high quality PC document scanners, Biol. 39(7):1089–1101. more mainstream scanners have been used for portal film digiti- sation with promising results. Portal film digitisation Abbreviation: TWAIN = Standard for image acquisition (Radiotherapy) Portal film is used in many aspects of radiother- devices. apy verification and quality assurance. The advent of software Related Article: Port film that allows automatic analysis of film data has greatly increased Further Reading: Devic, S., J. Seuntjens, E. Sham, E. B. its versatility and ease of use. Such software requires the data in Podgorsak, C. R. Schmidtlein, A. S. Kirov and C. G. Soares. the film to be digitised as an image that can be analysed pixel by 2005. Precise radiochromic film dosimetry using a flat-bed docu- pixel. ment scanner. Med. Phys. 32(7):2245–2253. Film Scanners: Specialist medical quality film scanners are Hyperlink: Vidar website, http://www .vidar .com/ available. The most common are probably the systems produced by Vidar (Figure P.76). These can scan either a single or stack Portal image (Radiotherapy) Portal imaging involves acquiring images using a radiotherapy treatment beam. The images are often used to ver- ify treatment accuracy by measuring the positions of anatomical Radiation structures relative to the radiation field edge. In many cases the source structures used are bony landmarks. More recent attempts have been made to use portal imaging to obtain dosimetric information. Comparison with Diagnostic Radiology: Radiotherapy treat- Dose ment beams use typical x-ray energies of 4–20 MV. This com- distribution pares with the typical energies used for diagnostic radiology of 50–120 kVp. The Compton effect dominates for portal imaging energies and the photoelectric effect is more dominant for diag- nostic energies. The Compton effect has a low Z (atomic number) CT data dependence, whereas the photoelectric effect has a high Z depen- dence (typically Z4) which means that tissues such as bone, which have a high calcium component (Z = 20), have greater inherent contrast than in portal imaging. Thus portal images are generally of lower contrast than diagnostic images. In addition the energy is selected to achieve optimal imaging quality for each anatomic site Image in diagnostic radiology, but the energy for portal imaging is the energy used for treatment which is selected based on dosimetrical Back-projected pixel considerations. Figure P.77 shows a diagnostic energy image of a phantom and Figure P.78 shows a portal image of the same phan- FIGURE P.75 Volumetric transit dosimetry. tom. The difference in image quality is evident. Portal radiography 732 Positive contrast media Imaging Technologies: Film has been traditionally used for portal imaging. Detector/screen combinations have been opti- mised to improve the image quality for a given dose. Competing technologies include the electronic portal imaging device (EPID). Early EPIDs were based on camera systems or liquid ionisation chamber arrays. Recently amorphous flat panel systems have been applied to both diagnostic radiology and portal imaging. Abbreviations: Z = Atomic number and EPID = Electronic portal imaging device. Related Articles: Port film, Portal film digitisation, Electronic portal imaging Further Reading: Herman, M. G., J. M. Balter, D. A. Jaffray, K. P. Mcgee, P. Munro, S. Shalev, M. van Herk and J. W. Wong. 2001. Clinical use of electronic portal imaging: Report of AAPM radiation therapy committee Task Group 58. Med. Phys. 28:712–737. Portal radiography P (Radiotherapy) Portal radiography is synonymous with portal imaging. It involves acquiring images using a radiotherapy treat- ment beam and using these images to verify treatment accuracy by measuring the positions of anatomical structures in the treat- ment beam. The positions are most commonly measured relative to the edge of the radiation field. The accuracy of set-up is evalu- ated in comparison with a reference image, which is usually a digitally reconstructed radiograph (DRR), a simulator image, or a previously acquired portal image. Two approaches to portal radiography are generally followed: 1. A low dose image to check the set-up accuracy at the FIGURE P.77 Diagnostic energy image of the head of a humanoid start of treatment, phantom. 2. An image over a longer time, with a larger dose, dur- ing the treatment to verify that the treatment has been delivered accurately. Abbreviation: DRR = Digitally reconstructed radiograph. Related Articles: Portal imaging, Port film, Electronic portal imaging Position-sensing photomultiplier tubes (Nuclear Medicine) This is a special photomultiplier (PM) tube design that allows the user to determine the interaction point on the photomultiplier tube surface. These PM tubes are mainly used for nuclear medicine imaging purposes. Examples of position-sensing photomultiplier tubes are a micro-channel plate multiplier where the electron cloud created in the gamma interaction is confined in a single small channel, and a hybrid photomultiplier tube where the electrons from the photocathode are accelerated onto a corresponding position on a silicon detector. Positive contrast media (Magnetic Resonance) This term refers to contrast agents that make particular tissues more conspicuous by increasing the sig- nal from them. In MRI, image contrast results from the interplay between the NMR properties of hydrogen nuclei (protons) in tissue and pulse sequence parameters. Positive contrast is usually achieved by using a T1-shortening agent, usually a paramagnetic agent such as a gadolinium chelate, together with a T1-weighted pulse sequence. The T1 of protons in regions receiving a high concentration of the agent is shortened, resulting in faster recovery of longitudinal FIGURE P.78 Portal image of the head of the same humanoid phantom. magnetisation and hence increased signal. Positive-ion cyclotron 733 Positron emission tomography (PET) Most contrast agents are administered intravenously, but there PET imaging is the

abbreviation for positron emission tomog- are also orally administered positive contrast agents to increase raphy imaging. The radionuclides used in PET are β+-emitters. the conspicuity of the bowel. These agents are sometimes based The system is optimised for detecting the annihilation photons on gadolinium chelates, but also include compounds such as ferric and the localisation process takes advantage of the fact that the ammonium citrate and oil emulsions (one study demonstrated the two annihilation photons are emitted at an angle of 180° to each utility of ice cream for this application). other. The localisation process is called annihilation coinci- Related Articles: Contrast agent, Gadolinium chelate, dence detection (ACD) and with this technique the use of high Paramagnetic contrast agents, Negative contrast media absorbing collimators is avoided, thus most PET systems have high count rates relative to SPECT and conventional scintigraphy Positive-ion cyclotron (Figure P.80). (Nuclear Medicine) In a positive-ion cyclotron, positively charged PET Clinical Applications: Conventional x-ray and particles are accelerated in circular paths in a magnetic field to CT-scanners produce morphological images that give information very high energies. Particles to accelerate are for example protons about the inner structure of the human body. PET on the other or alpha-particles. When the particles have achieved a very high hand produces images that describe functional processes inside energy they are extracted from the cyclotron using a negatively the body. charged electrode and directed against a target for production of There are many clinical uses of PET: oncology (e.g. locali- radionuclides. sation of tumours and metastases using F18 – FDG as a tracer), mapping human heart function and clinical diagnosis of brain Positron decay diffusion diseases like Alzheimer’s. It is also used, together with P (General) Positron decay is a radioactive process in which a pro- functional magnetic resonance imaging, to map human brain ton in the nucleus changes into a neutron and a positively charged activation. electron (positron or beta plus particle, β+) is emitted, together Today, PET is more often being applied in a combination with with a neutrino, ν: p+ → n + β+ + ν. Gamma rays may also be a CT or MR-scanner to produce both morphological and func- emitted. tional images. These systems are combined into one camera so The resultant nucleus has one less proton and one more neu- they share the same system of reference. If two image sets are tron. Positron decay is analogous to beta decay. The positrons acquired in two different image systems they do not share the emitted have a spectrum of energies with a definite maximum same system of reference (even though it is possible to use mark- energy, characteristic of the radionuclide involved, again analo- ers, etc.) and the matching between the two images can be a gous to the beta particles emitted in beta decay. source of error, for example when using the matched images to An example of positron decay is the decay of fluorine-18 to select a target volume to irradiate in radiotherapy. A combina- oxygen-18: tion between PET and other scanners will minimise the match- ing source of error and produce an image with both physical and 18 18 F ® O + b+ + n functional information. 9 8 Dedicated PET Systems: A number of dedicated PET scanners have been developed and most new scanners have discrete detec- The maximum energy of the positron is 0.63 MeV. The decay tors placed in a ring. These scanners have the advantage that they scheme is shown in Figure P.79. Ninety per cent of F-18 atoms are able to collect all projections simultaneously, thus decreasing decay by this process, the remainder by electron capture. the acquisition time and probability for patient movement. After emission, the positron loses kinetic energy and is then One of the most common scintillation detectors used in PET annihilated by combining with an electron to form two photons, is BGO which is well suited for this purpose because of its high referred to as the annihilation radiation: β+ + e− → γ + γ. The density and Z-value allowing it to effectively stop high annihi- annihilation is an example of the conversion of mass into energy lation photons. These detectors are arranged in rings along the and the energy of the two photons is given by applying Einstein’s equation (E = mc2) with the consequence that each photon has an energy of 0.51 MeV. Related Articles: Beta decay, Beta+ radiation, Electron cap- ture, Radioactive decay Positron emission tomography (PET) (Nuclear Medicine) 18 9F 110 m β+ 18 8O FIGURE P.80 Registration of two annihilation photons in a ring PET. The registration of the two photons are near-simultaneous. Because of FIGURE P.79 Decay scheme of Fluorine-18 (3% decay by electron cap- characteristics of the annihilation radiation coincidences can be sepa- ture – not shown). rated by a coincidence processor, thus there is no need for collimators. Post-processing 734 Posterior enhancement trans-axis direction and a PET scanner can have several rings. application MRI can give more information about the cancer tis- These scanners can run in either 2D or 3D mode. In a 2D mode sue, while CT data is necessary for calculating the distribution of the rings are separated by a lead shielding (septa) that prevents radiation in the tissues. Combination of several imaging devices, photons that are not parallel to the plane of the detector ring to such as PET-CT, helps in co-registration of two modalities. pass. A number of modifications in the 2D mode can be made Related Articles: Image segmentation, Gated acquisition to increase the sensitivity of the scanner (read more in PET data Further Reading: Hsieh, J. 2003. Computed Tomography, acquisition article). In 3D mode there are no septa between the SPIE Press, Washington, DC. rings, this increases the scanner sensitivity but the number of false and scatter coincidences are increased. Post-processing Related Article: Data acquisition PET (Nuclear Medicine) Post-processing refers to the processing of Further Reading: Cherry, S. R., J. A. Sorenson and M. E. image data following an acquisition. Common post-processing Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, procedures include smoothing, realignment and a number of cor- Philadelphia, PA, pp. 346–348, 358. rections, for example, background and uniformity correction. Post-processing Post-processing (Diagnostic Radiology) After the reconstruction and corrections (Ultrasound) In diagnostic ultrasound post-processing controls for the imaging conditions medical images are often post-pro- are used to alter the grey or colour levels in B-mode images, cessed for further analysis (Figure P.81). Post-processing may con- colour flow images and the spectral Doppler sonogram. By alter- P sist of several operations such as image enhancement, for example ing the grey level for different echo levels, the appearance of the by filtering, segmentation of the image into various regions and image is altered (Figure P.82) and the difference between spe- co-registration of various images. Also 3D reconstruction of dif- cific echo levels can be emphasised. Colour maps can be used ferent structures by using surface or volume rendering methods for B-mode or spectral Doppler displays. In colour flow imaging, is increasingly used in image analysis and planning of operation post-processing ascribes different colour maps to the CFI map of or therapy. velocity vectors. In CT imaging many corrections are often made already in The curves (lower right in each greyscale image) on Figure P.82 connection with the image reconstruction. These include correc- show the greyscale level (y-axis) against the echo levels received tions for back-projection artefacts, aliasing, partial volume effect (x-axis). Colour maps are also available on some systems although and beam hardening. For instance, the beam hardening error can they are not widely used in clinical practice. be quite sufficiently compensated for by remapping the projec- tion samples based on known water attenuation characteristics, Posterior because more than 80% of the human tissue is water. Some (General) Posterior: After, behind, following, towards the rear movement artefacts can be corrected for after the reconstruction. (e.g. the shoulder blades (scapula) are located on the posterior side Some body movements can be detected and corrected manually, of the body). breathing movement can be compensated by holding the breath, See also article on Anatomical relationships but movements affected by the heart can only be corrected by using gated image acquisition and adequate post-processing of the Posterior enhancement images. (Ultrasound) The ultrasound beam, unless travelling through Because the human visual system can only process a few grey fluid, is continually attenuated during its pulse-echo cycle and so levels at a time and the monitors are able to show a restricted it is typical for echoes received from deeper (distal) reflections to number of grey levels, the CT image must be windowed. While have weaker signals than proximal ones. Time-gain-compensation the whole dynamic range of the image is c.4000 HU (Hounsfield (TGC) is often used to increase the gain of the received signal units), only for instance 100 HU are displayed at a time, depend- from distal regions. ing on the imaging application. As the ultrasound beam propagates, it is likely to encounter Image matching of several imaging modalities is mostly based areas/objects of high attenuation (e.g. bone), and conversely, areas on standard anatomical landmarks. Image matching plays an of low attenuation (e.g. a water-filled cyst). The ultrasound beam important role for instance in radiation therapy planning. In this cannot accurately account for the non-uniformity of the various – and ever-changing – media it will travel through. If a section of the beam travels through a (cyst) region of low attenuation – but the rest does not – this section will retain more power (have increased echogenicity) at depths posterior to the cyst compared with the surrounding beam. Echoes received from this post-cyst section will therefore be relatively large compared with those from the Imaging surrounding regions and as a result have an increased brightness by CT Post Image using reconstruction processing analysis on the display image. x-rays Posterior enhancement can be considered the opposite to the shadow artefact, which is conversely the effect of a section receiving greater attenuation to the surrounding beam. It can be used to determine if a cyst is liquid filled or not. A transverse image of a bladder exhibiting posterior enhancement (bright echoes below the bladder, marked by the bracket) is shown in FIGURE P.81 Steps of CT imaging process. the figure. Posteroanterior (PA) projection 735 Power amplifier P FIGURE P.82 Image of a carotid artery with different post-processing curves applied. Ohm’s Law: The potential difference (V) is related to the cur- rent (I) and resistance (R) using Ohm’s law: V = IR SI Units: The International System of Units the SI unit of electri- cal potential difference is the volt (V). Related Articles: Potential energy, Voltage Potential energy (General) Definition and Units: Potential energy is the energy stored within a particular system, which results from the systems con- figuration, for example, the position of a charge within an electric field, or a mass in a gravitational field. Potential energy is equal B-mode ultrasound image of a bladder exhibiting posterior enhance- to the work done in moving an object from a reference point to a ment (bright echoes below the bladder, marked by the bracket). given position. It is commonly represented by the symbols PE, U and V. Like all other forms of energy the SI unit of potential Posteroanterior (PA) projection energy is the Joule (J). The force acting on an object (F) at a point (General) There is a convention where the radiographic technique (x) is related to the potential energy (V) by projection is identified by the direction of the x-ray beam. In the posterior–anterior projection, the x-ray tube produces an x-ray F beam which passes through the front to the back of the patient to ( ) dV x = - dx produce an image. Related Article: Technique projection Examples: Examples of potential energy include elastic, as in a stretched spring, magnetic, gravitational and electrical potential Potential difference energy. (General) Related Article: Potential difference Definition: In general, potential difference is defined as the difference in potential between two points in a scalar field. In Power amplifier electrodynamics the potential difference or voltage (V) between (General) A power amplifier is intended for delivery of power to a two points in an electrical circuit is defined as the amount of elec- load connected

to the output stage of a system. In medical imag- trical energy (W) changed to other forms when unit charge (q) is ing systems, power amplifiers are often used to supply actuators transferred, i.e. V = W/q. for movable parts of the equipment. Power breaker 736 PRDSR (Patient Radiation Dose Structured Report) Power breaker (General) See Circuit breaker Power Doppler (Ultrasound) Power Doppler is a generic name for a flow imaging technique where the power, or energy, of the Doppler is displayed, rather than the actual Doppler shift as a measure of velocity. The technique can be used with other trade names such as colour power angio, or energy Doppler. The advantage of power Doppler is that it is more sensitive to slow and small flows than colour Doppler, that is, when velocity is estimated. The same data is used but as the velocity information is disregarded, the gain can be increased by as much as 10 dB before the displayed signal is lost in noise. In colour Doppler, the first lag of the autocorrelation is used to cal- culate the velocity information, but in power Doppler, the zeroth lag is used. Essentially, this corresponds to squaring the Doppler time signal, that is, I2 + Q2, where I represents the in-phase signal, and Q the quadrature component, as obtained from a quadrature P demodulation. Another advantage is that power Doppler is virtu- ally angle independent as the power in the Doppler signal is inde- FIGURE P.83 Power Doppler image of the cortex and medulla in a kid- pendent of the Doppler angle (as opposed to the Doppler shift). In ney transplant. The orange/yellow scale is graded to show the strength of practice however, flows perpendicular to the sound propagation the colour signal which is an indication of the number of moving scatter- direction are not displayed properly, due to influence of the wall ers in the sample. filter that suppresses low velocities. That the power of the signal is displayed means, however, that there is a depth dependence, or rather, a dependence of attenu- ation. The displayed power is also dependent on such factors as gain, transmitted power, pulse repetition frequency/scale and aperture size. This fact has prevented power Doppler from becom- ing a quantitative tool, and is instead used qualitatively as a means to determine if there is there blood flow present or not. This often gives sufficient information, and has been used to verify flow in kidney transplants for instance. Care should be taken however, since power Doppler is a measure of movement and can give flow images in pulsatile arteries and veins even when there is no net flow through the vessel. When the technique was introduced at the beginning of the 1990s, approaches were suggested to normalise for the aforemen- tioned dependencies. Since then these have successfully been employed to evaluate flow in fetal lungs (normal, intrauterine growth restricted, and in cases with congenital diaphragmatic hernia), and the brain of normally grown and growth restricted FIGURE P.84 Spectrum analyser. foetuses. Some vendors also offer the total power in a volume of a 3D-data-set as a measure of the amount of flow, as a tool in off- line analysis of recorded data (Figure P.83). commonly, the power spectrum is calculated by using the discrete Related Articles: Autocorrelation, Colour flow imaging Fourier transform. The power spectrum of periodic signals is rep- resented by peaks at discrete frequencies, quasiperiodic signals by Power gain peaks at linear combinations of irrationally related frequencies. (General) See Power amplifier Stochastic signals have a continuous spectrum. Theoretically, one should have an infinitely long sequence of continuous data to Power injector calculate the spectrum. Since that is not possible, limitations in (Diagnostic Radiology) Power injector is a device used in x-ray calculation accuracy are present due to restricted amount of data angiography to inject contrast media (with specific debit) into and due to sampling frequency (Figure P.84). the blood vessels and the heart. Normally the contrast media is Further Reading: Weisstein, E. W. 1999. Power Spectrum. administered at the required anatomical place through a catheter, From MathWorld-A Wolfram Web Resource. http: / /mat hworl d which is connected to the power injector. The x-ray generator (as .wol fram. com /P owerS pect r um .ht ml (accessed on 9 July 2012). well as ECG monitor) is connected to the power injector, allowing imaging of specific phases of blood circulation. PPI (partial parallel imaging) Related Article: Angiography (Magnetic Resonance) See Partial parallel imaging (PPI) Power spectrum PRDSR (Patient Radiation Dose Structured Report) (General) The power spectrum gives a plot of the portion of the (Radiation Protection) See Patient Radiation Dose Structured power (energy per unit time) of a signal against frequency. Most Report (P-RDSR) Preamplifier 737 Presampling MTF Preamplifier z (General) A preamplifier is an electronic amplifier at the input of a measurement and processing channel or system. Most preampli- fiers have high gain and high input impedance. The preamplifier ω0 circuit may have a separate housing positioned near the signal M source to be measured. Related Article: Operational amplifier Precession (Magnetic Resonance) Precession is the change in the direction of the rotation axis of a rotating object and, in MRI, precession refers B0 to the precession of nuclear spins in the presence of an external magnetic field. The magnetic moment of a nucleus rotates (spins) around its axis and, if positively charged, its spin generates a mag- y netic field and a magnetic dipole moment parallel to the rotation axis. The magnetic dipole moment can be expressed by a mag- netisation vector M. When placed in a magnetic field B = B0n, where B0 is the magnetic field strength and n is the direction of the field, M will experience a torque. The equation of motion for x P M is then given by FIGURE P.85 When a magnetisation vector M is the subject of a time- dM = gM ´ B (P.28) independent magnetic field along the z-axis, it will start to rotate in the xy dt plane with the Larmor frequency. In a time-independent static magnetic field along the z-axis, the solution to Equation P.28 is given by the so-called Bloch-equations: Frequency of measurements High M Accuracy accuracy–low x (t ) = Mx (0)cos(w0t ) + My (0)sin (w0t ) precision My (t ) = -Mx (0)sin (w0t ) + My (0)cos(w0t ) Low Mz (t ) = Mz (0) Measured accuracy–high Actual precision cision value value Pre where the Larmor frequency is given by ω0 = γB0. Above, T1 and T2 relaxation are neglected. FIGURE P.86 Illustration of precision and accuracy. The precessional motion of the magnetisation vector in the transverse plane (xy plane) is the basis for MRI experiments where an external rotating magnetic field of the same frequency a constant stand-by filament current through the cathode (>1 A). as the magnetisation vector is applied to produce the magnetic This way the time to heat up the filament from the pre-heating resonance phenomenon (Figure P.85). to the requested temperature is much shorter (less than a sec- Related Article: Larmor frequency ond). When performing radiography the operating x-ray exposure switch has normally two phases (two-steps button). The pre-heat- Precision ing produces some thermal electrons, hence it is important not to (Nuclear Medicine) The measured spread in observations of a mea- switch the high voltage on before the cathode is heated to its high surable quantity is referred to as the systems precision. Precision temperature. However some simple dental x-ray equipment allow is also referred to as reproducibility or repeatability. Precision is for switching of the high voltage while the cathode heats up – see intimately related to accuracy, which refers to a system’s ability Figure V.21 from the article Voltage waveform. to obtain the actual value. As illustrated in Figure P.86 accuracy Related Articles: Filament heating, Voltage waveform of a measurement can be high while the precision is low and vice versa. Preparation (first trigger) (Diagnostic Radiology) Usually the exposure switch in x-ray radi- Pre-heating of cathode ography has two stages. The first stage (preparation or first trig- (Diagnostic Radiology) The cathode of an x-ray tube can work at ger) supplies the necessary filament current to the cathode and high temperature (necessary for production of the thermal elec- rotates the anode to the desired speed (rpm). The second stage trons – the anode current) for not more than 1000 working hours. (exposure or second trigger) supplies the high voltage to the x-ray Due to this reason the cathode is heated to this high tempera- tube and produces the exposure. ture for a limited time only (during the x-ray exposure). However to heat the cathode from room temperature to more than 2000 Presampling MTF K takes significant time. This means that an exposure will start (Diagnostic Radiology) The modulation transfer function several seconds after the exposure switch is pressed. In order to (MTF) describes the process of signal amplitude modulation for keep the heating time short the cathode stays always pre-heated to the different frequencies by the system transferring this signal temperature around 1500 K. The pre-heating is made by applying – in particular the response of an imaging system to an input Prescribed dose 738 P retargeting signal of varying spatial frequency. This modulation in an imag- Pressure and temperature correction factor ing system is the combined influence of all its components – (Radiotherapy) In an ionisation chamber open to the ambient air that is each component has its own MTF. For example in an the mass of air which is inside the cavity chamber is subject to x-ray imaging system we have the influence of the focal spot > variation. The Standard Laboratories report the calibration fac- x-ray detector > monitor, etc. (such system is a cascaded linear tor of an ionisation chamber for reference conditions for pressure, imaging system). The summary MTF of such a system is the temperature and humidity. If the ambient conditions at the user product of the MTFs of all separate components. However this are different from the standard conditions, the user has to make requires that the system is shift invariant (i.e. the image of the a correction for the pressure and temperature differences because object is independent of its position within the imaging field). the ionisation per unit of mass of air varies with the density of air This assumption is true for analogue imaging systems; however, inside the chamber cavity. The correction factor to convert the it cannot be applied in digital systems. This is due to the fact cavity air mass to the reference condition is given by that the imaging chain in such systems includes another func- tion, the digitising of the image. The digitising process samples 273.2 + T (°C) P ft, p = 0 the signal and divides it into pixels (sampling). According to 273.2 + T0 P (kPa) the Nyquist theorem this inevitably introduces aliasing (i.e. the digital image of an object depends on its spatial frequency con- where tent relative to the pixels.). The sampling frequency influences P0 and T0 are the reference pressure and temperature at the the MTF and we cannot use the summary MTF as an overall Standard Laboratory P characteristic of the imaging system. However the MTF of the T and P are the ones at the user imaging system before the digitising (the presampling MTF) is not affected by the undersampling phenomenon. The presam- Generally the reference value for the pressure is 101.3 kPa and pling MTF can be used to compare the performance of imag- for the temperature is 20°C. The correction factor ft,p is calculated ing systems. Normally the presampling MTF contains high from the ideal gas law and it is strictly valid without a variation spatial frequencies, which otherwise would be affected by the with temperature of the ionisation chamber volume. Chambers undersampling. The presampling MTF can be measured either with plastic walls may expand their volume with temperature directly before the digitising, or by obtaining and analysing the whereas graphite wall chambers exhibit a very small volume vari- line spread function of the system (LSF). To do this a narrow ation with temperature. When measurements in a water phantom metal slit is placed at slight angle to the sampling direction and are performed the chamber waterproof

sleeve should be vented to is imaged. The resulting image is a bright line on dark back- permit a fast equilibrium between the ambient air and the air in ground, produced by the x-rays passing through the metal plate the cavity chamber. with slit. Each sampling of the image of the line has small offset, relative to the next sampling, what results in reduced sampling Pressure parameters gaps (when the samplings are combined) – that is having the (Ultrasound) The instantaneous change in pressure at a single effect of increased sampling frequency. This way the resulting field point exerted by a sound wave can be seen as the ampli- presampling MTF includes higher spatial frequencies. tude waveform signal from a pressure sensitive device such as a Related Articles: Modulation transfer function, Aliasing, hydrophone. Where there is non-linear propagation, the positive Analogue to digital converter (ADC) going pressure changes may differ significantly from the negative going changes. These amplitudes may be described independently Prescribed dose as the peak positive pressure P+ and the peak negative pressure P (Radiotherapy) The prescribed dose is the dose specified by the (Figure P.87). radiation oncology team as being appropriate to achieve the pur- Related Articles: Intensity, Hydrophone, Acoustic pressure pose of the treatment, for example tumour eradication or pal- liation, within the bounds of acceptable complications. Dose Pretargeting prescribing and reporting is described in various recommen- (Nuclear Medicine) An approach to minimise the radiation dose dations, for example ICRU Reports. Prescribed dose should be to normal tissue in patients undergoing radioisotope therapy. related to the ICRU reference point. The dose at the ICRU refer- Pretargeting refers to the method of sending chemical compounds ence point is called the ICRU reference dose and should always that target and bind to specific tumour cells. After some time has be reported. Abbreviation: ICRU = International Commission on Radiation Units and Measurements. Related Articles: Treatment planning system, Critical struc- P tures, Fractionation, Planning target volume, Normalisation point + Further Readings: ICRU. 1985. Dose and volume specifi- cation for reporting intracavitary therapy in gynecology, ICRU Report 38, ICRU, Bethesda, MD; ICRU. 1993. Prescribing, record- ing and reporting photon beam therapy, ICRU Report 50, ICRU, Bethesda, MD; ICRU. 1998. Dose and volume specification for reporting interstitial therapy, ICRU Report 58, ICRU, Bethesda, 0 MD; ICRU. 1999. Prescribing, recording and reporting photon beam therapy (Supplement to ICRU report 50). ICRU Report P– 38, ICRU, Bethesda, MD; ICRU. 2004. Report 71. Prescribing, recording and reporting electron beam therapy. ICRU Report 71, ICRU, Bethesda, MD. FIGURE P.87 Pressure parameters. Primary barrier 739 Primary standard passed the chemical compounds will have bound to the tumour Target cells and all excessive chemical compounds are cleared from the body. It is then possible to inject radiolabelled tracer that targets the chemical compounds attached to the tumours. One example of a two-step pretargeting method is to use streptavidin-conjugated Primary primary antibodies that target the tumour. The antibodies are collimator then allowed to bind to the tumour cells and the excess antibodies are cleared from the body. Thereafter radiolabelled biotin can be injected. The radiolabelled biotin binds to the tumour because of Flattening the streptavidin-biotin bond. filter Related Articles: Radionuclide uptake in tumour cells, Extracorporal elimination FIGURE P.89 Illustration of the primary collimator position with Further Reading: Carlsson, J., E. F. Aronsson, S.-O. Hietala, respect to the target and flattening filter. T. Stigbrand and J. Tennvall. 2003. Tumour therapy with radionu- clides: Assessment of progress and problems. Radiother. Oncol. 66(1):107–117. conical-shaped piece of tungsten as illustrated in Figure P.89. The primary collimator defines the largest circular field size available. Primary barrier Further modifications to the beam, for example flattening, further (Radiotherapy) The design of a radiotherapy treatment depends beam shaping, etc. are carried out prior to it exiting the linac head. P on adequate shielding if the radiation exposure to those outside Related Articles: Linear accelerator, Collimation, Collimator, of the room is to be kept below the regulatory requirements. That Treatment head part of the room that the primary beam falls on directly is called the primary barrier and will be the most heavily shielded part of Primary colour the room (e.g. for a 10 MeV linear accelerator this will be of the (General) See RGB (red, green, blue) order of 2.5 m of regular concrete). Outside of this is a region where lower intensity radiation aris- Primary display ing from scattered or leakage radiation falls needs less in the way (Diagnostic Radiology) See Digital display of shielding and the barrier here is known as the secondary bar- rier. The thickness of the wall in this region is of the order of Primary radiation 1–1.5 m depending on beam energy (Figure P.88). (Radiation Protection) The primary radiation beam is the Related Article: Maze x-ray beam directed from the x-ray tube towards the patient. Interactions within the patient lead to scatter and attenuation of Primary collimator the primary beam. The scattered x-rays are referred to as second- (Radiotherapy) X-rays are produced in all directions whenever ary radiation. the electron beam impacts on the target in the head of a linac. To provide a useable beam some form of collimation is required Primary standard to attenuate those x-rays passing through the material. Typically (Radiation Protection) The primary standard for exposure mea- the thickness is such that the leakage radiation is approximately surement is a free air ionisation chamber. The x-ray beam or 0.1% of the open field value. The primary collimator is typically a gamma radiation entering the ionisation chamber passing through air interacts (photoelectric effect, Compton effect, pair produc- tion) with its molecules. As a result of these interactions electrons Primary barrier and ions are produced and electrical current appears between electrodes. The flow of current is proportional to the exposure Secondary barrier rate. The total charge is proportional to the overall exposure. If we assume that kerma in air (Ka) does not differ from absorbed dose ~2.5 m Concrete 2350 kg/m3 in air DA (Gy), the conversion of exposure in air DE (C/kg) to DA (Gy) can be expressed as Treatment machine DA (J/kg) = DE (C/kg)´W (J/C) where W(J/C) ≈ 34 J/C, the energy to form one ion pair in air. Then the primary standard can be used for absorbed dose (or ~1.5 m dose rate) measurement too. Maze entrance The accuracy of measurement depends on the electric field distribution (field lines should be normal to the electrodes), and the pressure and temperature of air in the chamber. If the mea- surement is not performed at STP (T0 = 273.15 K, P0 = 1.013 × 105 Pa) the corrections must be applied. Abbreviations: STP = Standard temperature and pressure and kerma = Kinetic energy released per mass. FIGURE P.88 Treatment room showing secondary and primary barri- Related Articles: Compton effect, Dose, Ionisation chamber, ers and maze corridor. Pair production, Photoelectric effect Primary tumour 740 Projected range Primary tumour the need for health, prevalence of diseases, comparison between (Radiotherapy) A term used to describe the original, or first, different technologies or services, the presence of similar tech- tumour in the body. Cancer cells from a primary tumour may nologies in the area, etc. spread to other parts of the body and form new, or secondary, Procurement often makes use of bidding or tendering pro- tumours. This is called metastasis. These secondary tumours are cesses. This is very common for public structures, but also for the same type of cancer as the primary tumour. private ones with medium-high dimensions. The procurement process, often supported by dedicated soft- Pristine Bragg peak ware, is concluded with a purchasing order or a contract. (Radiotherapy) Protons are positively charged particles and have Related Articles: Bidding, Tendering process, Disposal mass. Therefore, they behave very differently to photons when Further Readings: Iadanza, E. 2019. Clinical Engineering they interact with the medium. Handbook, 2nd edn., Academic Press, Elsevier, ISBN: Protons lose energy mainly due to inelastic Coulomb interac- 9780128134672; Miniati, R., E. Iadanza and F. Dori. 2016. tions with atomic electrons, causing ionisation and excitation of Clinical Engineering (from Devices to Systems), Academic atoms in the medium. The linear energy transfer (LET) of a proton Press, Elsevier, ISBN 9780128037676; Willson, K., K. Ison and or average rate of energy loss of a proton per unit of path length S. Tabakov. 2014. Medical Equipment Management, Taylor & is directly proportional to the square of its charge and inversely Francis Group, Abingdon, ISBN: 9780429130373. proportional to the square of its velocity. The LET becomes maxi- mum as the proton comes to the end of its range. As a result of P Production of radiopharmaceuticals this, protons deposit their maximum dose of radiation at the point (Nuclear Medicine) The preparation process whereby a radioiso- in the medium at which they stop. This sharp radiation dose depo- tope is labelled to a targeting agent. sition is called the Bragg peak (Khan and Gibbons, 2014). The radiopharmaceuticals typically used clinically can be However, the sharp Bragg peak of a mono-energetic proton divided into four groups, where each group represents a different beam, called the pristine peak, is too narrow to cover most clini- labelling technique: cal tumours. The Bragg peak can be extended by combining sev- eral mono-energetic proton beams to generate a spread-out Bragg 1. Solutions ready for dispensation peak (SOBP), which can cover the proximal and distal end of the 2. Prepared kits that only require an addition of for tumour (Horton and Eaton, 2017). instance 99mTc Related Articles: Bragg peak 3. Using prepared kits and an extra effort, for example Further Readings: Horton, P. and D. Eaton. 2017. Design and heating Shielding of Radiotherapy Treatment Facilities, IPEM Report 4. Advanced preparations that involve biological material 75, 2nd edn., IOP Publishing; Khan, F. M. and J. P. Gibbons. 2014. Khan’s the Physics of Radiation Therapy, 5th edn., Wolters Solution ready for dispensation refers to solutions produced outside Kluwer Health. the department, for example purchased from an external manu- facturer or prepared by a dispenser at a pharmacy prior to deliv- Probability of cell survival ery. There are two basic labelling techniques: ion exchange and (Radiotherapy) The probability that a cell will survive a dose of introduction of a foreign radioisotope into a molecule. The former radiation generally depends on its radiosensitivity and the mag- technique involves an exchange from a naturally occurring ion nitude of the dose it receives. This may be obtained from cell to a radioactive ion of the same element. Radiopharmaceuticals survival curves where the probability of survival is given by the prepared this way will have identical bio-kinetic properties as surviving fraction as a function of absorbed dose. The linear qua- the original radiopharmaceutical. The latter method is used for dratic model describes this function. most 99mTc labelling processes. The downside to this is that, due to Related Articles: Cell survival, Cell survival curve, Linear the introduction of 99mTc, the final radiopharmaceutical will have quadratic (LQ) model, Radiosensitivity, Surviving fraction. different bio-kinetics compared to the radiopharmaceutical it is intended to imitate. Probability of complications (Radiotherapy) Radiation treatment inevitably affects normal Production rate of radioactivity tissue and so may cause radiation-induced adverse effects. It is (Nuclear Medicine) See Activation formula generally the case that the total dose that can be tolerated depends on the volume of tissue irradiated – the dose–volume effect. Programmed radiographic technique However, the tissue architecture is also thought to be important in (Diagnostic Radiology) See Automatic exposure control determining the tolerance dose for partial organ irradiation (see articles on Parallel organs and Serial organs). Progression-free survival Several radiobiological models have been proposed that relate (Radiotherapy) See Clinical trial endpoints the probability of normal tissue complications to the dose distri- bution. For further information see the article on Normal tissue Projected range complication probability. (Nuclear Medicine) The projected range of a given type of a Related Articles: Adverse effects, Normal tissue complication charged particle, with a given energy, in an absorbing medium probability, Parallel organs, Serial organs, Tolerance can be defined as the expectation value of the farthest depth of penetration in the particle’s initial direction. In contrast to pro- Procurement jected range, the CSDA (continuous slowing down approxima- (General) The process of acquiring goods or services. It starts tion) range describes the true track pattern from starting energy with evaluations and

assessments of the specific setting, such as down to rest (Figure P.90). Projectile 741 Propagation-based imaging In many cases, the daughter nuclei or the reaction products will be in excited states after the decay or reaction has taken place. If a b those excited states have very short half-lives, often defined as less than 10−9 s, they are referred to as having a ‘prompt’ half- life. Metastable excited states, or nuclear isomers, are defined as those that have half-lives longer than 10−9 s. Excited states with a ‘prompt’ half-life will decay back to the ground state or a lower P nuclear state and the decay can take place through several pro- R cesses such as gamma emission or internal conversion. Therefore, if a nuclear excited state with a ‘prompt’ half-life FIGURE P.90 CSDA range describes the total track from a to b, mean- decays via gamma emission, this is referred to as ‘prompt’ gammas. while projected range describes the farthest depth PR in the absorbing medium. Many radionuclides used in nuclear medicine emit prompt gammas after their decay. For example, iodine-131 decays in 89.9% of the decays to an excited state of xenon-131, which has Further Reading: Frank, H. 1983. Radionuclides Production, an excitation energy of 364.5 keV. This excited state promptly Vol. 1, CRC Press, Boca Raton, FL, pp. 93–94. decays with a half-life of several nanoseconds to the ground state of xenon-131 under the emission of the characteristic 364.5 keV gamma ray. Projectile Related Articles: Gamma radiation, Radioactive decay (Magnetic Resonance) Ferromagnetic metal objects in the pres- P Further Reading: NNDC. Brookhaven National Laboratory. ence of the very strong static magnetic fields used in MRI (i.e. www .nndc .bnl .gov /nudat2/, last accessed 31 October 2019. up to 60,000 times the earth magnetic field at 3 T) act as projec- tiles. At reasonable distances from the magnets the field strength falls off according to the dipole approximation at about 1/r3. Long PROMs (Patient-reported outcome measures) metallic objects show a considerable torque when placed in the (Radiotherapy) See Patient-reported outcome measures (PROMs) static magnetic field and, after the subsequent rotation, their induced dipole moment is parallel to the applied magnetic field. Prone Thus, they experience an attractive force given by F = M · |gradB|, (General) There are a series of terms used to describe the position where M and B are, respectively, the magnitude of the induced of an individual when undertaking different imaging examination. dipole and the magnetic field. In case of a spherical object it has ‘Prone’ means lying on the front the smallest induced dipole moment and, therefore, the minimum See also Patient position force per unit mass of ferromagnetic material is exerted. The attractive force varies inversely with the seventh power of dis- Propagation tance and this results in a sudden increase in the force exercised (Ultrasound) Propagation describes the passage of ultrasound on a ferromagnetic object brought near the magnet. The force per through tissue. Propagation parameters including speed of sound unit of mass varies inversely with the fourth power of the dis- and attenuation are fundamental to the design and performance of tance along the axis of the magnet. Accidents where ferromag- medical ultrasound devices. netic objects are attracted to the centre of the magnet may result Related Articles: Speed of sound, Attenuation in injury and death of the patient and, therefore, ferromagnetic objects and devices are prohibited in proximity to the MRI scan- Propagation-based imaging ner. MR safety is based mainly on the construction of physical (Diagnostic Radiology) Propagation-based imaging (PBI), also barriers, metal detectors and the use of strict administrative rules referred to as free-space propagation imaging or in-line hologra- to prohibit accidental introduction of ferromagnetic objects into phy, is the simplest phase-contrast imaging technique, as it only the magnet room. requires to put some propagation distance between the sample and detector without using any optical element. On the contrary, Prompt gamma PBI has stringent requirements on the spatial coherence of the (Radiotherapy) Nuclei of some elements absorb neutrons, becom- x-ray source, thus, most of its applications are so far limited to ing transformed into isotopes of higher mass number. Prompt synchrotron radiation facilities or low-power microfocal sources. gamma rays are emitted during or immediately after such fission Within ray-tracing approximation, which is well suited to events. High-resolution gamma-spectrometers allow qualitative describing most phase-contrast imaging applications, the intensity identification and quantitative analysis of neutron capturing ele- detected at some propagation distance after the sample by using ments present within the sample (prompt-gamma activation anal- PBI is: ysis, PGAA). As prompt gamma emission is strictly correlated to ion path in matter, prompt gamma analysis holds potential for dl in-vivo range verification in hadron therapy, potentially reducing I(x, y) = é I 1 2 0 ê - Ñ F ù (x, y) T (x, y) ë 2p ûúrange uncertainties in treatment planning. Related Articles: In vivo range verification Where I0 is the intensity of the impinging x-ray beam, d the propa- gation distance, λ the radiation wavelength, ∇2 the Laplace opera- Prompt gammas tor on the object plane x-y, Φ the phase shift due to the sample (Nuclear Medicine) Prompt gammas are gamma rays that are and T the transmission function accounting for the attenuation of emitted within a very short time frame, typically defined as less the sample. than 10−9 s, after, for example, a radionuclide has decayed or after This equation fully describes how both attenuation and phase- a nuclear reaction has taken place. contrast mechanisms come into play: the Laplacian of the phase Propagation speed 742 P roportional counter P FIGURE P.91 Simulations of conventional attenuation based (top left) and propagation-based phase contrast (top right) images of a plastic wire. On the bottom, the corresponding intensity profiles matching the theoretical prediction of the previous equation. shift is significantly different from zero across sharp interfaces, preventative treatment. In the context of radiotherapy, it usually producing a strong phase-contrast signal (i.e. the edge-enhance- refers to the prescription of a radiation treatment to prevent pos- ment effect), conversely, within smooth regions of the sample, sible metastatic spread or to treat uninvolved nodes. the Laplacian vanishes, and only absorption contrast is observed One of the most common applications is prophylactic cranial (see Figure P.91). Of note, within the validity condition of the irradiation (PCI) to prevent brain metastases. This is predomi- ray-tracing model, the phase signal is linearly dependent on the nantly used in patient cohorts who have extensive primary small propagation distance, which can be regulated to obtain different cell lung cancer (SCLC). SCLC patients who respond well to their magnitudes of phase contrast. primary treatment, usually chemo-radiotherapy, are at less risk of Either by acquiring images at multiple distances or by insert- thoracic recurrence and brain metastasis becomes one of the main ing suitable approximations on the object composition, the previ- types of relapse (2-year cumulative risk >50%). A series of studies ous equation can be inverted to yield a map of the phase shift (Φ) demonstrated that PCI not only decreased the incidence of devel- introduced by the sample: such procedure falls under the name of oping brain metastases but also prolonged overall survival (Yin et phase-retrieval. al., 2019). However, although these findings resulted in changes Thanks to the relative simplicity and robustness of its set- in clinical practice, there is still a debate into if PCI should be up, PBI is the only phase-contrast technique that has been so far used, or if routine brain MRI would be a suitable alternative (Yin applied in a human clinical trial in the field of mammographic et al., 2019). imaging. The prophylactic irradiation of uninvolved nodes is usually Related Articles: Phase-contrast imaging referred to as elective nodal irradiation (ENI). The role of lymph Further Readings: Brombal, L. et al. 2019. Monochromatic node irradiation remains unclear, with randomised trails still try- propagation-based phase-contrast microscale computed-tomog- ing to address the value of ENI. Example treatment sites of ENI raphy system with a rotating-anode source. Phys. Rev. Appl. are being investigated, but include head and neck, non-small cell 11(3):034004; Burvall, A. et al. 2011. Phase retrieval in X-ray lung and anal cancers. phase-contrast imaging suitable for tomography. Opt. Express Related Articles: Metastasis 19(11):10359–10376; Castelli, Edoardo et al. 2011. Mammography Further Reading: Yin et al. 2019. Prophylactic cranial irra- with synchrotron radiation: First clinical experience with phase- diation in small cell lung cancer: A systematic review and meta- detection technique. Radiology 259(3):684–694; Endrizzi, M. analysis. BMC Cancer. https :/ /do i .org /10 .1 186 /s 12885 -018- 5251- 3. 2018. X-ray phase-contrast imaging. Nucl. Instrum. Methods Phys. Res. Sect. A 878:88–98. Proportional counter Propagation speed (Radiation Protection) The proportional counter is a gas detec- (Ultrasound) See Speed of sound tor which can operate as a sealed counter with window or a gas flow windowless tube. The voltage applied to the counter is larger Prophylactic irradiation than that in an ionisation chamber (Figure P.92). The primary (Radiotherapy) Prophylactic, which is derived from the Greek ion pairs, created during the interaction of ionising radiation for ‘on advance guard’, is a term used within healthcare for within the gas and then accelerated by the electric field, create the Protective earth terminal 743 Protective earth terminal e d c b a High voltage FIGURE P.92 Dependence of the current (collected charge) on the high voltage value: (a) region of recombination; (b) ionisation chamber region; (c) proportional counter region; (d) limited proportionality region; (e) Geiger–Müeller counter region. secondary ionisation process. The gas multiplication M occurs as a Townsend avalanche. P The total charge Q created by n0 primary ion pairs is Q = n0 ´ e ´ M FIGURE P.94 Box proportional counters, side-windows, single: large type PXAr (Kr, Xe) Be38 × 76, small type PXAr (Kr, Xe) Be19 × 38 where and double: large type PX2Ar Be38 × 76, different dimensions of cath- × ode body (38 × 76 mm or 19 × 38 mm). (Photo courtesy of Dr. Tadeusz e = 1.6 10−19C Kowalski.) M is the gas multiplication (≈105) The electric field of about 106 V/m is required which for HV of 2 kV can be achieved using a cylindrical geometry (Figure P.93). The anode is a thin wire placed inside a cylindrical or rectangular tube which is a cathode. The pulse amplitude is proportional to the number of primary ion pairs, i.e. to the energy of radiation. It depends on the kind of radiation. The proportional counters are used to detect x-rays, gamma rays of low energy, α and β− particles, neutrons (detector filled with BF3). In Figure P.94 through P.99 different types of proportional counters are shown. Related Articles: Geiger–Müller (GM) counters, Ionisation chamber Further Readings: Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, pp. 159–197; Hobbie, R. K. 1997. Intermediate Physics for Medicine and Biology, 3rd edn., Springer-Verlag, New York, pp. 424–425; Dendy, P. P. and B. Heaton. 1999. Physics for Diagnostic FIGURE P.95 Box proportional counters, side-windows, single and Radiology. 2nd edn., Institute of Physics Publishing, Ltd., Bristol, double. (Photo courtesy of Dr. Tadeusz Kowalski, AGH University of UK, p. 139. Science and Technology.) Protective earth terminal (General) See Grounding K A R – + FIGURE P.93 Scheme of a proportional counter (A – anode; K FIGURE P.96 Cylindrical, end-window (mica) proportional counter. – cathode). (Photo courtesy of Dr. Tadeusz Kowalski.) Log A Protective grounding 744 Proton arc therapy Protective grounding (General) See Grounding Protocol (Nuclear Medicine) A protocol is a set of guidelines or rules. In nuclear medicine protocols are used to guarantee qualitative examinations and treatments. For example, a protocol can contain guidelines for examination procedures or for quality control of imaging systems. Proton (General) Symbol P Mass 1.672 × 10−27 kg (938.27 MeV/c2) FIGURE P.97 Small box proportional counters single and double, side Charge 1.602 × 10−19 C P window. (Photo courtesy of Dr. Tadeusz Kowalski.) Nuclear spin ½ Gyromagnetic ratio 2.67 × 108 rad s−1 T−1 Radiation weighting factor 5 Protons are stable subatomic particles that were discovered in 1918 by Ernest Rutherford. They are composed of two up quarks and one down quark which are held together by the strong nuclear force. Alongside neutrons, protons form the constituent parts of the atomic nucleus. In magnetic resonance imaging

the hydrogen nucleus is often referred to as a proton. Medical Applications: Proton therapy – Due to their Bragg peak, proton beams can offer therapeutic advantage over photon beams in certain clinical situations (most notably in the treatment of anatomically awkward tumours). Proton radiotherapy requires a large initial investment (for a cyclotron or similar accelerator), but the technique is rapidly gaining ground worldwide. Magnetic resonance imaging – The magnetic moment of the protons in water molecules in the human body is exploited in Magnetic resonance imaging (MRI). Related Articles: Atom, Magnetic resonance imaging, Proton density, Proton therapy, Charged particle therapy, Bragg peak, Cyclotron, Radiation weighting factor FIGURE P.98 Straw proportional counters. (Photo courtesy of Dr. Proton arc therapy Tadeusz Kowalski.) (Radiotherapy) The concept of proton arc therapy involves con- tinuous proton beam delivery as a gantry rotates around a patient. Proton arc therapy is not in clinical use at the time of writing (June 2019). Proton arc therapy first appeared in the literature in 1997 (Sandison et al., 1997). A few more studies were reported from 2012 to 2016 (Zhang et al., 2013; Seco et al., 2013; Ding et al., 2016; Sanchez-Parcerisa et al., 2016). Related articles: Arc therapy, proton therapy Further Readings: Ding, X. et al. 2016. Spot-scanning proton arc (SPArc) therapy: the first robust and delivery-efficient spot- scanning proton arc therapy. Int. J. Radiat. Oncol. Biol. Phys. 96(5):1107–1116; Sanchez-Parcerisa, D. et al. Range optimization for mono-and bi-energetic proton modulated arc therapy with pen- cil beam scanning. Phys. Med. Biol. 61(21):N565; Sandison, G. A. et al. 1997. Phantom assessment of lung dose from proton arc therapy. Int. J. Radiat. Oncol. Biol. Phys. 38(4):891–897; Seco, J. et al. Proton arc reduces range uncertainty effects and improves FIGURE P.99 Pill-box (pancake) proportional counter. (Photo courtesy conformality compared with photon volumetric modulated arc of Dr. Tadeusz Kowalski.) therapy in stereotactic body radiation therapy for non-small cell Proton CT (pCT) 745 Proximal inversion with a control for off-resonance lung cancer. Int. J. Radiat. Oncol. Biol. Phys. 87(1):188–194; 100 Zhang, R. et al. 2013. Comparison of risk of radiogenic second cancer following photon and proton craniospinal irradiation for a pediatric medulloblastoma patient. Phys. Med. Biol. 58(4):807. Photon beam 6 MV Proton CT (pCT) (Radiotherapy) To date, calibration curves have been used to 50 Modified proton beam transform patient x-ray CT scans to the proton stopping power 250 MeV maps needed for proton treatment planning. This process, known as stoichiometric calibration, introduces range-uncertainties Native proton beam which must be considered in the treatment planning process. 250 MeV Proton CT (pCT) has been proposed as a means to reduce these range uncertainties by imaging and treating using the same par- 00 10 20 30 ticle. Measurements of proton energy/trajectory before and after Depth in tissue (cm) a beam traverses an object can be used to reconstruct a 3D image of the object where the voxel values represent proton stopping FIGURE P.100 Illustration of the Bragg peak (native proton beam) powers. However, protons undergo many small-angle deflections and spreadout Bragg peak (modified proton beam) for a 250 MeV proton as they pass through matter, primarily due to multiple Coulomb beam in comparison with that for a 6 MV photon beam). scattering. Statistical models must be applied to consider this in P the pCT reconstruction process, leading to lower spatial reso- lution than x-ray CT. Additionally, pCT requires substantially the dose deposition of protons in tissue displays a Bragg peak higher beam energies than clinical proton therapy; specialised with the majority of the energy deposited in the last few mil- accelerators must be considered. limetres of each proton’s path. Range and energy modulation Related Articles: Stopping power, Cyclinac, Proton radiogra- of the proton beam allows the generation of a spread-out Bragg phy, stoichiometric calibration, Treatment planning system, CT peak (SOBP) to target the tumour at all depths. The depth dose reconstruction, In vivo range verification characteristics of a Bragg peak and SOBP are compared with a 6 MV x-ray depth dose curve in Figure P.100. One of the major Proton density advantages of a proton beam is the absence of an ‘exit dose’. (Magnetic Resonance) One of the parameters modulating the However, range uncertainties (range straggling) limit the accu- amplitude of the MR signal is the number of the protons (or spins) racy and precision with which a proton beam can be stopped in within the volume of the sample. Certain tissues have more pro- front of an organ at risk. tons per unit volume than other tissues (e.g. water < fat). Voxels Related Articles: Charged particle therapy, Hadron therapy, with high proton density or hydrogen concentration appear bright. Ion therapy, Neutron therapy, Heavy particle beams, Spread-out MR signal amplitude depends on the presence or absence of pro- Bragg peak (SOBP), Range shifter, Intensity modulated proton tons (hydrogen nuclei) and is also sensitive to the environment of therapy (IMPT), Passive beam scattering, Energy selection sys- the hydrogen nuclei. In fact it is not only the absolute number of tem, Pencil beam scanning (PBS), Range straggling, Relative protons in the tissue that is important but also the number of pro- biological effectiveness, Cyclotron, Synchrotron, Single room tons within the volume that are sufficiently mobile to be able to particle therapy systems, Bragg peak, Magnetic beam steering, line up their spins with the external magnetic field. Tightly bound Degrader, Range straggling, In vivo range verification, Organ at hydrogen creates a weak MR signal. This effect explains why the risk, Stoichiometric calibration, Proton arc therapy, Pion therapy, cortical bone appears black on an MR image since it does not emit Carbon ion therapy much MR signal. This is not due to the absence of hydrogen but Hyperlinks: http://en .wikipedia .org /wiki /Image :BraggPeak because the hydrogen nuclei are tightly bound to the molecule. On .png the other hand, medullary bone is visible only indirectly due to the fat located in the space between the trabeculae and in marrow Proximal cavities. Consequently, on pure proton density images (i.e. very (General) Near, closer to the origin (for example, the proximal short TE, very long TR) fat and liquids appear brightest, soft tis- end of the femur joins with the pelvic bone). sue medium grey, and bone dark. See also article on Anatomical relationships Proton radiography Proximal inversion with a control for off- (Radiotherapy) Radiographic (2D) imaging using proton beams. resonance effects (PICORE) For more information, see the Proton CT (pCT) article, which (Magnetic Resonance) The PICORE arterial spin labelling covers tomographic (3D) implementation. (ASL) preparation technique is a derivative of the EPISTAR Related Articles: Proton CT (pCT) concept. The labelling experiment in PICORE is equal to that of EPISTAR. In the control experiment, however, an off-resonance Proton therapy and non-selective pulse, with a frequency offset identical to that (Radiotherapy) In proton therapy, high energy proton beams of of the labelling experiment, is employed (while a slab-selective energy ~250 MeV are used to treat deep-seated tumours. The distal inversion pulse is applied in EPISTAR). Assuming axial beams are usually produced using accelerators such as cyclo- imaging of the brain, with known inflow direction, the advantage trons or synchrotrons. Typically, lateral `dose-painting’ of of PICORE is a minimal contribution to the vascular signal from the tumour is achieved by scanning the beam in a raster pat- draining veins (i.e. from inflowing distal spins), with maintained tern using a magnetic scanning system. As for other hadrons, similar cancellation of magnetisation transfer effects. EPISTAR Dose (%) Pseudo CT 746 PSIF (a time reversed FISP) FIGURE P.101 Examples of pseudo CT generation. Figure is adapted from original (https :/ /do i .org /10 .1 016 /j .ijro bp .20 18 .10 .002). P should be considered in cases where eddy currents from slice- effect of frequency encoding: If the echoes are lined up next to selective gradients are significant. each other in order of phase encoding gradient amplitude, they Related Articles: Perfusion imaging, Arterial spin labelling, form a ‘pseudoecho’ along the phase encoding axis that resem- FAIR, EPISTAR, QUIPSS – QUIPSS II – Q2TIPS bles the echo that forms during frequency encoding, and this Further Reading: Wong, E. C., Buxton, R. B., Frank, L. R. ‘pseudoecho’ appears to be composed of signals with different 1997. Implementation of quantitative perfusion imaging tech- frequencies originating from different distances along the phase niques for functional brain mapping using pulsed arterial spin encoding axis. labeling., NMR Biomed. 10:237–249. Related Articles: B0 gradients, Frequency encoding, Phase encoding Pseudo CT (Radiotherapy) A pseudo CT is a way of generating electron den- PSF (point spread function) sity information from an MRI image used in MR-only treatment (General) See Point spread function (PSF) planning. CT values are assigned to areas in the MRI image based on the MR signal. There are different methods of pseudo CT gen- eration. The bulk density correction method assigns a CT value PSIF (a time reversed FISP) of either water, air or bone to areas of known MR signal strength (Magnetic Resonance) PSIF, the time reversed FISP sequence, which correspond to the four materials commonly found in the belongs to the family of SSFP sequences, a gradient echo body. sequence combing short repetitions times and low flip angles. The atlas-based method uses machine learning and a library However, PSIF is designed to employ the SSFP-echo instead and, of non-rigidly registered MRI and CT images to form the pseudo maybe, it is not correct to call it a GRE sequence, since it actually CT. The patient MRI is compared to the library of MR images, behaves more like a spin echo sequence. and the pseudo CT is generated by fusing the registered CTs in the The diagram in Figure P.102 shows the timings of a PSIF atlas and applying the fusion map of the patient MR to the atlas sequence. By carefully examining the scheme it can be seen that of MR images. it is a time reversed FISP. Note that in the PSIF sequence, the TE There are other methods of pseudo CT generation that are still is longer than the TR. being researched. Figure P.101 shows what an atlas-based method The SSFP-echo sequence carries a larger portion of T2- and bulk density correction method of pseudo CT generation weighted signal, and a combination of short TRs and medium makes. The pseudo CT is then used for dose calculation. The method of pseudo CT generation influences the accuracy of the dose cal- culation. It is also important to note that an atlas-based method RF or a bulk density correction method will not be an accurate CT representation of the patient, but a best estimate. Related Articles: Deformable image registration, MR only Gs planning Gp Pseudoecho (Magnetic Resonance) In conventional 3D MRI, gradient rever- sal is used to generate an echo which is subjected to Fourier transformation to yield a projection along the frequency encod- Gf ing gradient direction. The process is repeated many times with Echo time (TE) a different phase encoding gradient amplitude applied along the orthogonal direction so that the phase of the signal from magne- Repetition time (TR) tisation at a given position along this direction is incremented. The phase evolution across the set of acquired echoes mimics the FIGURE P.102 Schematic illustration of a SSFP-echo sequence, PSIF. Public exposure 747 Pulse flip angles provides images similar to strongly T2-weighted SEs. In SSFP sequences, which consist of several rapidly repeated RF pulses, the transverse magnetisation from moving Vmax PSV spins is spoiled, leading to very low signal from moving spins and signal behaviour similar to the spoiled GRE sequence. If Peak to peak no motion compensation is available, ghosting and signal loss velocity can occur. EDV The SSFP-echo sequence is strongly T * 2 -weighted and has the Vmin potential for fast imaging of lesions with long T2 compared with the surrounding tissue. It is presently not widely used clinically. Acronyms for the SSFP-Echo Sequences: Balanced in one Vmax direction: CE-GRASS (General Electric), CE-FFE-T2 (Philips) PSV and PSIF (Siemens) Peak to Related Articles: Fast imaging with steady state preces- peak sion (FISP), Spin echo (SE), SSFP, Steady state free precession, velocity T EDV 2-weighted Further Readings: Scheffler, K. 1999. A pictorial descrip- tion of steady-states in rapid magnetic resonance imaging. Concepts Magn. Reson. 11(5):291–304; Nitz, W. 2002. Fast and V P min Maximum reverse ultrafast non-echoplanar

MR imaging techniques. Eur. Radiol. velocity 12:2866–2882. FIGURE P.103 Variables used in the pulsatility index. The white bro- Public exposure ken line represents the maximum velocity in the waveform – the time (Radiation Protection) Public exposures are incurred by members averaged mean is taken over a number of complete cardiac cycles. of the public from radiation sources, excluding any occupational or medical exposures and the normal local background radiation but including exposures from authorised sources and practices z B and from intervention situations. 0 See also allowed values in Dose limits. Related Article: Dose limits M Further Reading: IAEA. 1996. International basic safety standards for protection against radiation and for the safety of radiation sources, Safety Series No. 155, International Atomic Energy Agency, Vienna, Austria. θ Pulsatile (Ultrasound) See Pulsatility index (PI) Pulsatility index y (Ultrasound) The pulsatility index is a simple, commonly used B1 measure of arterial waveform shape. The index reflects changes in the presence of a proximal stenosis and peripheral resistance. The pulsatility index (PI) is non-dimensional and independent of the ultrasound beam/blood flow direction angle: V x PI = max - Vmin Vmean FIGURE P.104 When an RF pulse is applied, the longitudinal magneti- The PI applied to two different shaped waveforms is shown in sation M is tilted away from the main magnetic field B0 by a flip angle θ. Figure P.103. The numerator is the total peak-to-peak excursion of the waveform. The mean velocity (Vmean) is the time averaged mean of the maximum instantaneous velocity envelope over a Pulse number of complete cardiac cycles. (Magnetic Resonance) To create MR images of an object a radio- Abbreviations: EDV = End diastolic velocity, Vmin = Minimum frequency (RF) pulse must be applied. This pulse must have a velocity, Vmax = Maximum velocity and PSV = Peak systolic frequency equal to the Larmor frequency of the spin system. velocity. The magnetic field B1 generated by the pulse is rotating in the Related Articles: Resistance index, Peak systolic velocity, transverse (xy) plane and application of the pulse causes the lon- End diastolic velocity gitudinal magnetisation to tilt away from the main magnetic field Further Reading: Gosling, R. G. and D. H. King. 1974. (Figure P.104). The deviation, or flip angle θ. of the longitudi- Continuous wave ultrasound as an alternative and compliment to nal magnetisation from the main magnetic field depends on the x-rays in vascular examination. In Cardiovascular Applications strength and duration of the RF pulse. Directly after the RF pulse of Ultrasound, ed., R. S. Reneman, pp. 266–281, North-Holland is turned off, the longitudinal magnetisation begins to realign Publishing Co., Amsterdam, the Netherlands. with the main magnetic field via relaxation processes. Pulse average intensity (IPA) 748 P ulse inversion The duration and amplitude of the RF pulse determines the B-mode and Doppler imaging and the time averaged intensity of flip angle. Flip angles between 0° and 90° are often used in gradi- the ultrasound in tissue. ent echo pulse sequences and a flip angle of 90°, followed by a Pulse duration can be determined from the pulse intensity inte- 180° pulse (spin inversion), is used in a spin echo pulse sequence. gral where it is defined as 1.25(t2 − t1). t1 is the time at which the The shape (the time-dependant distribution of amplitudes) of pulse intensity integral has reached 10% of its final value and t2 is the RF pulse determines the homogeneity of the magnetisation the time at which it reaches 90% of the final value (Figure P.106). of the spins and the spatial extension of the excitation when a gradient is applied. For example, a sinc (mathematical function Pulse echo defined by sin(x)/x) pulse will ideally excite a slice with a rectan- (Ultrasound) Name of the underlying principle used to generate gular profile. ultrasound images. This is also the same principle for instance Related Articles: Flip angle, Larmor frequency, bats and dolphins use to navigate and locate prey. A short sound Radiofrequency, Relaxation, Spin echo (SE) pulse (for animals we can describe it as a shriek) is emitted and the time until an echo returns is a measure of the distance, pro- Pulse average intensity (IPA) vided that the sound speed is known. Human tissue types actually (Ultrasound) For most diagnostic applications ultrasound is have a range of velocities but the range is fairly small so that an transmitted into the body in a series of pulses whose length and average sound speed of 1540 m/s can be assumed. amplitude envelope will depend on the application. The pulse average intensity is the average intensity of ultra- Pulse generator P sound over the duration of the pulse (Figure P.105). It is derived (Ultrasound) A device that outputs pulses, for instance used to from the pulse-intensity integral and the pulse duration. The trigger a piezo-electric transducer. To emit an ultrasound pulse, a pulse-intensity integral is the time integral of the instantaneous necessary voltage is often in the range of 100 V. To avoid multiple intensity in the pulse: voltage outputs from the transformer, a step-up converter may be used in the circuitry for single element equipments, where the Pulse-intensity integral IPA = power consumption is lower. Pulseduration Pulse inversion Or expressed mathematically, (Ultrasound) Pulse inversion is an imaging modality that capi- talises on non-linear effects to produce images with improved T ò i (t )dt contrast and/or resolution. The development of the pulse inver- I (PA) = 0 sion technique was motivated by the limited spatial resolution T and image contrast that was attained with harmonic imaging for contrast imaging. Harmonic imaging suffers as the harmonic can Related Articles: Intensity, Pulse duration, Duty cycle only be filtered out over a limited band, otherwise the fundamen- tal energy leaks into the detection bandwidth. This means that Pulse duration only relatively narrowband pulses can be used in harmonic imag- (Ultrasound) The pulse duration is the length of time over which ing, and thus a limited spatial resolution is attained. the pulse is transmitted. The pulse duration has important con- The solution to this problem as proposed with pulse inversion sequences for the resolution of the image, the performance of is to transmit not one pulse, but two. When the echoes from the first pulse have been received, a second pulse, which is a phase- inverted copy of the first, is transmitted (Figure P.107). The two scattered pulses from a linear scatterer will produce two signals Pulse 100 amplitude 90 80 70 60 50 40 P2 sum 30 20 10 0 Δt t1 t2 Time FIGURE P.105 Pulse average intensity is obtained by integrating the FIGURE P.106 Measurement of pulse duration from the pulse intensity pulse pressure and dividing it by time. integral. Percentage of final value Pulse pileup 749 P ulse sequence Linear scatterer Non-linear scatterer PET: An increase in the count rate will lead to an increase in random coincidences. These random coincidences will degrade Pulse 1 the spatial resolution and activity quantification. One method to deal with this problem is the delayed window method. In normal image acquisition mode coincidences occur when an event is fol- Pulse 2 lowed by a corresponding event in an opposite detector within a coincidence timing window (CTW). At high count rates two or more events occur in the opposite detectors and in some cases the line of response is misplaced. Sum In the delayed window method, the coincidence timing win- dow is delayed. When an annihilation photon is detected the photon from the corresponding annihilation will arrive within a FIGURE P.107 Principle of pulse inversion. The left column of wave- few nanoseconds at an opposite detector, and if detected it would forms shows the resulting echoes from a linear reflector, and the right be considered a true coincidence. When moving the coincidence column, from a non-linear reflector. The bottom row shows the sum of timing window (∼5 times the width of the window itself, the the echoes for the two cases. CTW width being typically 12 ns) no true coincidences can be registered. Instead, most coincidences registered in such a mode that are mirror variants of each other, and their sum will be zero. are random coincidences. The data from the delayed mode is used If, on the other hand, the scatterer is non-linear, as is the case of as an estimation of the random coincidence rate and it can be sub- P a contrast bubble, the signals will not sum to zero, but leave a tracted from images acquired in the image mode. detectable signal. It can be shown that this corresponds to filtering An obvious drawback is the fact that the coincidences recorded out the fundamental and odd harmonics from the received signal. in the delayed mode are not the same as the ones in image mode. The drawback is a sensitivity to motion artefacts. Variants of this Therefore the subtraction of random coincidences will cause an scheme have been suggested, such as pulse inversion Doppler or increase in statistical noise. differing amplitudes for the two transmissions rather than phase Related Articles: PET, Coincidence timing window, Line of inversion. Pulse inversion techniques are also used in native tissue response, Annihilation coincidence detection harmonic imaging to reduce the influence of fundamental echoes Further Reading: Cherry, S. R., J. A. Sorenson and M. E. from the harmonic image derived through non-linear pulse propa- Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, gation of the ultrasound pulse. Philadelphia, PA, pp. 114, 231–234, 357. Pulse pileup Pulse repetition frequency (PRF) (Nuclear Medicine) Pulse pileup refers to the problem with high (Ultrasound) See Pulse repetition rate count rate detection. At high count rates imaging systems can be incapable of handling the excessive information and separat- Pulse repetition period ing individual pulses. Pileup effects occur in both single photon (Ultrasound) See Pulse repetition rate imaging and PET but the way the effects are manifested is differ- ent in each case. Pulse repetition rate Planar Imaging: Two simultaneously registered events will (Ultrasound) Pulse repetition rate (1/s) or pulse repetition fre- produce a signal where the signal amplitude (proportional to quency is the number of pulses transmitted per second. It can also energy) is the sum of the two individual signals. Using standard be referred to as the inverse of the pulse repetition period (Figure pulse-positioning logic the events will be localised as one event P.108). The diagnostic ultrasound imaging pulse repetition rate is somewhere between the two original events. One solution is to limited by the depth used as this affects the transmit–receive time. have a discriminator circuit that discards all events registering A faster pulse repetition rate gives a higher update frequency of signal amplitude higher or lower than certain thresholds, that is the measured object, which can be relevant to a measurement of a an energy window. But if two photons are Compton-scattered fast developing or moving object. and then simultaneously registered the sum of energy might fall Related Article: Aliasing within the boundaries of the energy window, hence creating a false event. Pulse sequence Modern imaging systems use so-called pulse-tail extrapola- (Magnetic Resonance) The pulse sequence, that is the timing tion circuitry. If two photons are registered with only a small table for the radio-frequency (RF) pulses, magnetic field gradients difference in time then the signal induced by the first interac- tion has not fully decayed before the new interaction is regis- tered, hence the signals overlap. It is possible to use estimation Pulse circuitry to extrapolate the first pulse and later subtract this estimated signal from the registered signal, theoretically leav- ing just the signal contribution from the second event. With this method it is possible to decrease the number of pileup events and Waiting time at the same time the separated events contribute to the image information. A second approach is available in modern cameras. If the sig- T nal distribution of two simultaneously registered events does not overlap they can be separated and read independently from one FIGURE P.108 T = Pulse repetition period. (Courtesy of EMIT project, another. www .emerald2 .eu) Pulse sequence 750 Pulse sequence and signal sampling periods, is one of the most essential links gradient). The duration and amplitude of the refocusing gradi- between MR physics and clinical MRI. The possibility to change ent should be designed so that the area

under the gradient equals pulse sequence codes and thereby to significantly alter image con- the area covered by the original slice-selective gradient pulse, trast is unique for MR as a diagnostic imaging tool, and during the counted from the centre of the RF pulse. twenty years of MRI in clinical practice that have elapsed, a wide The phase-encoding gradient is switched on after the slice variety of pulse sequences have been suggested and tested. selection and this gradient is used to create spatial resolution by In order to fully characterise a pulse sequence, normally a encoding of spin phases in one of the directions in the imaging graph consisting of five diagrams is used (Figure P.109 through plane (Figure P.110). In conventional MRI, the pulse sequence is P.111). These diagrams describe the time frames for execution of repeated with different amplitudes of the phase encoding gradi- RF pulses, gradient pulses and for signal detection (signal sam- ent until a sufficient resolution is obtained in the phase-encoding pling period/ADC interval). direction, for example 128–256 times. Since the phase-encoding RF pulses and slice selection (Figure P.109): The RF pulse gradient utilises the spin phases to create spatial encoding, it is transmits energy to the spin system. The energy exchange flips the not possible to prevent spin dephasing in this direction, and there- spin population’s longitudinal magnetisation vector Mz towards fore the signal will be reduced when large amplitudes of the phase the transverse xy plane, and the energy transferred to the system encoding gradient are used. is characterised by the so-called flip angle. The RF pulse is, in The frequency encoding (readout) gradient is used to create most imaging sequences, combined with the slice selective gra- spatial resolution in the other of the two orthogonal imaging plane dient pulse, in order to excite only spins within a specific slice directions (Figure P.111). This gradient is used to create a spa- P or slab. The slice thickness can be adjusted by changing the RF tial frequency dependence of the signal. When the readout gradi- bandwidth (the selected frequency interval) and/or the amplitude ent is on, a specific frequency interval in the readout direction is of the slice selective gradient. Slice-selective RF pulses typically defined (readout bandwidth). Typical bandwidth values are a few are switched on for a few milliseconds (ms) and their shape is hundred Hertz (Hz) per picture element (pixel). From the math- designed to create, ideally, a rectangular slice profile and homo- ematics behind the imaging process, increased duration of the geneous flip angles over the whole width of the slice. Gaussian readout interval causes a decrease in bandwidth and an increase and, more frequently, sinc pulses are used for this purpose. For the latter pulse type, the slice profile (with a penalty in pulse duration) is improved by adding more lobes to the sinc function. Normally, the slice shape is also improved by using smaller flip angles. By combination of physical gradients in three orthogonal directions, the slice orientation can be freely selected. For maxi- mum signal, it is of great importance to refocus the slice selective Gphase gradient: When the slice selection gradient is turned on, spins at different positions in the slice direction obtain different resonance k-space frequencies ranging over the whole RF bandwidth. Hence, when the RF power and the slice selective gradient are turned off, spins Time within the slice at different position in the slice direction have dif- ferent phase. Since the signal from each volume element (voxel) is obtained by vector summation, so-called spin dephasing and signal loss will be the consequence. In order to compensate for this signal dephasing, a second lobe of the slice-selective gradi- ent with opposite polarity is added after the RF pulse (refocusing FIGURE P.110 Phase-encoding gradient is switched on after the slice selection and this gradient is used to create spatial resolution by encoding of spin phases in one of the two (orthogonal) directions in the imaging plane. RF Time Gslice Slice selection Gread k-space Time Refocusing Time ADC Time Rise time Refocusing FIGURE P.111 Frequency encoding gradient is used to create spatial FIGURE P.109 RF pulse is in most imaging sequences combined with resolution in one of the two orthogonal imaging plane directions. A refo- the slice selective gradient pulse, in order to excite only spins within a cusing procedure is undertaken prior to signal readout to suppress signal specific slice or slab. loss due to dephasing. Pulse sequence optimisation 751 Pulsed dose rate (PDR) in SNR. Reduction of bandwidth, however, enhances chemical After Fourier transformation of that relation, j(t) can be found as shift effects and susceptibility-related image artefacts, for exam- ple from metallic implants. ì FT é j t iFT ëw t ù ( ) ï ( ) ü = ï í û The readout gradient is only used to obtain frequency encod- ý ing and hence the phase differences caused by the gradient cre- îï FT éëh(t )ùû þï ate an undesired signal loss. Therefore, a refocusing procedure is applied also for the frequency encoding gradient, although where the spin refocusing in this case must be made prior to the actual iFT{ } represents inverse Fourier transformation encoding period. This means that the spins are defocused before FT[ ] Fourier transformation the signal-sampling period so that full gradient refocusing is achieved at the centre of the signal-sampling period. The encod- Alternatively it can be found by minimising the quantity (h(t) ing period is accompanied by signal sampling (the ADC inter- * j(t) − w(t))2 summed for all time samples. val). After the end of the signal sampling period, the sequence is repeated with a new value of the phase-encoding gradient until Pulsed cine a sufficient amount of k-space lines have been sampled. After (Diagnostic Radiology) Pulsed cine mode of operation uses a completed sampling, a 2D Fourier transform (FT) is used to short x-ray exposures for each cine frame. During this mode the decrypt or decode the raw signal data. x-ray beam switches rapidly on–off, thus producing sequence of It should be noted that the procedure described previously images (most often with rate of 25–30 fps). Using pulsed cine only covers the basic strategy of pulse sequence design, and mode requires special design of the x-ray generator (most often P that additional RF as well as gradient pulses are frequently used associated with use of grid controlled x-ray tube). The exposures for, e.g. fat saturation, saturation bands, flow compensation and have to be synchronised with the cine camera shutter. Often such intended spoiling of magnetisation. More extensive overviews of mode can also be synchronised with the ECG of the patient, thus pulse sequence design can be found in Further Reading. filming specific phases of the cardiac cycle (in cardio angiogra- Related Articles: Bandwidth, Flip angle, Frequency encoding, phy). The fact that the x-rays are off between two frames leads to Gradient, RF pulse, Sinc function an overall reduced patient dose, compared with continuous mode Further Reading: Haacke, E. M., R. W. Brown, M. R. cineradiography. Thompson and R. V. Venkatesan. 1999. Magnetic Resonance Related Articles: Grid controlled x-ray tube, Cineradiography Imaging. Physical Principles and Sequence Design, John Wiley & Sons, New York, pp. 781–820. Pulsed dose rate (PDR) (Radiotherapy, Brachytherapy) Pulse sequence optimisation Dose Rates in Brachytherapy: Different dose rates are (Magnetic Resonance) The MR pulse sequence determines used in brachytherapy treatment techniques. The International the characteristics of the reconstructed images, for example, Commission on Radiation Units and Measurements, ICRU, contrast, resolution, signal-to-noise ratio, dimensions, loca- defined these dose rates in its Report No. 38 ‘Dose and tion and orientation in a patient, sensitivity for movement. It is Volume Specification for Reporting Intracavitary Therapy in thus important that the operator optimises the pulse sequence, Gynecology’: that is chooses image parameters (such as repetition time, echo time, inversion time and flip angle) to obtain the desired image 1. Low dose rate, LDR characteristics with as little distortions and artefacts as possible. a. 0.4–2.0 G/h This process may implicitly include changing RF or gradient- b. Traditional technique; 0.5 Gy/h, 60 Gy with treat- pulse design and/or timing, and/or changing the readout method. ment time 5 days Sequence optimisation can be performed using high-level soft- c. Large amount of clinical data ware tools provided by the manufacturer enabling changes in the d. (NOTE: Ultra low dose rate 0.01–0.3 Gy/h) sequence protocol, but it can also be made using specific pro- 2. Medium dose rate, MDR gramming languages (pulse programming tools), often available a. 2–12 Gy/h for research purposes. b. Seldom used Related Articles: Pulse sequence, Signal-to-noise ratio (SNR) 3. High dose rate, HDR a. >12 Gy/h = 0.2 Gy/min Pulse shaping b. Treatment times approximately 5–20 min (compa- (Ultrasound) Pulse shaping describes a predistortion of a trans- rable to external beam therapy) mitted pulse in order for it to appear in a desired way either after c. Clinical data available having propagated a distance, or to appear similar to a pulse 4. Pulsed dose rate, PDR produced with a different source (transducer). For instance pulse a. Mimics LDR, using many small ‘HDR pulses’ dur- shaping can be used to minimise unintentional second harmonic ing a longer treatment time transmission in harmonic imaging systems. A transmitted ultrasound pulse waveform p(t) is the convolu- Example: 1 Pulse per hour during 24 h, 0,5 Gy per pulse given in tion of the excitation waveform e(t) with the transducer impulse 5 min; total dose 12 Gy/day. response function h(t). In order to produce a transmitted pulse The radiobiological effects in the tissues irradiated depend with the desired shape w(t), the transducer must be excited with a on the type of applicator used, on the fractionation scheme and voltage waveform: on both dose and dose rate distributions. As stated in the ICRU Report 38: ‘the clinical experience accumulated with radium tech- j(t), i.e. w(t) = h(t)* j(t). niques cannot be applied to new irradiation conditions without Pulsed laser 752 Pulsed wave Doppler careful consideration’. This includes consideration of both tumour following a pulsed stimulation with a duration of hundreds of effects and effects on normal tissues. nanoseconds and a time gap between the two pulses of a few hun- Abbreviation: ICRU = International Commission on Radiation dred microseconds. The optical stimulation is periodically turned Units and Measurements. on and off and the OSL delayed emission, the afterglow between Related Articles: Brachytherapy, Dose rates in brachytherapy, the pulses, is recorded and added up to the previous value until an see also articles under radiobiology equilibrium value for a POSL signal is obtained. The reader detec- Further Reading: ICRU. 1985. Dose and volume specification tor (PM tube) is blocked for a few tens of microseconds. This is for reporting intracavitary therapy in gynecology. ICRU Report an interval time longer than the duration of the stimulating pulse 38, ICRU, Washington, DC. so that no signal is detected during the stimulation. The stimula- tion and detection of the emitted OSL signal are separated by time Pulsed laser discrimination. The POSL signal could be increased by increasing (Non-Ionising Radiation) See Laser output mode the power of the stimulating laser source and decreasing the pulse width. Short and intense pulses result in larger POSL signals. The Pulsed mode POSL technique has been successfully applied in α-Al2O3:C radia- (Diagnostic Radiology) Pulsed mode of operation is used in some tion dosimetry using 4 kHz frequency pulses and a pulse width of contemporary x-ray fluoroscopic systems. During this mode the 300 ns for stimulation and by blocking the PM tube for 15 μs. x-ray beam switches rapidly on–off, thus producing sequence of It is possible for the OSL readout to extract the necessary sig- images. Each image is recorded in the system memory. Using nal, without exhausting all trapped electrons, and could ensure a P pulsed mode requires special design of the x-ray generator (most consecutive OSL reading. In comparison, the TLD readout nor- often associated with use of grid controlled x-ray tube). The mally has only one reading. resulting images are with rate of 25–30 fps, or more. The fact Related Articles: Optically stimulated luminescence, that the x-rays are off between two frames (acquisitions) leads to Thermoluminescent dosimeter (TLD) an overall reduced patient dose, compared with continuous mode fluoroscopy. Pulsed radiation Pulsed mode fluoroscopy has another advantage, related

(Diagnostic Radiology) See the article on Pulsed cine mode to image quality. Due to the short exposures of each frame the resulting images have less motion artefacts from movement of the Pulsed ultrasound anatomical organs – that is these are sharper (with better resolu- (Ultrasound) Pulsed ultrasound uses the emission of a time-lim- tion and less noise). ited burst of sound, as opposed to continuous wave ultrasound. Related Article: Grid controlled x-ray tube With pulsed ultrasound, range detection of echoes is possible and Further Reading: Beutel, J., H. Kundel, R. van Metter, eds. pulse echo techniques are the basis of B-mode imaging. Pulsed 2000. Handbook of Medical Imaging, Volume 1, Physics and ultrasound is also used for pulsed Doppler, using slightly lon- Psychophysics, SPIE Press, Washington, DC. ger pulses where the phase shift between successively received echoes is sensed. This forms the basis of colour flow imaging and Pulsed mode pulsed wave spectral Doppler in clinical ultrasound scanners. (Ultrasound) Pulsed mode describes a mode of operation where Related Articles: B-mode, Pulsed mode, Pulsed wave Doppler, the ultrasound is transmitted intermittently, that is in bursts or Colour flow imaging pulses, by the transducer. Between each pulse, the echoes from targets within the body are collected. Using the time between the Pulsed wave Doppler transmitted pulse and the received signals, the distance at which (Ultrasound) Pulsed wave Doppler techniques are used for the echoes were generated can be determined: colour flow imaging and spectral Doppler in ultrasound scanners (Figure P.112). By transmitting pulses and measuring phase shifts c ´ t d = 2 where d is the distance of the target from the transducer c is the speed of sound in the tissue t is the round trip time for the burst of ultrasound Example: A transducer is placed on the surface of the body over tissue in which the speed of sound is 1540 m/s. An echo is received 1 × 10−4 s after the transmitted ultrasound pulse. The depth of the structure that generated this echo is (1540 × 10−4)/2 = 0.077 m or 7.7 cm. All ultrasound imaging and almost all other diagnostic appli- cations of ultrasound rely on pulsed mode for their function. Related Articles: B-mode, Pulsed ultrasound Pulsed OSL readout FIGURE P.112 Ultrasound image of flow in a renal artery. The colour (Radiation Protection) In the pulsed optically stimulated lumi- flow image and spectral Doppler sonogram both use pulsed wave Doppler nescence (POSL) readout, the output is recorded intermittently techniques. Pulse-height analysers (PHAs) for radiation detectors 753 P ulse-height analysers (PHAs) for radiation detectors between successive echoes to reconstruct the Doppler shifts, pulsed Doppler can examine movement at particular depths. The length of the pulse and the interval over which analysis is undertaken dictates the axial resolution, the time between pulse transmitted and received echoes determines the depth at which analysis is made. The processing of pulsed wave Doppler is different for spec- tral Doppler and colour flow imaging and reflects the require- ments for each. For colour flow imaging, rapid calculation of movement in several sample volumes along each of several lines is undertaken by auto or cross-correlation. For spectral Doppler displays, sample and hold of narrowband pulses phase shifts are used to reconstruct the Doppler shift for audio output and onward processing for the sonogram. Pulsed Doppler suffers from one limitation which has pro- found consequences for colour flow imaging and spectral FIGURE P.114 High velocities in a renal artery stenosis lead to high Doppler. The maximum Doppler shift that can be measured Doppler frequencies. The high PRF necessary to measure this unambigu- unambiguously is half the pulse repetition frequency (PRF), the ously leads to an additional superficial sample volume (ringed) and an Nyquist limit. This limit can be almost doubled if flow is unidi- increase in noise in the sonogram. P rectional but the frequencies used for ultrasound, blood velocities in the arterial circulation and depths in the abdomen combine to provide practical constraints. If Doppler frequencies exceed half a major practical problem. For colour flow imaging little can be the PRF then aliasing occurs whereby the displayed flow appears done other than lowering the transmit frequency so that inher- to be in the wrong direction, either on colour flow or spectral ently lower Doppler shifts are easier to image unambiguously. For Doppler (Figure P.113). This can be useful, for example in identi- spectral Doppler some scanners allow a high PRF (HPRF) mode fying areas where blood suddenly accelerates at a narrowing but whereby a second pulse is transmitted before the first has returned can lead to difficulties in measuring high velocities at increasing (Figure P.114). This creates a second, albeit unfocussed, sample depths where the PRF is constrained by the round time necessary volume in superficial tissue where attenuation is lower, increasing for transmission and reception. noise and reducing the clarity of the sonogram from the deeper For bidirectional flow, the limit of velocity V, depth d and sample volume. beam flow angle θ is given by Related Articles: Doppler ultrasound, Colour flow imaging, Sonogram, Nyquist limit, Aliasing Vmax d cosq = c2 /8 ft Pulse-height analysers (PHAs) for radiation detectors where (Radiation Protection) If the ionising radiation is measured with f use of scintillation, semiconductor or proportional detectors, the t is the transmitted frequency of the pulse c is the speed of sound in tissue height (amplitude) of the pulse (electric signal) is proportional to the energy of radiation absorbed by the detector. A pulse-height For a frequency of 3 MHz the limit of Vmax.d cos θ is approxi- analyser (PHA) is a device selecting pulses depending on their mately 0.1 meaning that for a sample at 10 cm depth and with amplitude by means of a preselected voltage window (V, V + ΔV). flow aligned to the beam, the maximum measured velocity is ±1.0 The PHA (Figure P.115) can be used in differential or integral m/s. Since velocities in disease arteries can exceed 4 m/s this is counting mode. In differential mode the PHA selects pulses by two levels (height window): lower level (baseline) V and a window size ΔV (upper level). Only the pulses with amplitude (V, V + ΔV) pass to the recording device. In integral counting all pulses with a height above a certain value V, selected with the lower level discrimina- tor, are counted. The measuring system is calibrated, i.e. the proportionality between a pulse height V and energy E of detected radiation is known, and with use of PHA one can separate counting corre- sponding to the chosen energy of radiation. If it is for only a single energy, for example a photopeak, the PHA is called a single-chan- nel analyser (SCA). Abbreviations: PHA = Pulse-height analyser and SCA = Single-channel analyser. Related Articles: Proportional counter, Scintillation detector, Semiconductor detector, Single-channel analyser Further Readings: Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, FIGURE P.113 Colour flow and spectral Doppler aliasing. At the site of Chichester, pp. 626–647; Brown, B. H. et al. 1999. Medical narrowing (arrow) there is aliasing of the colour flow image. Aliasing is Physics and Biomedical Engineering. Institute of Physics also evident in the spectral Doppler trace where peaks exceed the maxi- Publishing, Bristol, UK, pp. 154–156; Saha, G. P. 2001. Physics mum of the scale. and Radiobiology of Nuclear Medicine, 2nd edn., Springer-Verlag, Pulse-less generator 754 PZT Detector Preamplifier Amplifier Pulse high Recording analyser device FIGURE P.115 Counting system with scintillation or proportional detector and PHA to select energy of measured radiation. Regulation unit HV devices + – 50 Hz Reference HV control signal unit + P – Regulation HV devices unit FIGURE P.116 Block diagram of pulse-less generator. New York, pp. 78–79; Graham, D. T and P. Cloke. 2003. Pulse-pressure-squared integral Principles of Radiological Physics, 4th edn., Elsevier Science (Ultrasound) The time integral of the square of the instantaneous Ltd., Edinburgh, UK, pp. 380–381. acoustic pressure at a particular point in an acoustic field inte- grated over the acoustic pulse waveform (IEC 61157). Unit: Pa2 s Pulse-less generator (Diagnostic Radiology) The pulse-less x-ray generator (also pi = ò p2dt called constant potential generator), is usually a 12 pulse x-ray the pulse generator with high voltage stabiliser (often using high voltage tetrodes or triodes as regulating elements). Figure P.116 shows a Related Articles: Beam width, Intensity typical block diagram of such a generator. It includes a feedback Further Reading: IEC (International Electrotechnical circuit with high voltage dividers, which supply the regulating Commission), International standard, IEC 61157 ed2.0 2007. circuit with signal proportional to the kV across the x-ray tube. This signal is compared with a reference value (the set kV) and, PVDF in case of difference between these, changes the regulating volt- (Ultrasound) Polyvinylidene fluoride or PVDF is a highly non- age of the tetrodes/triodes, which regulates the high voltage over reactive and pure thermoplastic fluoropolymer. the tube. In ultrasound, PVDF is used as a membrane in hydrophones. The pulse-less x-ray generator was the highest level of devel- For further information see Hydrophone. opment of the classical x-ray high voltage generator. It was used for powerful x-ray equipment (as angiographs) and CT scanners. PZT After the 1990s all these are gradually being replaced by medium (Ultrasound) PZT, lead zirconate titanate, is the most commonly frequency x-ray generators. used transducer material. Related Article: High voltage generator For further information see Transducer. Q Q factor sensitisation is described by the wave vector, or the q-value, q = (Diagnostic Radiology) Sometimes image quality parameters and (γ/2π)Gδ [m−1], where γ is the gyromagnetic constant, δ the gra- related dose parameters can be linked in a formula, presenting the dient duration and G the gradient amplitude. Measurements are quality as a single figure. This new empirical figure is an imaging often carried out using several different diffusion times and this performance parameter, known as Q factor or Q value. However it allows for studies of restricted or hindered diffusion. is not an absolute figure, as it often depends on additional param- In MRI the q-space imaging method is used to obtain a prob- eters, such as type of test object, x-ray system parameters, etc. For ability displacement distribution, achievable from the Fourier example, one such Q value, used for the evaluation of some CT transform (FT) of the signal decay curve. Thereby the diffusion scanners, includes a formula incorporating the 50% value of the can be quantified with the mean displacement or full width at half MTF, the percentage of image noise, the nominal slice thickness maximum (FWHM) of the displacement distribution, a measure and the dose in water. which is dependent upon the diffusion time. In the Figure Q.1 a FWHM map obtained from q-space imag- Q-factor ing is seen. The image contrast is similar to an apparent diffusion (Ultrasound) The Q-factor of an ultrasound transducer is a mea- coefficient (ADC) map, however, depending upon the measuring sure of the ‘purity’ of the frequency spectrum of the ultrasound time, see ADC. beam. The presence of the damping block in the transducer allows Related Article: Apparent diffusion coefficient (ADC) Q the shortening of the vibration time (also called ring-down), but also results in a broadening of the frequency spectrum about the Q-switching (Laser) central transducer frequency. The Q-factor is defined as: (Non-Ionising Radiation) A mode of laser output, see Laser out- put mode. Q = f0 / Bandwidth Related Articles: Laser output mode where f0 is the central frequency and the bandwidth is the width QALYs (Quality-adjusted life years) of the frequency distribution. (Radiotherapy) See Quality-adjusted life years (QALYs) A ‘high-Q’ transducer has a narrow bandwidth with a narrow distribution of frequencies around f0. Such a transducer has a long QC (quality control) spatial pulse length and a better frequency response, which is (General) See various articles on the subject under the name important for accurate velocity determination. Continuous-wave Quality control (CW) ultrasound transducers have a high Q-factor. A “low-Q” transducer has a broad bandwidth with a wider dis- tribution of frequencies around f0. Such a transducer has a short QDE (detective quantum efficiency) spatial pulse length, which is important for higher axial resolution. (Diagnostic Radiology) See Detective quantum efficiency (DQE) Related Articles: Damping block, Spatial pulse length, Axial resolution, Doppler ultrasound, CW Quadrature Further Reading:

Bushberg, Seibert, Leidholdt and Boone. (Magnetic Resonance) See Phase quadrature 2012. The Essential Physics of Medical Imaging, 3rd edn., Lippincott Williams and Wilkins. Quadrature artefact (Magnetic Resonance) RF quadrature artefacts are caused by a Q-space (used in diffusion-MRI and NMR) disturbance in the two detector channels of the quadrature detec- (Magnetic Resonance) The q-space imaging technique utilises tor. DC offset of the output of one of the amplifiers will produce extremely high diffusion sensitivities which makes it possible to a bright point in the centre of the image. A higher gain of one study the geometrical shape and extension of several chemical com- detector than the other will result in a ghost appearing diagonally pounds, such as polymers and emulsions. The technique has been in opposite from the real object. use in the field of nuclear magnetic resonance (NMR) for a long time These artefacts can be caused by the technical faults within but has not been applicable within the area of MRI due to limita- the quadrature detectors and typically must be rectified by the tions of the MRI scanner hardware, mainly related to the maximum service engineer. achievable gradient strength. With increasing system performance Abbreviations: RF = Radiofrequency and DC = Direct current. it has now become possible to achieve higher diffusion sensitivities Related Article: Ghost artefact also in MRI, however, not as high as in the NMR environment (due to physiological limits like peripheral nerve stimulation). Quadrature coil The technique is based on the same method as for conven- (Magnetic Resonance) An RF quadrature coil (also called a cir- tional diffusion weighted (DW) MRI, but includes a large num- cularly polarised or CP coil) is a coil that is used to transmit and/ ber of measuring points with continuously increasing sensitivity or receive a circularly polarised B1 field. where the maximum sensitivity is inversely proportional to the A circularly polarised RF transmit coil requires only half the achievable q-space resolution. In q-space imaging the diffusion transmit power required by an equivalent linearly polarised coil. 755 Quadrature detection 756 Q uadrature detector linear polarisation. The two signals detected are 90° out of phase and are shifted relative to one another prior to recombination. Quadrature detection (Ultrasound) In quadrature detection, the received signal is mixed with two different signals, one in-phase component (cosine), and one quadrature component being 90° out of phase with the first (sine). Usually these components are referred to as the I-channel (or component) and the Q-channel (or compo- nent). This phase shift can for instance be accomplished with a high reactance capacitor. The reason for employing quadrature detection in Doppler systems is that the reverse and forward flow components can be separated, using additional circuitry shown in Figure Q.3. The 90° phase shift after the low-pass filter (which is there to remove the sum frequency components) is, however, difficult to achieve over the whole range of Doppler frequencies using analogue components. Cascaded RC-circuits may be used but if the signal can be digitised, a digital Hilbert transform can FIGURE Q.1 A FWHM map obtained from q-space imaging. be employed instead. The Hilbert transform performs a 90° phase shift of all frequencies up to the Nyquist limit. For the separa- tion of reverse and forward components to be successful it is very A quadrature coil creates a spinning B1 component that rotates in important the phase shift is exactly 90°. If not, a spurious compo- Q the same sense as the precession of spins. nent will show up at the mirrored frequency, i.e. if only a forward The conceptual operation of a quadrature transmit coil is Doppler shift of 400 Hz is evident, a weak reverse shift at −400 shown in Figure Q.2. The driving signal from the RF generator will also be apparent. Less than 2° of mismatch will keep the is split into two paths. A phase shifter creates a 90° phase shift spurious component under 36 dB compared to the wanted signal. between each path. The geometry of the coil is such that each of A similar error can arise from having unmatched amplitudes in the driving signals generates mutually orthogonal linearly polar- the low pass filters after the mixers. ised magnetic fields. As these component fields are both orthogo- Related Articles: CW Doppler, Doppler ultrasound nal and in quadrature, the resultant is a circularly polarised field. Quadrature coils also operate as RF receivers. The signal to Quadrature detector be detected (the spinning transverse component of magnetisation) (Magnetic Resonance) Quadrature detection is a signal process- is circularly polarised. Again, this spinning component can be ing method for extracting in phase and in quadrature signal com- thought of as two orthogonal, in quadrature linear polarisations. ponents from an input signal, relative to some reference signal. A linearly polarised receive coil would detect only one of these The process is illustrated in Figure Q.4. polarisations. A quadrature receive coil is more efficient, detecting both polarisations and giving an SNR improvement of a factor of √2 compared to an equivalent linearly polarised coil. The recep- Cos(2πf0t) tion process is essentially the reverse of the transmission process ∑ shown in Figure Q.2. The geometry of the coil is such that one Lowpass Doppler signal filter 90° Reverse flow received signal represents a linear polarisation of the spinning transverse component and the other represents the orthogonal Mixers Lowpass For ard flo From er 90° w w filt transmitter ∑ Sin(2πf0t) FIGURE Q.3 Block diagram of a unit for separation of forward and reverse Doppler components using quadrature detection. 90° Phase shifter Coil RF receiver coil × Low pass filter Real component In quadrature linearly polarised × Low pass filter Imaginary component B1 fields Resultant circularly RF 90° polarised B1 field oscillator phase To RF shift transmitter FIGURE Q.2 Quadrature coil in transmission. The coil is shown look- ing down the z-axis (i.e. along the bore in a conventional MRI). FIGURE Q.4 Quadrature detection. Qualified expert (QE) 757 Quality-adjusted life years (QALYs) In MRI, the ‘reference signal’ originating from the RF oscil- and regulations. In some countries a qualified expert is called lator can be written as Radiation Protection Adviser or Expert in Radioprotection. The certification of the QE is issued by appropriate boards or pro- Sref (t ) = Aocos wot (Q.1) fessional societies. The certification is based on the academic qualifications (usually a basic requirement) and tested – docu- This signal is driven with a frequency ωo, corresponding to the mented practical skills – expertise and knowledge of specific Larmor frequency. relevant fields such as medical physics, radiation protection, The signal component detected by the RF receiver coil from occupational health, quality assurance/quality control measure- any given voxel is ments, etc. In general, there are national lists of qualified expert autho- S rised to work. There might be various levels of work authorised detect (t ) = A1cos((wo + Dw)t + f) (Q.2) in the license of the QE, depending on his/her skills and the kind where of applications (kind of ionising radiation and energy used). The Δω is the local difference in precession frequency in that voxel duties of the QE are usually specified in the national laws/regula- relative to the centre frequency tions and include personal dosimetry, environmental dosimetry, ϕ is the local phase difference of the transverse component of record keeping, etc. magnetisation relative to the reference signal Further Reading: IAEA (International Atomic Energy Agency). 1996. International basic safety standards for protection Multiplication of the incoming detected signal by the refer- against radiation and for the safety of radiation sources. Safety ence signal yields Series No. 155, International Atomic Energy Agency, Vienna, Austria. Smullt (t) = Ao cos wot.A1 cos((wo + Dw)t + f) Quality-adjusted life years (QALYs) Q 1 (Radiotherapy) In order to have a systematic way to appraise = Ao A1(cos(Dwt + f)) (Q.3) 2 the benefit of treatment options, many health organisations use quality-adjusted life years (QALYs) as the comparison metric. + cos((2wo + Dw)t + f) In contrast to metrics of quantity of life (e.g. overall survival), The second term is eliminated using low pass filtering, yielding a a QALY is a measure of the state of health of a person or group demodulated signal: which incorporates quality of life. One QALY is equal to one year of life in perfect health. The condition of perfect health is S t A D measured in terms of the person’s ability to carry out normal demod(in- phase) ( ) = cos( wt + f) (Q.4) daily activates, freedom of physical pain and mental distur- bance. This is calculated by estimating the years of life remain- 1 (where A = Ao A1). A second demodulator is used to implement ing for a patient after treatment or intervention and weighting 2 each year on a scale of 0 to 1 (0 = death, 1 = best possible health the same process, except using the in-quadrature reference signal: state). To ensure that QALYs remain systematic, the weighting Sref (t ) = Aosin wot (Q.5) valuations are usually derived using health utilities. EQ-5D is an example of a health utility used by NICE in the United The output of the second demodulator is Kingdom. Scores for the EQ-5D are generated from the ability of the individual to function in five health dimensions: mobil- Sdemod(out of phase) (t ) = Asin (Dwt + f) (Q.6) ity, pain/discomfort, self-care, anxiety/depression and usual activities. This systematic scoring allows QALYs to be used In general, Equations Q.4 and Q.6 represent the components of as a common currency to assess performance from a variety the detected signal which are in phase and in quadrature with of interventions; an example scenario is shown in Figure Q.5. the RF driving signal. For the particular case of a purely in- Effectiveness of interventions can then be combined with the phase detected signal (ϕ = 0) at centre frequency, only Equation corresponding cost in order to generate cost-utility ratios which Q.4 shows an output and only Equation Q.6 shows an output are heavily used in health economics. A cost-utility ratio is a for an in quadrature signal (ϕ = 90°). In complex notation the difference between the costs of two interventions divided by in-phase signal is assigned as the ‘real component’ and the the difference in the QALYs they produce (Equation Q.7). This quadrature signal as the ‘imaginary component’. Digitised sam- provides a cost per QALY of a given intervention, which can ples from each are written into k-space as complex number val- be used by health organisations to evaluate the availability of ues. In current systems the entire quadrature detection process intervention. may be implemented in the digital domain, while older systems Whilst QALYs provide a systematic way of evaluating the used analogue electronics. Note that in practice more complex quality of life benefits from a variety of medical procedures, demodulation and quadrature detection schemes may be imple- they are far from perfect as a measure of outcome. Quality of life mented, including demodulation to an ‘intermediate frequency’ is subjective and therefore difficult to systematically score; the less than ωo. importance attached to each of the health dimensions is highly dependent on age, life context and life responsibilities. QALYs are Qualified expert (QE) not well suited for evaluating the treatment of less severe health (Radiation Protection) The qualified expert is a professional problems. Finally, chronic diseases, where the quality of life is a who is duly authorised/certified to work in specific fields where major issue and survival can be less of an issue, are difficult to such figure is required by international/national authorities’ laws evaluate using QALYs. Quality assurance 758 Quality assurance relating to equipment reliability and/or data-accuracy. Patient- safety is of paramount importance to most equipment vendors (their sales depend on it), and safety issues should have been exhaustively tested by the manufacturer. Nevertheless, safety- tests form part of routine QA procedures; these tests are designed to check patient-critical systems which may not be encountered during routine clinical operation. Some equipment faults may be spotted first by clinical staff, and so fault-logs also form part of the QA system. Most QA activity is taken up by assessment of accuracy/reli- ability of equipment/processes, ensuring that data produced by a medical device/process is accurate. The tests undertaken to deter- mine accuracy and reliability

are defined under quality control (QC) procedures, which form part of the overall QA system. Process optimisation is another aspect of QA that aims to reduce errors in clinical procedures. This is of particular impor- tance in areas such as diagnostic imaging, where a variety of clinical and technical factors can invalidate the usefulness of a scan. As well as the economic impact of poorly optimised proce- FIGURE Q.5 Quality-adjusted life year comparison of two interven- dures, there are additional risks associated with ionising radiation tions. The QALYs of intervention A is being directly compared to that of that necessitate a ‘right-first-time’ ethos. Thus process review, QC intervention B. results and fault-logs are all aspects that contribute to a responsive Q QA system. Related Article: Quality control Cost utility ratio = ( Quality assurance Cost of Intervention A - Cost of Intervention B) (Q.7) (Ultrasound) Quality assurance is a term used to describe meth- (No. of QALYs produced by Intervention A ods and processes to ensure that products and services are of a - No. of QALYs produced by Intervention B) sufficient standard for their purpose. It is commonly used in man- ufacturing where there are internationally agreed terms to cover Related Article: Clinical trial endpoints various aspects of the quality process under ISO 9000 (www .iso .org). Quality assurance In clinical ultrasound the term quality assurance is commonly (Magnetic Resonance) Quality assurance is the management and used to describe performance testing of ultrasound equipment. review of the quality control tests that should be performed on a This may be done at various stages including regular basis on MRI equipment. The aim of quality assurance is to ensure that the quality control tests take place, the protocols • Comparative testing and evaluation prior to purchase and results are reviewed regularly and any necessary alterations • Acceptance testing prior to clinical use are made to the procedure. • Continued performance assessment in service Legal requirements to fulfil specific criteria in MRI quality assurance will vary from country to country but a best practice Assessment of system performance can be performed clinically policy should be adopted to make sure quality assurance takes (e.g. is a system adequate for its task and does it remain so) or place. This will ensure that the equipment is maintained at a clini- by technical assessment of performance, usually using ultrasound cally useful level. phantoms (link). Related Article: Quality control Change in performance can often be gauged by users, for example if controls do not work or work inadequately or if there Quality assurance are obvious changes in the image appearance or new artefacts (Radiation Protection; General) Quality assurance (QA) is the appears. More subtle changes in performance may be difficult system that provides confidence in the safety, reliability and accu- to observe in everyday use. The need for routine quality assur- racy of a medical device or process. A QA system is based on a set ance is controversial. Many protocols were designed for systems of procedures that allow the monitoring of the medical device or with a high proportion of analogue circuitry; the increased use process, such that if a fault occurred it would be quickly noticed of digital processing and the ability for self-testing means that and the equipment removed from service/process suspended for beamforming and the processing of images is less likely to suffer review. degradation in performance. The transducer is still at risk from A QA system should encompass the life-cycle of a medi- deterioration or damage and some scientists argue that the focus cal device/process, instilling quality principles at specification, of QA testing should be on transducer performance. tendering, acceptance, commissioning, in-service and de- A large range of measurements can be made, for example in commissioning stages. A mature QA system should inform the B-mode: specification and tendering process, just as acceptance and com- missioning should inform the in-service QA. • Spatial resolution axial, lateral, elevation QA systems generally categorise faults relative to their impor- • Penetration, (sensitivity) tance, with patient-safety foremost, followed by lower-level faults • Contrast resolution Quality audit 759 Quality control • Measurement accuracy (callipers), linear and area wires show as a series of points permitting measurement of axial • Image uniformity (includes monitor) and lateral resolution and calliper accuracy. • Dead zone • Hard copy performance Quality audit (Radiation Protection; General) An assessment of the systems Doppler parameters include and procedures and the adherence to those procedures within a Spectral Doppler: department using radiation. An assessment of systems and pro- • Velocity direction indication cedures would check compliance against legal requirements (e.g. • Sample volume dimensions national ionising radiation regulations) as well as accepted codes • Directional discrimination of practice. Assessing adherence to local protocol is achieved • Penetration depth (sensitivity) through various quality control metrics, such as auditing patient • Velocity estimation accuracy administered doses, image quality and equipment maintenance. • Waveform index estimation accuracy There must be appropriate mechanisms in place to review the • Volume flow estimation accuracy results of quality audits, to feed back those results and apply cor- rective actions, and to ensure the actions are completed. Colour flow imaging: Related Articles: Quality control, Quality assurance • Lowest detectable velocity • Highest detectable velocity Quality control • Image spatial resolution (and registration accuracy) (Magnetic Resonance) A quality control system is put into place • Temporal resolution to ensure that the MRI system is maintained at a clinically useful • Velocity resolution level. System variability can be caused by drift in the electronics • Tissue colour suppression performance of the system, slow drift of the magnetic field, gradient system failures or the introduction of ferromagnetic articles into the scan- In practice, making consistent measurements from phantoms ner room, for example. These factors can lead to a variety of arte- Q requires considerable expertise, spatial resolution parameters, for facts or changes in image characteristics. The most important of example require rigorous attention to detail for the environment, these arguably is a decrease in SNR, caused either by a reduction scanner settings and use of the phantom. Doppler parameters are in signal or an increase in noise. The reduction of SNR will lead particularly difficult to measure and a simple measurement such to a less clinically useful image which, if not noticed, may affect as peak velocity is dependent on depth, power, gain, beam steer- clinical diagnosis. ing and position in the array. A clear schedule of quality control measures is important, with A diagrammatic representation of ultrasound transducer and the most important measures repeated more frequently. Typically, phantom and the resulting image is shown in Figure Q.6. a daily test is performed with a uniform test object and a stan- The insonated region on Figure Q.6 is shown by the box on the dard pulse sequence to inspect the SNR and image uniformity. side of the phantom beneath the transducer face. Images of the This gives a good overview of the performance of the scanner in FIGURE Q.6 Diagrammatic representation of a linear array used to image part of an ultrasound phantom. Quality control 760 Quality Index a short amount of time. The SNR can be monitored over time to Quality factor look for any trends or sudden problems in the scanner. (Magnetic Resonance) The equivalent circuit for a simple RF coil Other tests are performed on a less frequent basis to provide is shown in Figure Q.7a. The resistance represents the resistance a more in-depth and detailed analysis of the different parameters of the unloaded coil plus loading effects of the patient. that can be affected by system variability. These include SNR, The frequency response Figure Q.7b shows the variation of the spatial resolution, uniformity, geometric distortion, slice thick- current flowing in this coil with applied driving frequency. The ness, slice position, slice separation and ghosting. This set of tests coil shows a resonant peak at a frequency ω0. The quality factor requires a set of appropriate test objects. or Q-factor is a measure of the sharpness of this peak. Some quality control measures may be specific to the clinical For the series circuit in Figure Q.7b the quality factor is and research interests of the site, for example breast imaging or given by MR spectroscopy. Abbreviation: SNR = Signal to noise ratio. w L Q = 0 R Quality control The quality factor can also be calculated from the measured coil (Nuclear Medicine) Quality control refers to the activities per- frequency response: formed to acquire evidence to verify quality in specific parts of a business or department. w In nuclear medicine, quality control refers to the system- Q = 0 Dw atic control of imaging systems, patient and personnel routines to ensure that the service fulfils the predefined quality condi- where Δω is the measured coil bandwidth. tions. A specific quality control can be a step in the QA process The Q-factor is an important measure of coil performance as where the entire research and clinical activity are reviewed and SNR improves with the square root of Q. Q tested. One example of a quality control routinely preformed is the SNR µ Q uniformity measurements of a scintillation camera. Related Articles: Quality assurance, Uniformity Quality Index (Radiotherapy) Linear accelerator beams tend to be generically Quality control specified by a nominal accelerating potential (e.g. 4, 6 MV, (General) Quality control (QC) refers to the processes by which etc.). While this aids the user in a simple definition of the treat- the data-accuracy and safety of a medical device is assessed. ment beam, it by no means specifies the true nature or energy Quality control describes any set of tests that should be carried of the beam. This is important as two machines with the same out on an equipment to determine whether it is fit for purpose. nominal accelerating potential can produce very different beam Generally, a minimum set of tests that should be carried properties such as percentage depth doses, output factors, etc. out on a piece of equipment is defined by a professional body and the user must be aware of this before thinking about beam or a national regulatory authority. These will set-out required matching. tests and the acceptable performance for each test. Acceptable While in diagnostic radiology a beam can be specified in performance may be defined in terms of absolute accuracy, or terms of half value layers (HVLs) of materials such as copper a tolerance of variation from a previous measurement (e.g. at or aluminium, the spectrum of energies in a linear accelerator commissioning). treatment beam renders this beam qualifier irrelevant. The beam QC then concerns itself with monitoring of data-accuracy over specifier of choice in radiotherapy is called the quality index. the in-service use of the medical device, and informs the QA sys- The quality index is simply a measurement of the tissue phan- tem, which encompasses the broader clinical application of the tom ratio (TPR) at two specific depths, 10 and 20 cm, and this device, and all potential-related sources of error, over the life- is often denoted by TPR20 10 . The measurement is made with the cycle of the device. chamber positioned at the isocentre for both readings, and with a Related Article: Quality assurance set field size of 10 × 10 cm2. An illustration of the setup for each field is given in Figure Q.8. Quality control (Ultrasound) Quality control is part of the larger quality assur- ance process of a product or service. In manufacturing, quality C control could include monitoring and auditing the manufacturing 1 process and product. Quality assurance would include the larger I process including a company’s systems to ensure the design and 0.7071 manufacturer of systems lead to a high quality product. In indus- V L Δω try, quality is addressed by the ISO 9000 group of standards for quality management systems. In ultrasound imaging the term quality control might include R ω Frequency the testing of scanners to ensure that they achieve the designed (a) (b) 0 standards of performance and safety. The term is rarely used in clinical applications and departments where the phrase ‘quality FIGURE Q.7 (a) Equivalent circuit for an RF coil. (b) Frequency assurance’ is used when describing ways to ensure the perfor- response of coil. The bandwidth is the frequency

difference between the mance of scanners in clinical use. The use of this term is dis- frequencies where the response has fallen to −3 dB of the peak response cussed under quality assurance. (i.e. reduced in amplitude by a factor of √2). Coil current Quantitative imaging 761 Quantum number Related Articles: Molecular imaging, CT angiography, Magnetic resonance spectroscopy, Radionuclide imaging, Ultrasound contrast agents FSD = 100 cm Hyperlink: www .r sna .o rg /en /rese arch/ quant itati ve -im aging -biom arker s -all iance Quantum detection efficiency (QDE) 20 cm 10 cm (Diagnostic Radiology) See Detective quantum efficiency (DQE) Quantum efficiency 10 × 10 cm2 10 × 10 cm2 (Diagnostic Radiology) The quantum efficiency can generally be defined as the percentage number of input quanta that contrib- FIGURE Q.8 Illustration of the setup employed for the two measure- ute to the output. A quantum efficiency can be defined for each ments required to find the Quality Index of a linac x-ray photon beam. individual step in an imaging process where photon or charged particle interactions take place and often the input energy is con- verted from one form to another. A measurement of the quality index is essential for current For example, in the photocathode of a photomultiplier, the absorbed dose to water codes of practice for calibration of MV quantum efficiency is treatment beams. Typical values are 0.656 for 6 MV and 0.758 for 15 MV beams. (Number of photoelectrons emitted) QE = Alternatively, the quality index Q can also be determined from ( Number of incident light photons) depth dose measurements with fixed SSD of 100 cm at depths 10 cm and 20 cm with a formula derived from experimental data by Q Andreo et al. (1986): The QE may be expressed as a single figure for a quantum inter- face or, since the interactions will often have an energy depen- Q = 2.189 -1.308j + 0.249j2 dency, can be expressed as a function varying with energy or wavelength. The lower the quantum efficiency, the lower the sen- sitivity of detection and this leads to a higher noise component. In with j = M10 cm/M20 cm. a chain of conversion steps, a particularly low quantum efficiency Related Articles: Calculation of absorbed dose, Beam energy, is described as a quantum sink. Tissue phantom ratio (TPR), Beam quality Further Reading: Andreo, P., A. E. Nahum and A. Brahme. Quantum mottle 1986. Chamber-dependent wall correction factors in dosimetry. (Diagnostic Radiology) Quantum mottle refers to the pattern of Phys. Med. Biol. 31:1189–1199. quantum noise, or random variations, in a radiographic image which is due to the statistical fluctuation of x-ray photon absorp- Quantitative imaging tion and consequent light photon emission by an intensifying (Nuclear Medicine) Quantitative imaging is a whole branch of screen or digital radiology scintillator screen. The faster the medical imaging, which includes all imaging modalities and is intensifying screen, or the higher the kV, the more light photons associated with extracting additional information (mainly numer- are produced and so fewer x-ray photons actually contribute to a ical) from the visual image (mainly digital Image). According to final image of a desired optical density or pixel value. the Radiological Society of North America website: Since the emission and detection of photons are normally dis- Quantitative imaging is the extraction of quantifiable features tributed, the noise, or randomness, associated with the number from medical images for the assessment of normal or the severity, of photons is proportional to the square root of the number of degree of change, or status of a disease, injury, or chronic condi- photons. So with fewer x-ray photons contributing to the image, tion relative to normal. the greater is the noise, or mottle, seen in the image. Quantitative imaging develops, standardizes and optimizes: • Anatomical, functional and molecular imaging acquisi- Quantum noise tion protocols (Nuclear Medicine) Quantum noise refers to the fluctuations in • Data analyses detected signal because of the finite number of discrete signal • Display methods (energy) carriers (e.g. photons, electrons). These fluctuations are • Reporting structures manifested as noise in an image. By measuring a known intensity the average number of photons collected can be established but These features permit the validation of accurately and precisely will not lead to knowledge of the actual number of photons col- obtained image-derived metrics with anatomically and physi- lected per element (pixel). The number of photons collected per ologically relevant parameters, including treatment response element will have a mean value in accord with a Poisson distribu- and outcome, and the use of such metrics in research and patient tion (Poisson statistics). care. Related Articles: Pixel, Poisson distribution Clinically important metrics (e.g. left-ventricular ejection fraction, tumour SUV, blood volume, etc.) can be measured Quantum number using imaging modalities such as nuclear medicine (planar and (Nuclear Medicine) Quantum numbers are values describing the SPECT), PET, CT, MRI, US, which fall under the umbrella of dynamics of a quantum system such as an atom or a particle. ‘quantitative Imaging’. Quantum numbers can describe the energy state of an electron in Quartz 762 QUIPSS – QUIPSS II – Q2TIPS an atomic system or the angular momentum or spin of an electron. lithotripsy to obliterate kidney stones and HIFU (high intensity The electron orbits in an atomic structure is described by four focused ultrasound) for destruction of pathogenic tissue. quantum numbers: principle (n), angular (l), magnetic quantum Related Articles: Ceramics, HIFU, Piezoelectric crystal in (m) and spin (s) quantum number. n, l and m describe the size, ultrasonography, Potential difference, PZT, Ultrasound shape, orientation of the orbit respectively, s describes the spin of the individual electrons. Quench (quenching) (Magnetic Resonance) A quench is the sudden loss of supercon- Quartz ductivity of the magnet coils due to a local temperature increase (General) in the magnet above the critical temperature of the material used for the coils. The material most commonly used for the coil con- Molar mass: 0.060 kg/mol struction is niobium titanium (Nb-Ti) which has a critical tem- Density at STP: 2650 kg/m3 perature of 9 K. To keep constantly the coils at a temperature Melting point: 1850–2000 K lower than 9 K, the Nb-Ti filaments are immersed in liquid helium Boiling point: 2500–2750 K at a temperature of 4.2 K. To avoid the possibility of local transi- tions to the resistive state the filaments of Nb-Ti are enclosed in Refractive index: 1.53–1.56 a matrix of copper, good thermal conductor, to ensure a stable temperature state. During a quench the energy of the magnetic The mineral quartz (from German ‘Quartz’) is the most abundant field is converted into heat. The energy is equal to 1/2 LI2 where mineral in the Earth’s continental crust. Quartz is a rhombohe- L is the inductance and I is the coil current. After releasing a dral crystal consisting of a lattice of silica (silicon dioxide, SiO quench, the magnetic field strength drops to approx. 20 mT within 2) tetrahedra (Figure Q.9). Naturally occurring quartz crystals are approx. 20 s. The helium evaporates rapidly producing a huge normally twinned or distorted, lacking macrocrystalline struc- amount of helium gas that has the potential to displace all the Q ture. Pure quartz is usually colourless or white, although coloured oxygen in the magnet room if it is not carried away as the volume forms exist, such as rose, amethyst and smoky. The various types ratio of the gaseous to liquid helium is 730. The rapid escaping of quartz depend on the specific macro and microcrystalline of helium displaces the oxygen in the room and there is a real structure, though generally the more transparent varieties tend to danger of asphyxia if the lighter than air gas displaces all the air. be macrocrystalline in structure. The oxygen concentration in the room can be checked via an oxy- Quartz is widely used in the production of concrete and glass, gen meter and in case of low oxygen concentration they automati- and it is a major ore of silicon, which is used in integrated cir- cally command a dedicated forced ventilation of the examination cuits (‘chips’). Quartz exhibits the phenomenon of piezoelectric- room. In case the exhaust line fails in part or in full, the heating ity, meaning that it generates an electric potential when under and air-conditioning system is not capable of guaranteeing suf- mechanical stress. This property led to the use of quartz in spe- ficient air exchange. Heavy fog formation toward the ceiling of cific applications such as phonograph pickups and crystal oscil- the examination room impairs visibility. Also, the pressure in the lators in electric circuits of watches and computers. The quartz examination room will rise. Depending on the type of defect, e.g. clock utilises its natural resonant frequency to accurately mark large leaks, hazards such as acute hypothermia or frostbite are time. It is used in devices for accurate measurements, includ- present. When helium escapes, the condensation process at pipes ing microbalances, strain gauges and thin-film thickness moni- or at the magnet may lead to local oxygenation with an increase in tors. These measurements are accomplished through mechanical fire hazards in the vicinity of these components. loading which changes the resonant frequency of the quartz crystal oscillator. Piezoelectric sensors function by producing Quenching a force on the opposing faces of a sensing element due to the (Nuclear Medicine) Quenching refers to any process that inter- change in its physical dimension. These sensors are often used as feres with the counting performance of counting system. extremely sensitive microphones, such as those used in industrial In many liquid scintillators dissolved oxygen works as a non-destructive testing (NDT). They often act as both a transmit- quenching agent leading to a reduction in fluorescence efficiency. ter and receiver of sound because the piezoelectric property is Liquid scintillators are therefore sealed in a closed volume from reversible. which most of the oxygen has been extracted. There are two types Medical Applications: Quartz can be used in piezoelectric of quenching; colour quenching and chemical quenching. Colour sensors of ultrasonic transducers for medical imaging, although quenching means that the added sample induces a change in the the ceramic lead-zirconate-titanate (PZT) is more commonly optical properties of the solvent so that the scintillation light is used. Ultrasound transducers are also used in physiotherapy, unable to penetrate the solvent. Chemical quenching refers to sample-induced chemical changes that interfere with the scintil- lation energy transfer process. Quenching used in Geiger–Müller (GM) counters is neces- O O O sary to limit the amount of detector dead time. The quenching is accomplished with use of a small quantity of alcohol vapour Si Si Si in the gas. Related Articles: Liquid scintillation (LS) counting, Chemical O O O quenching, Geiger–Müller (GM) counters Si Si Si QUIPSS – QUIPSS II – Q2TIPS (Magnetic Resonance) The arterial spin labelling (ASL) tech- FIGURE Q.9 Chemical structure of quartz (SiO2). nique QUIPSS (quantitative imaging of perfusion using a single QUIPSS – QUIPSS II – Q2TIPS 763 Q UIPSS – QUIPSS II – Q2TIPS TI2 In-plane In-plane TI1 presaturation slab presaturation TI1S Inversion pulse Periodic saturation pulses EPI readout 90° 90° 90° 90° 90° 90° Imaging slices RF Gslice Gfreq G Periodic phase saturation slice × n Inversion slab (tagged region) (a) (b) Q FIGURE Q.10 Q2TIPS pulse sequence diagram (a) with positions of regions where the different pulses are applied (b). (Modified from Luh, W. M. et al., Magn. Reson. Med., 41, 1246, 1999.) subtraction), the modified version QUIPSS II (Wong et al., 1998) pulses are applied to the imaging slices for further reduction of and Q2TIPS (QUIPSS II with thin-slice TI1 periodic saturation) signal from the static tissue. The presaturation is followed by a (Luh et al., 1999) are all ASL saturation techniques which are 180° inversion, usually a pulse applied to a 10 cm tagged region. less sensitive to transit delays, and all provide quantitative perfu- After a delay TI1 the train of thin-slice saturation pulses is applied sion images based on measurements at a single inversion time. at the distal end of the tagged region. Finally, the images are QUIPSS and QUIPSS II differ in the application of the saturation acquired by a fast read-out sequence (e.g. EPI, spiral reconstruc- pulses. In QUIPSS, the saturation is applied in the imaging slice tion, HASTE or 3D GRASE)

in a region distal (superior) to the after a delay time (TI1), and the signal in the difference image is inversion slab and with a 1 cm gap between the inversion slab and based on the blood that arrives between saturation and the readout the imaging region. The control image is acquired according to pulse. In QUIPSS II, the saturation is applied in the inversion slab either the EPISTAR or the PICORE preparation method. instead, and only blood that has left the inversion slab before satu- Related Articles: Perfusion imaging, Arterial spin labelling, ration contributes to the signal in the difference image. Q2TIPS is EPISTAR, FAIR, PICORE a modified QUIPSS II saturation technique, where the QUIPSS II Further Readings: Luh, W. M., E. C. Wong, P. A. Bandettini saturation pulse is replaced by a periodic train of thin-slice satu- and J. S. Hyde. 1999. QUIPSS II with thin-slice TI1 periodic satu- ration pulses at the distal end of the inversion slab. This modifica- ration: A method for improving accuracy of quantitative perfu- tion is introduced to minimise the errors related to the incomplete sion imaging using pulsed arterial spin labeling. Magn. Reson. saturation of the spins of inverted blood occurring in a thick satu- Med. 41:1246–1254; Wong, E. C., R. B. Buxton and L. R. Frank. ration pulse. Figure Q.10 illustrates the Q2TIPS pulse sequence 1998. Quantitative imaging of perfusion using a single subtraction and the corresponding slice positions: Two in-plane presaturation (QUIPSS and QUIPSS II). Magn. Reson. Med. 39:702–708. R R (Roentgen) Related Articles: AORD, Irradiance, Light source, (Radiation Protection) The unit of exposure is the Roentgen, R. Photodiode, Phototherapy, Solid angle, UV dosimetry It is named after Wilhelm Röntgen, German physicist and discov- Further Readings: Czapla-Myers, J. S., K. J. Thome and S. erer of x-rays (1845–1923). 1 R is equivalent to 1 C of electrical F. Biggar. 2008. Design, calibration, and characterization of a charge of one sign, liberated in 1 kg of matter. field radiometer using light-emitting diodes as detectors. Applied Related Articles: Roentgen (R) Optics 47(36):6753–6762; Ihrke, I., J. Restrepo and L. Mignard- Debise. 2015. Principles of light field imaging: Briefly revisiting Rad (radiation absorbed dose) 25 years of research. IEEE Signal Processing Magazine 33(5):59– (Radiation Protection) See Radiation absorbed dose (Rad) 69; Kitsinelis, S. and S. Kitsinelis. 2015. Light Sources: Basics of Radial k-space sampling Lighting Technologies and Applications. CRC Press. (Magnetic Resonance) The majority of MRI acquisition sequences use Cartesian sampling techniques that fill k-space row by row. The Radiant exposure k-space data (which are a set of Fourier coefficients) are equally (Non-Ionising Radiation) See Radiance spaced in rows and columns which can be easily transformed into image space using the Fast Fourier Transformation (FFT). k-space Radiation data can be acquired in other patterns as well, including spiral and (Radiation Protection) Radiation is a process by which energy is radial patterns. Radial k-space sampling alters x- and y- direction transferred from one point in space to another. It usually refers gradients simultaneously to sample k-space with radial spokes to the transfer of energy by electromagnetic wave or photon. The which pass through the centre of k-space. One advantage of radial R higher frequency/shorter wavelength end of the electromagnetic sampling is that data sampling is denser in the centre of k-space radiation spectrum has enough energy to break biological molec- where most of the image contrast is concentrated. In addition, this ular bonds and/or to eject electrons from their atomic orbits – that radial sampling of k-space making image acquisition less suscep- is causing ionisation – this radiation therefore being called ionis- tible to motion artefacts, and the motion in radial sampling results ing radiation. in blurring and noise rather than coherent ‘ghost’ artefacts. Related Articles: Radiation, Gamma rays, Quanta, Ionisation, For Fourier-based reconstruction, the images need to be re-grid- Ionising radiation, Non-ionising radiation ded onto a Cartesian coordinate system and this results in compu- tational complexity and loss of data integrity. Radial sampling has Radiation absorbed dose (Rad) recently seen a resurgence in popularity as it is amenable to use in (Radiation Protection; General) Originally defined in the first sparse sampling techniques such as compressed sensing. 3D imple- recommendations of the ICRP in 1958 (Publication 1), this was mentations of radial sampling include a ‘Koosh-ball’ sampling tech- the unit of absorbed dose until superseded by the Gray. 1 Gy = nique where all radial spokes pass through the centre of k-space, or 100 rad. The unit is still used, particularly in the United States. a stack-of-stars where multiple 2D planes are acquired (Figure R.1). Related Article: Absorbed dose Related Articles: k-space, k-space trajectories, Spiral sam- pling, FFT, Compressed sensing Radiation, alpha Radiance (Radiation Protection) Alpha radiation is composed of alpha par- (Non-Ionising Radiation) This is a measure of the flux of a light ticles emitted during radioactive decay of nuclides that decay by source per unit solid angle. The radiance is measured in Wm−2sr−1 alpha decay. and, for an extended light source is independent of the distance An alpha particle (α) is a helium nucleus and consists of two from the source. neutrons and two protons. It therefore has a total mass of four. Light exposure of the eye is usually expressed as a radiant Alpha particles are typically emitted with energies between 3 and exposure. 7 MeV. Alpha radiation has a very short range due to the mass and double charge of the alpha particles. This causes a very high ioni- sation density and LET. Alpha decay normally occurs when very large nuclei decay – an example is the decay of radium 226: 226 222 Ra ® Rn + a 88 86 Abbreviation: LET = Linear energy transfer. Related Articles: Alpha particle, Alpha particle emitter, Radioactive decay 765 Radiation, beta 766 Radiation detection systems FIGURE R.1 Methods of sampling k-space. A Cartesian approach acquires k-space data in a line-by-line fashion (left). Radial sampling acquires k-space data as a series of spokes which pass through the centre of k-space. Radiation, beta biological effect on the cell. For more details, see the article (Radiation Protection) See Beta radiation on Radiolysis. Repair of Radiation Damage: When cells are exposed to ion- R Radiation biology ising radiation, this damage does not necessarily lead to adverse (Radiation Protection) Both ionising and non-ionising radiation effects or bioeffects. Cells have proteins and enzymes whose may cause damage to cells, leading to biological effects. These main function is to act as part of the mechanisms to repair dam- effects are dependent on the type and energy of the incident age to DNA. Ordinarily these mechanisms work very well to radiation and are described under Bioeffects. The science study- repair radiation damage. ing such radiation damage and biological effects on the human Classification of Radiation Damage: The damage caused by body is known as radiation biology, or radiobiology, leading to the radiation may be described in terms of the effect it has upon the description of radiobiological models to predict the harm caused individual (somatic damage), the reproductive cells (genetic dam- by exposure to ionising radiation. age) and the unborn foetus (teratogenic damage). Related Articles: Bioeffects, Radiation damage, Somatic Damage: There is a latent period associated with Radiobiological models somatic damage. The latent period is the length of time between irradiation and tumour development and varies depending on the type of tumour that develops and the age of the exposed person. Radiation, bremsstrahlung There are two risk models that may be applied to describe the (Diagnostic Radiology) development of cancer post-latency: additive (absolute) and mul- See Bremsstrahlung tiplicative (relative). Genetic Damage: Genetic damage is caused by damage to Radiation, Cˆerenkov the DNA in the gonads due to radiation exposure. Damage to the (Radiation Protection) See Cˆerenkov radiation DNA leads to an increase in the probability of hereditary dis- eases; these may be dominant (first generation), recessive (later Radiation damage generations) or x-linked (first generation). (Radiation Protection) Radiation damage may be classified into Teratogenic Damage: Effects have been observed in chil- two types of effects: dren who were in utero during severe radiation events such as the Japanese atomic bomb. These observations suggest that damage 1. Direct effects to the foetus caused by radiation may lead to effects such as men- 2. Indirect effects tal retardation. Related Articles: Radiolysis, Bioeffects, Deterministic Direct effects tend to be caused by charged particle ionising effects, Stochastic effects radiation such as protons, electrons and alpha-particles. The radi- ation is absorbed directly by the critical target, especially DNA in Radiation detection systems the nucleus of the cell. Direct damage to the DNA in the form of (Radiation Protection) Radiation detection systems are used for strand breaks may mean that cell functions are impaired and cell the detection of radioactivity, for measuring dose and for identifi- death may occur. cation of the type and the energy of the radiation. Indirect effects tend to be caused by photonic ionising The simplest radiation detection system, for detecting radio- radiation such as gamma rays and x-rays interacting with water activity, consists of a Geiger–Müller counter or semiconductor molecules in the cytoplasm of the cell. The physical interac- detector with a power supply and a number of counts or counts/s tion is followed by chemical reactive stages and a subsequent indicator. Radiation dose 767 Radiation field For example, the detection systems applied x-ray equipment These data objects comprise different templates, which incor- in diagnostic radiology with semiconductor detector measures porate an information set required to estimate the radiation dose an entrance dose, dose indicator (for digital radiography), and for to the patient. computer tomography, a dosimetre with a CT chamber and CT This information includes geometric and technique details adapter, measures the computed tomography dose index (CTDI100) (exposure parameters, beam geometry, collimation etc.) and com- and weighted computed tomography dose index (CTDIw). mon dose indicators adopted as the summary metric for a specific The detection system for identification of the type and energy modality (Kerma Area Product (KAP), CT Dose Index (CTDI) of radiation should consist of a detector not only sensitive to the and Dose Length Product (DLP) etc.). type of radiation (α, β, gamma or X) but also able to produce out- Depending on the modality, different sets of templates are put pulses with height proportional to the energy of the radiation. available: The detection systems generally used are proportional gas-filled detectors, scintillation and liquid scintillation counters and ger- • CT Radiation Dose Templates manium detectors. • Projection X-Ray Radiation Dose Templates (XA, RF, Pulse processing and shaping must be done with appropriate CR, DR and Mammography) electronics consisting of preamplifier, amplifier and pulse height • Radiopharmaceutical Radiation Dose Templates (PET, analysis system using single- or multichannel analysers. The energy SPECT, Planar Gamma Camera) spectrum is represented as a function of the measured intensity of radiation against its energy (corresponding to the channel number). RDSR template structure is described in detail for each modality. Abbreviations: CTDI100 = Computed tomography dose index and CTDIw = Weighted computed tomography dose index. Related Articles: Geiger–Müller (GM) counters, Germanium Radiation dosimetry detectors, Liquid scintillation (LS) counting, Proportional (Radiation Protection) See Dosimetry counter, Pulse-height analysers (PHAs) for radiation detectors, Scintillation detector Radiation, electromagnetic Further Readings: Easy Quality Checks with – PTW (Radiation Protection) Electromagnetic (EM) radiation is the Equipment, Code of Practice – Constancy tests of x-ray equipment propagation of energy by simultaneous vibration of electric and in diagnostic radiology. 2006. PTW Freiburg. www .ptw .de; Knoll, magnetic fields. EM radiation is characterised by its wavelength G. F. 2000. Radiation Detection and Measurement, 3rd edn., John (10−13–103 m). Radiation wavelength, frequency and quantum R Wiley & Sons, Inc., New York, pp. 103–119, 307, 619, 685–691. energy are used to define the electromagnetic spectrum, differ- ent ranges of the spectrum are recognised by different names: Radiation dose radio waves, infra-red, visible light, ultraviolet, x-rays and (Radiation Protection) Radiation dose is defined as the energy gamma rays. absorbed per unit of mass. When dose is measured using an The different EM radiation ranges in the EM spectrum are fur- ionisation chamber to calculate radiation exposure in Roentgens ther grouped into ionising and non-ionising radiation. About 34 (R), then the exposure is calculated by dividing the charge in eV of energy is required to produce an ion pair by the process of Coulombs (C) detected by

the ionisation chamber, divided by the ionisation. A quantum of x-rays or gamma rays will always pos- mass of air (kg) within the chamber, that is C/kg. sess at least 34 eV of energy and are referred to as ionising radia- The interactions that occur when an x-ray photon passes tion. However, a quantum of radio waves, infra-red, visible light through an ionisation chamber lead to the absorption of kinetic or ultraviolet rays will possess less than 34 eV and these wave- energy (measured in Joules) within the air molecules. This energy length ranges are referred to as non-ionising radiation. Ionising absorption is termed absorbed dose. Absorbed radiation dose is radiation is used more frequently than non-ionising radiation for measured in Gray (1 R = 8.7 × 10−3 Gy), 1 Gy = 1 J/kg. medical purposes. Absorbed dose may be converted into equivalent dose by The properties of the electromagnetic spectrum and its parts, applying factors to take into account the type of incident radiation. ionisation, ionising and non-ionising radiation are discussed in These radiation weighting factors are defined by the International greater detail elsewhere. Commission on Radiological Protection (ICRP). Equivalent dose Related Articles: Radiation, Gamma rays, Quanta, Ionisation, is measured in Sieverts (Sv). Ionising radiation, Non-ionising radiation, Ultraviolet radiation Equivalent dose, in turn, may be converted to an effective (UV), Infrared radiation (IR) dose by applying factors to take into account the tissue that the Further Reading: Dendy, P. and B. Heaton. 1999. Physics for radiation is incident upon. These tissue weighting factors are also Diagnostic Radiology, Taylor & Francis Group, Boca Raton, FL. defined by the ICRP. Effective dose is also measured in Sieverts. Related Articles: Ionisation chamber, Absorbed dose, Radiation exposure Equivalent dose, International commission on radiation protec- (Radiation Protection) See Exposure tion, Effective dose Radiation field Radiation Dose Structured Report (RDSR) (Radiation Protection) The radiation field is the area irradiated (Radiation Protection) RDSR is the acronym for Radiation Dose by the incident radiation beam and is defined at a set distance Structured Report. from the radiation source. The intensity of the radiation may vary To define a model for the recording and reporting of dose from across the field and may be altered by the use of filtration. The medical procedures, the DICOM Standard has specified a set of size and shape of the radiation field is defined and altered by the Radiation Dose Structured Report (RDSR) data objects. use of collimation. Radiation force 768 Radiation, particle Radiation force (Ultrasound) The radiation force is defined as time-average force acting on a body in a sound field and caused by the sound field. More generally it is the time-average force in a sound field, appearing at the boundary surface between two media of different acoustic properties [1]. Unit: Newton, N. Radiation force is measured by an ultrasound force balance and the relationship between the force F and acoustic power W can be expressed as F = hW/c, where c is the speed of sound and h is a constant depending on the geometry of the used target. Related Articles: Force balance, Radiation pressure, Acoustic power Further Reading: International Electrotechnical Commission, Report IEC 61161 ed2.0 2006. Radiation force balance (Ultrasound) See Force balance Radiation, gamma (Radiation Protection) See Gamma radiation FIGURE R.2 An example of a star-shot film acquired to confirm the Radiation hazard radiation isocentre of a linac. (Radiation Protection) Hazard, or health hazard, is defined as anything that has the capacity to cause harm to a human being. If the hazard relates to radiation exposure, it may be called a radia- isocentre (use the in-room lasers for this). The jaws should be tion hazard. closed to a very small field size (e.g. 1 × 1 cm2 or smaller), and R Related Articles: Hazard, Risk assessment sufficient MUs delivered for each field to give a readable optical density on the film. Without moving the film irradiate at a range Radiation-induced secondary malignancies of different gantry angles (four or five exposures from the differ- (Radiotherapy) Radiation is a carcinogen: it can cause mutations ent quadrants of gantry angle) taking care to avoid any overlap in previously healthy cells, rendering them vulnerable to carcino- between angles. The processed film will have a series of lines genesis i.e. the formation of cancer. Consequently, radiotherapy crossing it and is commonly referred to as a ‘star-shot’ film (an patients can develop secondary malignancies as a result of their example is given below in Figure R.2). Mark the centre line of treatment. Radiation-induced malignancies typically take years each beam on the film. The point (or rather small region) where or even decades to develop. Consideration of secondary cancer the central axis of all the fields overlap is the radiation isocen- risk from radiotherapy is becoming more important as life expec- tre, and should ideally be no more than 1 mm in width, that is a tancy post-treatment increases, particularly amongst paediatric radius less than 0.5 mm. patients, who are more inherently sensitive to radiation-induced Related Articles: Isocentre, Mechanical isocentre carcinogenesis than adults. Long-term follow-up will be essential in helping to determine whether changes in therapeutic technique Radiation monitoring and radiation doses affect the incidence of radiation-induced sec- (Radiation Protection) In addition to what is described under ondary malignancies. Dose monitoring, radiation monitoring requires a more general Related Articles: Adverse effects (Radiotherapy), Repair, approach including: identification of sources and the level of haz- Free radical, Secondary malignancies ard and risk, the availability of adequate instrumentation and of know-how and trained personnel, the identification of roles and Radiation, infrared responsibilities, the formulation and implementation of programs (Radiation Protection) See Infrared radiation and the evaluation of program effectiveness. Related Articles: Dose monitoring Radiation, ionising (Radiation Protection) See Ionising radiation Radiation, nuclear (Radiation Protection) Radiation describes any process in which Radiation isocentre energy emitted by a source in the form of particles or electro- (Radiotherapy) This is the point (rather small spherical volume) magnetic waves travels through a medium or through space. It in space about which all the radiation beams from different gan- also refers to the energy itself. ‘Nuclear radiation’ is when the try angles overlap, and should be in good agreement with the energy emitted is the result of a nuclear process, such as radioac- mechanical isocentre. Prior to performing this test the photon tive decay or by the nuclear transformation of matter by bombard- collimator symmetry should be checked for two collimator angles ment with particles. 180° apart. Related Articles: Radioactive decay, Radiation To find the radiation isocentre place a piece of radiographic film vertically in-between some sheets of Perspex on the treat- Radiation, particle ment couch such that the centre of the film is roughly at the (Radiation Protection) See Particle radiation Radiation, penetrating 769 Radiation protection supervisor (RPS) Radiation, penetrating generating equipment operating below 30 kV and with a dose (Radiation Protection) See Penetrating radiation rate no greater than 1 μ/Svh at 1 m from any surface of the equipment. Radiation, positron Although the employer may wish to consult the RPA about (Radiation Protection) See Positron decay any aspect of the Regulations, it is compulsory for the employer to consult the RPA on the issues listed in Schedule 5 of the Radiation pressure Regulations, which are reproduced here: (Ultrasound) In acoustics, the radiation pressure is normally defined to be the pressure force that a beam of sound inflicts on a 1. The implementation of requirements as to controlled target, or at the interface between two media. The force exerted and supervised areas. on the target can be plane, conical or spherical and is related to 2. The prior examination of plans for installations and the the ultrasound intensity or total power. There are two main mod- acceptance into service of new or modified sources in els for describing the theoretical background of the force. relation to any engineering controls, design features, The Rayleigh radiation pressure is known to be the force per safety features and warning devices provided to restrict unit area on an absorber. The theory is based on a model repre- exposure to ionising radiation. senting oscillations in a sealed tube, where one end of the tube is 3. The regular calibration of equipment provided for a sound source and the other end the absorber. monitoring levels of ionising radiations and the regular If the source and absorber are placed in a fluid, unlimited in checking that such equipment is serviceable and cor- all directions, the resulting force on the absorber will differ. This rectly used. force is known as the Langevin radiation pressure. The magni- 4. The periodic examination and testing of engineering tude of the radiation pressure is proportional to the kinetic energy controls, design features, safety features and warning density in the sound wave. devices and regular checking of systems of work pro- Related Articles: Radiation power, Force balance vided to restrict exposure to ionising radiation. Radiation, primary The appointment of an RPA must be made in writing by the (Radiation Protection) See Primary radiation employer. The employer is expected to provide the RPA with adequate information and facilities to enable him to undertake Radiation protection his/her work. R (General) Radiation protection is a rather general term. Radiation The qualifications and experience required for an individual is a form of energy that has always been around in nature and will to become an RPA are described by the health and safety execu- always influence human life (natural radiation). In addition to the tive (HSE). A person wishing to gaining certification as an RPA general presence in the environment, humans have found many needs to demonstrate to an HSE recognised certification body forms of specific use of ionising radiation (manmade), for exam- that he/she possesses the knowledge and experience to enable ple energy production, industry and medicine (not to talk about him/her to undertake the work satisfactorily. Once a certificate atomic bombs). When the effects of ionisation and non-ionisation is issued to the individual, it must be renewed at least every 5 radiation were studied, a necessity to put in place regulations and years, involving the person demonstrating continuing profes- control systems arose as a consequence, in order to limit the nega- sional development. tive effects of radiation and optimise its use. In order to assess the impact of radiation exposure properly, it Radiation protection officer (RPO) is essential to introduce appropriate quantities and units, for the (Radiation Protection) The radiation protection officer (RPO) various kinds of radiation, which can then be used for the quanti- is a professional of proven competence in radiation protection fication of exposures (see Dosimetry). In principle, the main aim matters relevant for a given type of practice, who is appointed by of radiation protection is the control of radiation exposure, while the registrant or licensee in order to ensure that the requirements radiation dosimetry deals primarily with the measurements of rel- of the national standards are applied. evant radiation quantities, in particular doses. National radiation protection laws and regulations normally In summary, in relation to the different kinds of radiation, it include role, duties and responsibilities of the RPO and/or is essential to know the relevant radiation characteristics defini- qualified expert (see also Qualified expert) or other designated tions, quantities, interpretations and effects. The study of short persons with specific responsibilities in radiation protection. and long-term effects on the environments and human beings Different national regulatory authorities and national laws and along with cost-benefit evaluation (exposure of population, radia- regulations can use different terminologies. tion workers and patients) are the basis for the implementation of Related Articles: Qualified expert, Regulatory authority an appropriate radiation protection system. Further Reading: IAEA (International Atomic Energy Agency). 1996. International Basic Safety Standards for Radiation protection adviser (RPA) Protection against Radiation and for the Safety of Radiation (Radiation Protection) In the United Kingdom, the Ionising Sources, Safety Series No. 155, International Atomic Energy Radiations Regulations 1999 (IRR99) requires employers to Agency, Vienna, Austria. appoint and consult a radiation protection adviser (RPA) if the work they undertake involves the use of ionising radiation Radiation protection supervisor (RPS) (whether electrically produced in x-ray tubes, etc. or from work (Radiation Protection) In the United Kingdom, the imple- with radioactive substances). There are a few exceptions to this mentation of the Basic Safety Standards Directive (96/29/ requirement based

on very-low-hazard work – for example, Euratom) is achieved by The Ionising Radiations Regulations radioactive substances with very low specific activity, or x-ray 1999 (IRR99). Regulation 17(4) of the IRR99 requires a Radiation quality 770 Radiation, secondary ionising radiation protection supervisor (RPS) to be appointed for any A personnel monitor or personal radiation alarm is a dose-rate designated radiation work area that has been made subject to monitor to be carried in a pocket. Local Rules – a document setting out procedures for ensuring A survey meter is a compact, portable radiation intensity mea- that no one entering the area receives an unacceptably high suring instrument. It can be used for monitoring adequate shield- radiation dose. The role of the RPS is to ensure anyone work- ing, locating areas that can have been contaminated, etc. ing in the area does so in compliance with the Local Rules. Further Reading: McCall, R. and J. A. Wall. 1967. Radiation However, legal responsibility for compliance remains with the safety instruments, Chapter 8, in G. Hine (ed.), Instrumentation employer. in Nuclear Medicine, Academic Press, New York, pp. 163–179. The RPS is appointed by the employer and should be named in the Local Rules. There may be more than one RPS for an area and Radiation safety officer (RSO) an RPS may be responsible for more than one area. (Radiation Protection) The term radiation safety officer is used The Approved Code of Practice suggests that a suitable RPS specifically in the United States of America. The RSO is the person should within an organisation responsible for the safe use of radiation as well as regulatory compliance. An organisation licensed by the • Know and understand the requirements of the Nuclear Regulatory Commission to use radioactive materials Regulations and Local Rules relevant to the work with must designate a radiation safety officer in writing. ionising radiation The requirements for a radiation safety officer (RSO) • Command sufficient authority from the people doing vary with the type of license and types of materials used. the work to allow them to supervise the radiation pro- Requirements are laid down in the regulations of the US Nuclear tection aspects of that work Regulatory Commission (NRC) regulations and Regulatory • Understand the necessary precautions to be taken and the guides (NUREG), reports and brochures from the US Nuclear extent to which these precautions will restrict exposures Regulatory Commission (USNRC). • Know what to do in an emergency The term radiation safety officer is sometimes used in other countries for a person within an organisation who has some In general the RPS will be an employee of the employer undertak- level of responsibility for radiation protection. The duties may ing the work, in a line management position and closely involved be similar to those of a radiation protection supervisor or of a with the work being done. However the RPS does not have to be R broader responsibility comparable to a radiation protection present and directly supervising the work being done. The RPS officer. National legislation and guidance should be checked to should receive appropriate training that reflects the complexity of ensure that the correct national terminology is used. the work undertaken, so they can fulfil their task. Related Articles: Nuclear regulatory commission (NRC), In practice the RPS often acts as the focus point for radiation Radiation protection officer (RPO), Radiation protection protection issues that may arise in the work area and will need supervisor (RPS) to liaise with the employer and radiation protection advisor to Hyperlink: US Nuclear Regulatory Commission (NRC). ensure compliance with regulations and safe working practices http://www .nrc .gov/ are maintained. Related Articles: Local rules, Radiation protection adviser Radiation, scattered Radiation quality (Radiation Protection) Scattered radiation refers to all pho- (Radiation Protection; General) This is the quantity given to tons and charged particles resultant from scattering interaction each type of ionising radiation to quantify the relative effective- between an incident photon or particle of ionising radiation and ness in causing damage when the radiation interacts with cells. the medium. This is also known as the radiation weighting factor. Scattered radiation may travel in any direction away from Related Articles: Radiation weighting factor the site of interaction. Therefore further ionisations and energy deposition in the medium may occur outside the primary Radiation safety instrument radiation beam. Interactions between scattered radiation and the (Nuclear Medicine) In each laboratory where one is using ion- medium are important for radiation protection purposes – they are ising radiation, one needs to have radiation safety instruments, important in radiotherapy treatment planning when attempting to designed to monitor the radiation environment in each laboratory. deliver the prescribed dose to the target volume (tumour) whilst There are several types of instruments. minimising the peripheral dose to surrounding healthy tissues A dose-rate monitor gives continuous indication of the radia- due to such scattered radiation. tion intensity at one or more points in the laboratory where the Scattered radiation may also exit the medium altogether. Such radiation intensity may change owing to the work carried out in transmitted scattered radiation is important in diagnostic radiology the laboratory. where it may be detected by the imaging receptor, contributing to An area monitor is a dose-rate monitor that is usually perma- the ‘fogging’ of the x-ray image. Scattered radiation is minimised nently installed in a nuclear medicine laboratory. Its purpose is in diagnostic radiology by the introduction of anti-scatter bucky to check the radiation intensity in such laboratories as hot lab, grids between the patient and the image receptor. accelerator rooms, storage rooms, etc. Related Articles: Secondary radiation, Secondary electron, A laboratory bench monitor is a semi portable dose-rate Peripheral dose, Bucky grids monitor that is ordinary kept in the laboratory work area. This is especially important when working with open radiation sources Radiation, secondary ionising as in a radio chemistry laboratory. It is also useful for making (Radiation Protection) Secondary ionising radiation refers to rough radioassays, for checking the worker’s hands for contami- all photons and charged particles resultant from an interaction nation, as well as the laboratory working area for radioactive between an incident photon or particle of ionising radiation and spill. the medium, which still have enough energy to cause further Radiation shielding 771 Radioactive decay ionisations. Therefore secondary ionising radiation does not The composition is mostly clay-earth with a density of 1.6 g/cm3. include any secondary photons or charged particles which are Iron/Steel: The density is circa 7.8 g/cm3 and its contribution non-ionising, for example ultraviolet radiation. has to be considered. The contribution from these materials in the Secondary ionising radiation may travel in any direction away equipment must be considered. from the site of interaction. Therefore further ionisations and Ledite: Ledite has a density of about 4 g/cm3 and is a mixture energy deposition in the medium may occur outside the primary of lead in concrete bricks. The availability in bricks makes it pos- radiation beam. Such secondary interactions are important for sible to use this material also as additional self-supporting shield. radiation protection purposes – they are important in radiother- Different materials can be used; the thickness required varies apy treatment planning when attempting to deliver the prescribed very much. Just to give an example, for energy in the 500 kVp dose to the target volume (tumour) whilst minimising the periph- spectrum, the 10th value layer thickness, TLV, is 1.19 cm for lead eral dose to surrounding healthy tissues. and 11.7 cm for concrete with a density circa 2.4 g/cm3. Related Articles: Secondary radiation, Secondary electron, The presence of corners might present particular problems and Peripheral dose it is important to create overlapping of the protective material. The same applies for the frames around windows and doors. Radiation shielding The simple calculation of shielding is not enough; it is (Radiation Protection) Shielding is one of the methods, together compulsory to verify this with measurements at the barriers with reducing time and increasing distance (see also Time–dis- at the moment of their construction and also regularly, with tance–shielding), to reduce radiation exposure. When the opti- adequate measuring equipments and procedures, as part of the misation of time and distance is not sufficient, there is a need departmental monitoring. to provide shielding barriers. The shielding barriers protect staff, Related Articles: Time–distance–shielding (TDS) rules, Lead patients (when not being examined/treated), visitors and public, content person working adjacent to or nearby. The intensity of radiation Further Reading: IAEA (International Atomic Energy falls exponentially going through the shielding. Agency). 1998. Design and Implementation of a Radiotherapy In general the basic information needed to calculate a barrier Programme: Clinical, Medical Physics, Radiation Protection are: (1) equipment type, (2) workload (W), (3) target dose (D), and Safety Aspects, IAEA TECDOC-1040, International Atomic (4) use factor and direction of primary beam (U), (5) distance to Energy Agency, Vienna, Austria. the area of interest (d), (6) occupancy of area to be shielded (T), Hyperlinks: International Radiation Protection Association (7) limit values in area to be shielded (P). The evaluations are (IRPA). http://www .irpa .net/; International Atomic Energy R made on the basis of possible working conditions, in such a way Agency (IAEA). www .iaea .org/ that, there should be a reasonable safety margin. Under shielding is worse than over shielding, but shielding is usually expensive, Radiation, ultraviolet therefore it should be reasonably evaluated. There are many (Radiation Protection) See Ultraviolet radiation computer programs available to calculate the thickness of various materials. The 10th value layer thickness (TVL) for different Radiation weighting factor material is usually given. (Radiation Protection) Different types of ionising radiation (e.g. The problems related to shielding in diagnostic radiology and x-rays, alpha particles, beta particles and gamma rays) each have a nuclear medicine are clearly different from those in radiotherapy different relative biological ability to induce damage to biological where a potentially lethal dose is given to patients. molecules, including DNA, dependent on the density of ionising The choice of material depends very much on the setting events caused by the radiation. This ability, or effectiveness to and kind of radiation. There are also situations when improper cause damage, is used in the conversion of the absorbed dose to shielding can make the situation worse. a tissue or organ to the equivalent dose to that tissue, and can be Typical shielding situations are as follows: represented numerically in the form of a radiation weighting fac- tor. For more information, see the articles on Equivalent dose and 1. Low-energy gamma and x-ray can easily be shielded by Relative biological effectiveness. lead or by any other material if used in sufficient amount. Related Articles: Equivalent dose, Relative biological 2. For high energy (more than 500 keV) gamma and x-rays, effectiveness concrete (high-density concrete) is recommended, as it is cheaper and self supporting. 3. Electrons are shielded appropriately when photons are Radio waves accounted for. (General) See Electromagnetic energy spectrum 4. Neutrons from high-energy linacs need special consideration. Radioactive decay (General) Nuclides that do not have a stable combination of neu- Lead: Lead is a high-density (11.3 g/cm3) material with a high trons and protons undergo spontaneous radioactive decay. This atomic number, requires small space and is a good shielding for transformation involves either the expulsion of a charged particle low-energy x-rays. Wooden panels with lead inside can be added or the capture of an electron by the nucleus, with the result that the to already existing walls. It is relatively expensive and difficult to nucleus is either stable or more stable (lower-energy state). This work with as it can easily slide. process alters the balance of neutrons to protons and results in an Concrete: Concrete is the best material when used during con- atom of a different element. The nucleus may be left in an excited struction. The density can vary from 1.8 to 3.7 g/cm3 for the high- state by the transformation and will in most cases transition to density kind, and has to be tested. It is a cheap material. lower-energy (ground) state by emission of one or more gamma Brick: The density of bricks can vary very much depending on rays; in some cases, however, internal conversion (IC) occurs with the shape, for example solid or with holes. the emission of one or more electrons from the atom (Figure R.3). Radioactive materials 772 R adiobiology Particle

Gamma ray usually divided into different categories such as high-level waste (from the reprocessing of spent nuclear fuel), intermediate-level waste and low-level waste (generally in the form of radioactively contaminated industrial, medical or research waste). Special categories are also used such as spent nuclear fuel from reac- Parent Daughter in Daughter in tors and naturally occurring radioactive materials. The detailed excited state ground state specification of these categories is laid down by national regula- tory authorities. The Basic Safety Standard of the International Atomic Energy FIGURE R.3 Radioactive decay. Agency (1996) defines radioactive waste as ‘Material, whatever its physical form, remaining from practices or interventions and for which no further use is foreseen (i) that contains or is contami- Abbreviation: IC = Internal conversion. nated with radioactive substances and has an activity or activity Related Articles: Isotope, Radioisotope, Gamma ray concentration higher than the level of clearance from regulatory requirements, and (ii) exposure to which is not excluded from the Radioactive materials Standards’. (Radiation Protection) Radioactive materials are composed of, or Related Articles: Radioactive material, IAEA contain, one or more radioactive substances in sufficient quanti- Further Reading: International Atomic Energy Agency ties to constitute a hazard to health. The radioactive substance, or (IAEA). 1996. International Basic Safety Standard for Protection substances, may be naturally occurring or artificially produced against Ionizing Radiation and for the Safety of Radiation by nuclear fission or by bombardment by neutrons or ionising Sources, Safety Series No. 115, IAEA, Vienna, Austria. radiation. National authorities define the levels of radioactivity above Radiobiology which a material is classified as being a radioactive material. Also (Radiotherapy) Radiation biology or radiobiology is the study of some national authorities will exclude radioactive waste from the the interaction of radiation with biological systems, ranging from definition of radioactive waste. cells to biological tissues and whole organisms. It brings together Radioactive materials are often classified in different types R the disciplines of physics, biology and chemistry and covers the such as unsealed radioactive sources and sealed sources. effects of both ionising and non-ionising radiation. Therefore it Related Articles: Radioactive substance, Radioactive waste has relevance in many areas of medical physics, including diag- nostic radiology, radiation protection, MRI, ultrasound, nuclear Radioactive series medicine and radiotherapy. (Nuclear Medicine) See Parent–daughter decay Radiation can be both of benefit and harm to the population, for example it can both cure and cause cancer. Due to the large Radioactive sources amount of radiobiological data collected since x-rays were discov- (Radiation Protection) A radioactive source is any object or mate- ered in 1895 by Wilhelm Roentgen, the balance of benefit and det- rial which emits ionising radiation because it contains a radioac- riment can be estimated with a good degree of accuracy. Further tive substance. Radioactive sources are manufactured or produced information on the biological effects of radiation from a radiation for a specific purpose and are divided into two classes – sealed protection point of view can be found in the article on Bioeffects. sources that utilise radioactive materials that are firmly contained In radiotherapy, radiobiology has aided in or bound within a suitable capsule or housing and unsealed radio- active sources that may have a solid, liquid or gaseous form. • The identification of the mechanisms and processes National regulatory authorities specify licensing, labelling underlying the response of tumours and normal tissue and other requirements for the manufacture, use and disposal of to irradiation (for further details see the article on The radioactive sources. 5 Rs of radiobiology) The International Atomic Energy Agency (2003) has catego- • The development of new treatment strategies such as rised radioactive sources to identify those types that require par- hyper- and hypofractionation, the use of radiosensitisers ticular attention for safety and security reasons. and high-LET radiation, all intended to maximise the Related Article: Radioactive materials therapeutic effect of radiotherapy Further Reading: International Atomic Energy Agency • The selection of treatment schedules using dose– (IAEA). 2003. IAEA TECDOC 1344, Categorization of radioac- response models, for example linear quadratic model tive Sources (revision of TECDOC 1191), IAEA, Vienna, Austria. based formulae, to calculate the effect of changes in fractionation or dose rate Radioactive tracer (Nuclear Medicine) This is the common name for tracer com- While the linear quadratic model and its associated formulae have pounds labelled with a radioisotope, for example 99mTc, 125I and been relatively successful in providing a quantitative evaluation 18F. In nuclear medicine radioactive tracer is often referred to as of the clinical situation, it is not without its limitations (see the ‘tracer’. A more elaborate explanation on tracers can be found in article on Linear quadratic [LQ] model) and clinical trials are the separate article Tracer. still essential for the final selection of treatment protocols. Related Article: Tracer Abbreviations: LET = Linear energy transfer and MRI = Magnetic resonance imaging. Radioactive waste Related Articles: Radiation biology, Bioeffects, The 5 (Radiation Protection) Radioactive waste is radioactive mate- R’s of radiobiology, Hyperfractionation, Hypofractionation, rial for which no further use is foreseen. Radioactive waste is Radiosensitisers, Therapeutic effect, Relative biological Radiobiological models 773 R adiochemical purity effectiveness, Dose–response models, Linear quadratic (LQ) L = Latent initiation model, Fractionation, Dose rate dependence period Further Readings: Hall, E. J. and A. J. Giaccia. 2006. Radiobiology for the Radiologist, 6th edn., Lippincott Williams & Wilkins, Philadelphia, PA; Nias, A. H. W. 1998. An Introduction to Radiobiology, 2nd edn., John Wiley & Sons Ltd, Chichester, UK; Steel, G. G. 2002. Basic Clinical Radiobiology, 3rd edn., Arnold Publishers, London, UK. (a) L Time Radiobiological models (Radiation Protection) Radiobiological models refers to a series of models of varied complexity designed to describe the interac- tion of ionising radiation with cellular DNA and other biological molecules, and to predict the damage that will be caused by a given radiation dose, whether delivered acutely at one time, or chronically over a period of time. The models are based on an understanding of the interaction of ionising radiation with biological molecules, which can be (b) L Time classified as either ‘direct action’ or ‘indirect action’. Direct action interactions are those where the incident photon FIGURE R.4 (a) Additive risk model. (b) Multiplicative risk model. or charged particle directly ionises an atom in the DNA molecule, leading to molecular bond dissociation and a strand break. Indirect action involves the physical interaction of the radiation Additive Risk Model: Post-latency, exposure to radiation will with water molecules in the cytoplasm of the cell, leading to a lead to the development of cancer independent of the spontaneous chemical reaction which in turn causes biological damage to the rate of development. DNA and the cell. For more information on this indirect action, The risk associated with radiation exposure varies according see the article on Radiolysis. to sex and the age at exposure. When there is damage to DNA as a result of the direct or indi- Multiplicative Risk Model: Post-latency, the time for cancer R rect action of ionising radiation, the likelihood is that the damage to develop depends on a constant or time-varying factor based on will be repaired. However, if the damage is not repaired, then that the age dependant incidence of natural cancers in the population. damage may be expressed as either prompt deterministic effects The multiplicative risk model appears to be more accurate due to cell killing or late stochastic effects due to cell mutation. when compared with exposure data. All current radiobiological models are based on the interpreta- Related Articles: Radiolysis, Radiation damage, Repair tion of data from exposure of persons to higher doses of radiation of radiation damage, Deterministic effects, Stochastic effects, – the survivors of the atomic bombs at Hiroshima and Nagasaki, Linear no-threshold model, Linear-quadratic model, Bystander and from radiotherapy treatments amongst others. This leads to a effects, Adaptive responses and hormesis, Latent period ‘gold standard’ of known radiation damage and response at doses from approximately 0.1–4 Sv. However much of the framework Radiochemical purity of radiation protection is intended to be used for persons being (Nuclear Medicine) The radiochemical purity (RCP) of a radio- occupationally exposed at much lower doses and dose rates. Thus pharmaceutical preparation is defined as the ratio of the activity it has been necessary to develop models to extrapolate back from between the desired chemical form and the total activity of the the gold standard levels of dose down to zero above background. preparation. The RCP value is expressed as a percentage. Examples of the models developed to extrapolate back from Radiochemical impurities are undesirable because it may the gold standard of dose–response include the linear no-threshold result in an unwanted distribution in the body, influencing the use- model used as the basis of the framework for radiation protection, ful information in the study due to high background activities in and the linear-quadratic model used in the radiobiology of areas not of primary interest, as well as unwanted absorbed doses radiotherapy. to organs and tissues of the patient. Further refinements include currently developing descriptions Free 99Tcm – sodium pertechnetate present may readily be seen of the response of cells at low doses such as bystander effects, in the scintigraphy as increased uptake in the thyroid, salivary adaptive responses and also hormesis – the idea that a little glands, stomach and in the bladder. radiation exposure is beneficial – it ‘does you good’. A number of sources can cause poor radiochemical purity, for Each model is then a combination of the description of a dose– example oxidation–reduction reactions, chemical changes during response curve, and the existence, or non-existence of a threshold storage because of changes in pH, temperature or light, compet- for harm/risk. Examples of such dose–responses are the linear ing chemical reactions during labelling, preparation techniques no-threshold dose–response (LNT), the linear dose–response, and radiolysis. and the non-linear dose–response models. Acceptable RCP values are listed in the package insert or in One common aspect of the models is the description of a the European pharmacopoeia, and must be fulfilled throughout latent period before the biological effects in terms of damage are the useful life of the radiopharmaceutical. expressed in the person exposed. The excess risk expressed in an A number of chromatographic systems and methods are in use exposed population may be described as addition to the normal for testing the RCP. However, the most commonly used system is incidence rate, or multiplicative of the normal incidence rate. thin-layer chromatography (TLC). Figure R.4 illustrates the difference between additive and multi- Abbreviations: RCP = Radiochemical purity and TLC = plicative risk models. Thin-layer chromatography. Cancer incidence Cancer incidence Radiochromic film 774 R adiofrequency screening Related Articles: GMP, Quality control, Radionuclide purity, Radiofrequency heating Chemical purity, Biological purity (Magnetic Resonance) The majority of the RF power transmitted Further Readings: European pharmacopeia, European by the RF transmitting coils for MRI or MRS is transformed into Directorate for the Quality of Medicines (EDQM), Council of heat within the patient’s tissues as a result of resistive losses. This Europe. http: / /www .edqm .eu /s ite /H omepa ge -6 2 8 .htm l; Kowalsky, energy absorption is described in terms of the specific absorp- R. J. and S. W. Falen. 2004. Radiopharmaceuticals in Nuclear tion rate (SAR) measured in watts per kilogram. Significant whole Pharmacy and Nuclear Medicine, 2nd edn., American Pharmacists body heating has a major impact on the cardiovascular physiology Association, Washington, DC; Saha, G. B. 2004. Fundamentals and thermoregulatory ability and heating consequences depend of Nuclear Pharmacy, 5th edn., Springer, New York; Zolle, I. on the status of the patient’s thermoregulatory system and the ed. 2007. Technetium-99m Pharmaceuticals – Preparation and presence of cardiovascular disease, hypertension, diabetes, fever, Quality Control in Nuclear Medicine, Springer, Heidelberg, old age, and obesity. The thermophysiologic responses of the Germany. patient to heating involve complex factors such as: the duration of exposure, the RF energy absorption rate, and the heat conversion Radiochromic film in the presence of various events like blood circulation, diffusion, (Radiotherapy) This is a relatively new type of film (a.k.a bulk flow. A rise in core body temperature less than 1°C generally Gafchromic film) that may be used for radiotherapy dosimetry. It is no cause for concern because cell death does not occur until is a coloured film which has an almost tissue-equivalent compo- the temperature exceeds 42°C. The acute exposure of volunteers sition and

produces a colour change when exposed to radiation. to a 64 MHz RF for 30 min at whole body SARs of 2.7–4 W/kg This colouration results from the polymerisation of a dye. The or 16 min to 6 W/kg resulted in small increases of temperature polymer will absorb light and so by using a densitometer is it pos- (0.1°C–0.4 °C), heart rate and localised sweating and increased sible to measure the transmission of light through the film and skin blood flood. The International Electrotechnical Commission relate this to the level of radiation exposure. A major advantage of (IEC) establishes standards concerning particular requirements radiochromic film compared to conventional radiographic film is for SAR limits in magnetic resonance equipment for medical that it is self-developing and therefore negates any need for devel- diagnosis (IEC 601-2-33:1995) introducing also the concept of opers or fixers. different operational modes. In the following table the SAR limits Another useful feature is that the radiochromic film is grain- are reported for the different operation modes. R less, so has a very high spatial resolution and can be used where there is a very steep dose gradient such as brachytherapy and ste- Second reotactic radiotherapy. First Controlled Controlled Further advantages include ease of use (i.e. no need for a dark Normal Mode Operation Operation room, film processor and chemicals, etc.), can be used for very Region Limit Mode Limit Mode Limit high doses, the response is independent of dose rate and it has better energy characteristics except for very low energies (25 kV). Whole body 1.5 W/kg 4 W/kg averaged Above the first The disadvantages are the increased cost of each film and a averaged over over 15 min controlled reduced sensitivity. 15 min mode limits In common with radiographic film, radiochromic film is to be Head 3 W/kg 3 W/kg averaged Above the first used as a relative dosimeter, and with appropriate care of use and averaged over over 10 min controlled conditions it should be possible to achieve a precision of better 10 min mode limits than 3%. Local SAR 8 W/kg head 8 W/kg head and Above the first Related Articles: Radiographic film dosimetry, Gafchromic (averaged and torso, 12 torso, 12 W/kg controlled film over any 1 W/kg in in extremities mode limits g of tissue) extremities averaged over 5 Radiofrequency (RF) averaged over min (Magnetic Resonance) Radiofrequency (RF) is the term used to 5 min describe any oscillating field which an oscillation frequency in the range between 3 Hz and 300 GHz. Most often the term is associated with electromagnetic fields, since mechanical devices The imaging sequence influences the SAR as the latter increases cannot respond to oscillations in this range. The term ‘radiofre- with the square of the flip angle, the size of the patient and the quency’ refers to the fact that this is the range of the electromag- duty cycle of the RF pulses. The SAR also increases with the netic fields used for generating and detecting radio waves. square of the frequency and therefore depends on magnetic field In MRI, radio frequency pulses (RF-pulses) are used to strength. Thus a patient undergoing an MR imaging at 3.0 T will excite the protons. In this case the pulses are generated with a experience nine times the SAR experienced with MR imaging frequency corresponding to the Larmor frequency of the spins in at 1.0 T. The temperature changes associated with RF induced the sample. Note that this field can be generated without having heating depend on the environmental conditions that exist in and any electrical component. The field is often denoted the B around the MR system that include ambient temperature, rela- 1 field and has a magnetic field strength in the microtesla (μT) range, as tive humidity and airflow. Care should be taken when imaging determined from the duration of a 90° RF pulse. patients with metallic objects, which may result in an increased Related Articles: RF pulse, Larmor frequency heating compared with the heating in normal biologic tissues and this could result in burns. Radiofrequency absorption Radiofrequency screening (Magnetic Resonance) See Specific absorption rate (Magnetic Resonance) MRI scan rooms must be screened to pre- Related Articles: Radiofrequency, Specific absorption rate vent radiofrequency signals originating from outside the room Radiogenomics 775 Radiographic film dosimetry Object contrast Object Background Film density Contrast = Do – Dbg Image brightness Contrast = Bo/Bbg FIGURE R.6 Radiographic contrast and its formula. (Courtesy of Sprawls Foundation, www .sprawls .org) FIGURE R.5 Copper RF screening being fitted during an MRI room density values for a radiograph recorded on film or as the ratio of build. brightness values for radiographs on digital displays. Radiographic film dosimetry interfering with RF signal pick up from the body. MRI room (Radiotherapy) Film dosimetry has been used extensively as a designs incorporate a Faraday cage, which encloses the room with convenient and rapid means of measuring dose distributions of a conductive surface or mesh. The cage attenuates RF transmis- electron and photon beams and also for dynamic beams and for sion into the scan room to an acceptable level (Figure R.5). studying the combination of stationary beams treated sequentially. Related Article: Faraday cage Film dosimetry is in particular important for the dosimetry of scanning electron beams where the automated dosimetry systems Radiogenomics using diode or ionisation chambers cannot be easily employed R (Radiotherapy) The term radiogenomics is used in two contexts in because of long accelerator beam times. Singular advantages of radiotherapy. It can refer to: film as a dosimeter are its high spatial resolution, minimum per- turbation of the radiation field and acquisition of 2D data simulta- • Radiation genomics: how genetic variation impacts neously over a large area. upon a patient’s tumour or normal tissue response to Radiographic films consist of silver halide crystals (dimension radiation therapy. 0.2–2 μm), usually AgBr, embedded in gelatin and spread uni- • Imaging genomics: the relationship between a tumour’s formly on a polyester base in a thin sensitive layer. The electrons genomics and its appearance in imaging (i.e. its freed by ionisation process preferentially move into vacancies/ radiophenotype). imperfect sites in the lattice of the silver halide crystals where a silver ion is reduced to silver atom by the process Generally, genomic testing looks for gene alterations anywhere in the genetic code. It can be performed on blood samples (e.g. Ag+ + e- ® Ag to look for inherited genetic changes) or on cancer cells extracted via biopsy (e.g. to identify specific tumour mutations which may producing a ‘latent image’. These silver specks catalyse the con- allow doctors to recommend targeted therapies). version of the whole AgBr crystal when Ag is subject to an appro- Related Article: Radiomics priate developer solution. Then the image is fixed by dissolving the nonionised AgBr grains in a weakly acidic solution of sodium Radiograph thiosulphate. The elemental silver is black and its presence deter- (Diagnostic Radiology) A radiograph is an image produced by mines the blackening of the film that can be quantitatively evalu- projecting x-ray beam through an object, such as a human body, ated using optical densitometry. The optical density is given by and recording the image for viewing later. Radiographs are recorded either on film or various types of electronic or digital I media. Optical Density = log 0 10 I Radiographic accessories where (Diagnostic Radiology) Radiographic accessories include devices I0 is the incident light intensity and consumable supplies that are used along with x-ray equipment I is the intensity after passing through the film to perform radiographic procedures. The optical density is determined by the quantity of converted Radiographic contrast Ag grains and is therefore dependent on the quantity of radiation (Diagnostic Radiology) Radiographic contrast is the contrast of which was incident. The most common experimental set-up in objects or structures in a radiograph that determines their vis- relative dose measurements is to sandwich the radiographic film ibility (Figure R.6). within a phantom of water equivalent material with the film plane The contrast of a specific object or area relative to its back- parallel to the central axis of the radiation beam, taking care of ground can be measured and expressed either as the difference in the perfect alignment of the film edge with the surface of the Radiographic imaging chain 776 Radiographic kV control phantom and avoiding any air gap on either side of the film, that Phys. 34: 2228–2258; Suchowerska, N. et al. 2001. Directional is exerting a pressure on the phantom. To a first approximation dependence in film dosimetry: Radiographic and radiochromic the radiation induced optical density initially increases linearly film. Phys. Med. Biol. 46: 1391–1397. with the exposure to radiations. If the film is exposed to photons of different energies and the optical densities lie within the linear Radiographic imaging chain range it is found that to a first approximation the optical density (Diagnostic Radiology) The radiographic imaging chain is the is proportional to the absorbed dose. Figure R.7 shows a set of series of components and functions that produce a radiograph as dose–response curves in terms of optical density for the Kodak illustrated in Figure R.8. XV2 film for some beam qualities, perpendicular to the central beam axis and for a 10 × 10 cm2 field size at dmax. The films have Radiographic kV control been inserted in a water equivalent phantom for photons and poly- (Diagnostic Radiology) The high-voltage generator (HVG) of the styrene phantom for electrons. x-ray equipment has various functions, one of these being the Further Readings: Mota, H. C. et al. 1990. Film dosimetry: control of kV during radiography. The main parts of the classi- Linearization of dose-response for relative measurements of dose cal radiographic kV control circuit include the line-voltage com- distributions. Phys. Med. Biol. 35(4):565–569; Pai, S. et al. 2007. pensation (through the autotransformer) and kVp selection. One TG-69: Radiographic film for megavoltage beam dosimetry. Med. important sub-function of the line-voltage compensation is the compensation of the voltage drop due to the internal resistance of the mains. 4 a b c d Contemporary x-ray equipment with high-frequency generator e control the kV by varying the frequency of the DC–AC converter. These equipment use detectors to measure the output kV (usually 3 f by a high-voltage divider) and compare it with a value set by the radiographer. Any difference at the output of the comparator trig- gers the power switches of the DC–AC converter and produces 2 current with appropriate frequency. The high-frequency current feeds the high-voltage ferrite transformer which produces the nec- R 1 essary output kV. This is possible because U ~ cf , where c = An 0 1 10 100 1000 Here, Phantom dose (cGy) U is the voltage at the secondary winding of the transformer (kV) FIGURE R.7 Dose–response curves for KodaK XV-2 film in terms of f is the frequency of the current supplying the transformer (Hz) optical density. (a) 120 kVp, 2.5 mm Al HVL, (b) 280 kVp, 1.7 mmCu, (c) c is a constant depending on the cross section of the trans- 3.5 mm Cu HVL, (d) 20 MeV electron beam, (e) Co60, (f) 6 MV photons. former core (A) and the transformer ratio n (secondary/ (Image by Mota, H.C. et al., Phys. Med. Biol., 35(4), 565, 1990.) primary windings) X-ray image contrast Attenuation With Scatter image scatter removed Recorded Processed Displayed Perceived Object Grid Receptor with Viewing Processing physical contrast Anatomical environment conditions Display characteristics and X-ray beam spectrum controls KV 8 5 Filter Anode FIGURE R.8 A radiographic imaging chain. (Courtesy of Sprawls Foundation, www .sprawls .org) Net optical density Radiographic mode 777 Radioisotope cameras A good radiographic kV control would produce high voltage hold the radioisotope at the other end of the molecule. The more with less than 1% error. Usually it is measured during the quality arms in the cage, the more stable the chelation of the radioisotope. control procedure. Error above 5% is normally an indicator of a If the radioisotope breaks loose, then it may end up in the kid- problem in the kV control circuit. neys or bone marrow, causing delayed radiation damage to these The mA control works in a similar way. In this case the com- organs. parator changes the frequency supplying the filament transformer, A new approach to therapy is emerging where radioisotopes

which varies the filament current. This changes the temperature of that emit very short range (80 μm) alpha particles with high LET the cathode; hence, the flow of thermal electrons (tube current, mA). are tagged onto monoclonal antibodies for targeted alpha therapy Related Articles: High-voltage generator, High-frequency (TAT) to form alpha-immunoconjugates (AIC). generator, Voltage drop, High voltage control device Abbreviations: AIC = Alpha-immunoconjugates, LET = Linear energy transfer, RIC = Radio-immunoconjugates and TAT Radiographic mode = Targeted alpha therapy. (Diagnostic Radiology) Radiographic mode is the most often used Related Article: Targeted alpha therapy mode of operation of an x-ray equipment. During this mode the x-ray tube is supplied with pre-set kV, mA, ms or mAs and pro- Radioisotope duces a short x-ray exposure, resulting in a static (momentary) (Nuclear Medicine) Isotopes that are unstable and disintegrate x-ray image of the organ of interest. Usually radiographic mode (undergo radioactive decay) via emission of a particle and/or γ uses short exposures to minimise blurring due to patient organs’ radiation are called radioisotopes. The nuclear composition of a movement. The time of the exposure is usually from 1 to 1000 ms. specific element can vary depending on the number of neutrons In order to produce an image with sufficient grey shades (vary- in the nucleus. For elements with low atomic number (<20) most ing from black to white) this mode applies relatively high anode stable nuclei have a proton to neutron ratio close to 1. For stable current – usually from 10 to 500 mA. The kV selection depends nuclei with atomic number higher than 20 the proton to neutron on the x-rayed anatomical structures and the desired image con- ratio is greater than 1. The stable atoms form a cluster around an trast. Radiographic mode can use single exposure (most often for imaginary line of stability as seen in Figure R.9. a single x-ray image) or multiple exposures (most often sequence If the ratio deviates from the line of stability then the nucleus used in x-ray angiography). tends to be unstable. The process of radioactive decay brings the The other basic mode of operation of x-ray equipment is fluoro- isotope closer to the line. scopic mode (used for observing organs in their dynamics in real R It is possible to produce radioisotopes in facilities with appro- time). This mode uses very low anode current (usually 0.5–5 mA), priate equipment, for example by adding neutrons to the atomic because the fluoroscopic units have image intensifier. The length nucleus using a neutron reactor or removing a neutron by aiming of exposure depends on the diagnosis (often more than 60–200 s). a proton beam towards a target in a cyclotron. The kV applied in this mode also depends on the observed ana- Related Article: Cyclotron tomic regions and the desired image contrast. Although the main parts of an x-ray equipment are the same, it may have different control circuits for the different operating Radioisotope cameras modes. (Nuclear Medicine) In the 1950s and 1960s radioisotope scanners were replaced with radioisotope cameras. One of the first ideas was published by the Swedish group from Lund University where Radiographic rating Johansson and Skanse reported on the use of a multi channel col- (Diagnostic Radiology) Radiographic rating is generally a limator, solid sodium iodine crystal, and blue-sensitive film to clinical term referring to the process of assigning to anatomical or obtain pictures of radioactive distributions in vivo. The sensitivity pathological conditions some numerical values or ‘ratings’ based was very low and not practical for clinical use. on observations in radiographs. In 1956 Hal O. Anger proposed the first scintillation camera where the read out of the scintillation light from the scintillation Radiography (Diagnostic Radiology) In general terms radiography is the method which uses x-rays, which pass through an object (being modulated by the absorption of its structures) and after this 120 interact with a detector (such as film emulsion, or phosphor, etc.) this way producing a medical image. 80 Line of stability Radiography, digital (Diagnostic Radiology) See Digital radiography N= Z (straight line Radio-immunoconjugates 40 through origin) (Radiotherapy) Radioisotopes are used in nuclear medicine pro- cedures for imaging and therapy. Imaging isotopes emit gamma rays, therapeutic isotopes emit low-energy gamma rays or high- Carbon energy beta or alpha radiation. Radioisotopes are chelated to the (6.6) targeting monoclonal antibodies or proteins to form radio-immu- 20 40 60 80 noconjugates (RIC). Atomic number-Z A number of chelating agents are used. These bifunctional molecules have a covalent bond to the antibody and form a cage to FIGURE R.9 Line of stability. Number of neutrons-N Radioisotope scanner 778 RIS/HIS (Radiology/Hospital Information System) crystal was done electronically with PM-tubes and an electronic Radioisotope scanner collimator circuit. The camera was produced commercially and several cam- (Nuclear Medicine) For a radioisotope scanner the purpose of the eras were produced in the mid 1960s. The first scintillation cam- collimator is to limit the field of view of the detector. In this way era (or Anger camera) in Europe was installed at Lund University spatial resolution is achieved. Hospital in Sweden. For a collimator the point source response is determined. Most The first camera was equipped with a pinhole collimator and collimators are constructed such that the holes are focused to one with only nine PM-tubes. point and thus define a focal plane for the scanner. The modern scintillation cameras have the same basic princi- From point source response curves at different depths, a point ples as the Anger camera including the so-called Anger electronic source response pattern may be constructed. If the response is positioning system. Usually these cameras are called gamma symmetrical about the axis, this completely determines the spa- cameras. tial collimator response. Further Readings: Anger, H. O. 1952. Use of a gamma-ray Further Reading: Brownell, G. L., S. Aronow and G. J. pinhole camera for in vivo studies. Nature 170:200; Anger, H. Hine. 1967. Radioisotope scanning, Chapter 16, in G. Hine (ed.), O. 1967. Radioisotope cameras, Chapter 19. In: Instrumentation Instrumentation in Nuclear Medicine, Vol. 1, Academic Press, in Nuclear Medicine, Vol. 1, ed., G. Hine, Academic Press, New York, pp. 393–412. New York, pp. 485–552; Anger, H. A., R. K. Mortimer and C. Related Articles: Radioisotope scanner, Collimator A. Tobias. 1956. Visualization of gamma-ray emitting isotopes in the human body. Proceedings of the First International Radioisotope scanner collimator line source response Conference on the Peaceful Uses of Atomic Energy, Geneva, (Nuclear Medicine) For a radioisotope scanner the performance Switzerland, Vol. 14, p. 204, 1955, United Nations, New York; of the collimator to a line source in a plane perpendicular to the Johansson, S. A. E. and B. A. Skanse. 1953. A photographic axis of the collimator is important. The collimator scans on the method of determining the distribution of radioactive material in x-axis across the line source, yielding a bell-shaped response vivo. Acta Radiol. 39: 317. function as with the point source. This shape is identical with the Related Articles: Anger camera, Scintillation camera, Gamma point source function if we have a Gaussian distribution for the camera x- and y-dependence. Related Articles: Radioisotope scanner, Scintigraph, R Radioisotope scanner Collimator (Nuclear Medicine) Radioisotope scanning is also known as recti- Further Reading: Brownell, G. L., S. Aronow and G. J. linear scanning or scintigraphy. After the introduction of thyroid Hine. 1967. Radioisotope scanning, Chapter 16, in G. Hine (ed.), uptake measurements with radioiodine the scanning technique Instrumentation in Nuclear Medicine, Vol. 1, Academic Press, was developed for the thyroid to identify hot and cold nodules. New York, pp. 399–401. Thereafter scanners were constructed for scanning larger areas of the body. Then mostly the lungs, liver, kidneys, heart and brain Radiological technologist were investigated. (Diagnostic Radiology) See Radiographer For a scanner a high sensitivity and high spatial resolution are desirable. In general the higher the spatial resolution the lower the Radiologist sensitivity (mostly depending on the collimator design). (Diagnostic Radiology) See Diagnostic radiology A radioisotope scanner consists of a focusing collimator and a scintillation crystal to convert the photon energy into a signal. In the conventional mechanical scanner a system of motors and con- Radiology information system (RIS) trols moves the detection system with respect to the patient and (Diagnostic Radiology) Radiology information systems (RIS) are synchronises the readout with the position of the detector. The specialised types of information systems for application in radi- scanning speed and the line spacing are variable in the scanner. ology departments. This is a database driven distributed system The total information contained in a scan is proportional to intended to handle the data entry, storage, exchange and access of the scanning time and independent of the actual number of lines radiology related patient data and images. RIS works either as a so long as they are not spaced further apart than the resolution standalone system or as an integrated part of a hospital informa- distance of the collimator. tion system (HIS). RIS integrates different modules for patient The readout systems in the earlier models were based on data tracking, image handling and reporting: dot records where a stylus was moving over a piece of paper and stamping a dot or a bar for a fixed number of input pulses. • Patient registration data The density of dots was then proportional to the activity in the • Patient tracking records collimators’ field of view. • Result from examinations Another system of readout is photographic recording where • Scanning/imaging data a light is moved over a photographic film and focused to a point. • Other patient or examination related documents and The light intensity is proportional to the count rate. reports In recent systems the reading is stored in a computer memory • Administrative data and the usual image processing and imaging display tools can be used. Radiology information systems support data exchange and Further Reading: Brownell, G. L., S. Aronow and G. J. interfaces to HL7, DICOM, PACS. Hine. 1967. Radioisotope scanning, Chapter 16, in G. Hine (ed.), Instrumentation in Nuclear Medicine, Vol. 1, Academic Press, RIS/HIS (Radiology/Hospital Information System) New York, pp. 381–428. (Diagnostic Radiology) See Radiology information system (RIS) Related Articles: Scanning, Scintigraph, Activity distribution and Hospital information system (HIS) Radiolucent 779 Radiomics Radiolucent The heart of a radiometer is the light sensor which is basically (Diagnostic Radiology) Radiolucent is the characteristic of an a photodiode in an inverse polarisation mounting such as the object or material that permits radiation, typically x-radiation, to current induced in the device is proportional to the photons (i.e. penetrate or pass through it. energy) absorbed by the device. Photodiodes are designed as such as to be sensitive to light in Radiolysis a certain region, and their response is then further adjusted by the (Radiation Protection) Although DNA is the critical target for use of light filters. Calibrate hazard broadband radiometers can damage to cells caused by ionising radiation, most physical be used to estimate directly weighted irradiances of sources with interactions between the radiation and the cell occur in the cyto- known spectral output, such as phototherapy sources. plasm involving the absorption of the energy of the radiation by The most common types of radiometers used in healthcare cellular water molecules. The consequences of these interactions have detectors designed to match the response of either ultraviolet may be described by considering the radiolysis of water: or blue light therapy sources. Radiometers can of course also be In the initial physical stage there is an interaction involving used to measure solar light and to estimate factors such as the the ionisation or excitation of the water molecule – this produces ultraviolet exposure index (UV index). a radical ion and a free electron: Detectors of light in the visible and infrared region are also based on temperature probes, such as thermocouples H (Figure R.10). 2O ® H O+ 2 + e- Related Articles: AORD, Light source, Photodiode, Alternatively, the water molecule can simply be raised to an Phototherapy, Temperature probe, UV dosimetry excited state by absorbing the energy of the radiation: Further Readings: Coleman, A., F. Fedele, M. Khazova, P. Freeman and R. Sarkany. 2010. A survey of the optical haz- H ards associated with hospital light sources with reference to the 2O ® H * 2O Control of Artificial Optical Radiation at Work Regulations 2010. In the pre-chemical stage, the H

J. Radiol. Prot. 30(3):469; Czapla-Myers, J. S., K. J. Thome, and 2O+ and H2O* radical ions disas- sociate to produce hydroxyl (HO) and hydrogen (H or H+) ions. S. F. Biggar. 2008. Design, calibration, and characterization of a These hydroxyl and hydrogen ions are highly reactive. field radiometer using light-emitting diodes as detectors. Applied In the chemical stage the products of chemical reactions in the Optics 47(36):6753–6762; Kitsinelis, S. and S. Kitsinelis. 2015. Light Sources: Basics of Lighting Technologies and Applications. R pre-chemical stage migrate in pairs close enough to react chemi- cally. This process leads to a series of chemical reactions to pro- CRC Press; Li, T. H. and G. Y. Luo. 2015. Design of light source of duce, for example peroxide (H2O2) and hydroxide (OH) – highly agricultural UVALED pest control lamp in food production. Adv. oxidising chemical reagents – as well as water (H2O). If the prod- J. Food Sci. Tech. 9(1):36–39; Vanicek, K. et al. 2000. UV-Index ucts do recombine as water then it can be assumed that no further for the Public. Publication of the European Communities, harm is done. Brussels, Belgium. At the biological stage, once all the free radicals produced have reacted, the resultant oxidising agents can migrate to cause Radiomics direct damage to DNA by reacting and reorganising molecule (Radiotherapy) Historically, radiology has mainly relied on visual bonds between oxygen and hydrogen atoms in the DNA molecule. interpretation, with relatively few quantitative imaging metrics. This biological damage to the cell can be effected in hours, but Radiomics is the extraction of quantitative information from med- may take days, weeks, months or years. ical images, usually images acquired as clinical standard of care. Related Articles: Radiation damage, Radiobiological models Radiomic analyses can reveal image features which provide use- ful diagnostic, prognostic or predictive information, potentially Radiometer, light improving upon conventional imaging metrics such as tumour (Non-Ionising Radiation) A radiometer is an electronic device size and volume. Radiomic analyses have several stages, typically used to measure the total irradiance of an incoherent light source. including: image acquisition, segmentation, feature extraction, FIGURE R.10 Examples of radiometer with separate interchangeable detector head (a) and with built-in head (b). Radionuclide generators 780 R adionuclide purity predictive modelling, and model validation. It is highly desir- one can still obtain maximum activity. However a reduced yield able to harmonise or standardise image acquisitions for radiomic may be eluted before this time if there is need for more activity. analyses. Regarding segmentation, usually the region of inter- Due to the decay of 99Mo, generators are normally used for a week est in radiotherapy is the gross tumour volume which may be and then discarded. contoured manually or using auto-segmentation tools. Features extracted from radiomic analyses are can be divided into two cat- Radionuclide imaging egories: semantic (qualitative features often derived from existing (Nuclear Medicine) See Nuclear medicine imaging guidelines) and agnostic (computational metrics with predefined mathematical formulations). Algorithms from artificial intelli- Radionuclide production gence or statistical learning are often applied to detect correla- (Nuclear Medicine) The production of radionuclides used in med- tions between radiomic features and relevant clinical endpoints icine for molecular imaging or radiotherapy. or biological characteristics. Correlations with genomic traits can One of the fundamental properties of the nuclides used in also be considered, in so-called imaging genomic (or radioge- nuclear medicine is that they are radioactive, that is they decay nomic) analyses. As large numbers of possible radiomic features and radiate particles or photons. The radiation can be used for exist (typically relative to small numbers of patients), care must either imaging and/or therapy. Since there are only a limited be taken to avoid overfitting. Emphasis should be placed on the amount of radionuclides in a sample the activity will decrease validation of radiomic signatures using independent cohorts. with time. It is therefore important for a nuclear medicine centre Related Article: Radiogenomics to have a constant production of radionuclides. The clinical use is dominated by two radionuclides, namely Radionuclide generators 99mTc and 18F. 99mTc is the decay product of 99Mo and is extracted (Nuclear Medicine) Radionuclide generators are used to produce from a technetium generator which is refilled with new 99Mo short-lived radionuclides in centres remote from the site of pro- once every week. 18F on the other hand is cyclotron-produced, duction, using their longer-lived parent radionuclide. The most meaning that 18O-enriched water is irradiated by a proton beam common generator in use is the 99Mo/99mTc generator, which pro- producing 18F. duces NaTcO4 which can then be used directly in labelling pro- For more information regarding the two different radionuclide cedures. A schematic cross-section of the generator can be seen production methods, please see Related Articles. R in Figure R.11. Abbreviations: Mo = Molybdenum and Tc = Technetium. The decay series of 99Mo is as follows: Related Articles: Technetium generator, Cyclotron 99 Mo® 99m Tc + b- + u t1/2 = 66h Radionuclide purity 99m Tc® 99 Tc + g t1/2 = 6.02h (Nuclear Medicine) The radionuclide purity of a radiopharmaceu- tical preparation is defined as the ratio of the activity between the 99mTc is being continuously produced by the decay of 99Mo, but desired radionuclide and the total activity of the radiopharmaceu- continuously destroyed by its own decay. As t tical. It is expressed as percentage and should be below a certain 1/2 (Mo) >> t1/2 (Tc), a transient equilibrium is established where after reaching activity limit to be approved for use. a maximum value, the activity of the daughter follows that of the Impurities of other radionuclides than the one intended are parent. At this stage, the ratio of the activities of the parent and detrimental because it will results in unnecessary absorbed doses daughter becomes constant, and in effect the daughter radioactiv- to organs and tissues of the patient, especially if the impurity has ity decays with an apparent half-life of the parent radionuclide much longer half-life than the desired radionuclide. Unwanted rather than its own. The generator can be eluted every 24 h and radionuclides in radiopharmaceuticals can arise from many fac- tors, including production method, purity of the target (competing nuclear reactions) and incomplete or inefficient separation during the radiochemical processing. Vacuum vial Measurements of radionuclide impurities are best performed Sterile saline by gamma spectroscopy with the use of a Ge(Li) semiconductor detector due to its superior energy resolution in order to sepa- rate the photopeaks. The energy resolution of NaI(Tl)-detectors is not good enough in order to resolve the photopeaks. The Ge(Li)- detector is connected to a multi-channel analyser (MCA). The amount of impurity, expressed in percentage of the total activity, is determined by measurement of the number of counts of the area Mo-99 defining the photopeaks. Tc-99 m Examples of radionuclide impurities are 99Mo in 99Tcm–elu- ates, 110In and 114Inm in 111In-solutions, 124I in 123I and 202Tl in 201Tl. Column Relevant and acceptable radionuclide impurity levels are given in the insert packages and in the European pharmacopeia. The radionuclide impurity requirement for a radiopharmaceutical must be fulfilled throughout its useful life. Lead Most radiopharmaceuticals used in nuclear medicine have been tested for radionuclide purity by the manufacturer and is only performed at the hospital when new radiopharmaceuticals are introduced. The only routine test that is normally required is FIGURE R.11 The Mo-99/Tc-99m generator. measurement of 99Mo breakthrough in a new delivered generator. Radionuclide therapy 781 Radionuclides in radiotherapy Related Articles: GMP, Quality control, Radionuclide purity, Another desired feature of a radionuclide used for tumour Chemical purity, Biological purity therapy is that the uptake in normal tissue should be kept to a Further Readings: European pharmacopeia, European minimum. There are several approaches suggested to minimise Directorate for the Quality of Medicines (EDQM), Council the radiation dose to normal tissue, for example extracorporeal of Europe http: / /www .edqm .eu /s ite /H omepa ge -6 2 8 .htm l; elimination and pre-targeting. Kowalsky, R. J. and S. W. Falen. 2004. Radiopharmaceuticals in Related Article: Extracorporeal elimination Nuclear Pharmacy and Nuclear Medicine, 2nd edn., American Further Reading: Carlsson, J., E. F. Aronsson, S.-O. Hietala, Pharmacists Association, Washington, DC; Saha, G. B. 2004. T. Stigbrand and J. Tennvall. 2003. Tumour therapy with Fundamentals of Nuclear Pharmacy, 5th edn., Springer, New radionuclides: Assessment of progress and problems. Radiother. York; Zolle, I., ed. 2007. Technetium-99m Pharmaceuticals – Oncol. 66(1): 107–117. Preparation and Quality Control in Nuclear Medicine, Springer, Heidelberg, Germany. Radionuclides (Nuclear Medicine) Radionuclide is the term used to describe a Radionuclide therapy nuclide which is in the process of undergoing radioactive decay. (Nuclear Medicine) Radionuclide therapy refers to the therapy Associated with each radionuclide is a set of characteristic proper- technique of labelling specific targeting agents (or radiotracers) ties. These include mode of decay, transition energy and half-life. that participate in specific biological pathways or processes Diagnosis: For nuclear medicine diagnosis the aim is to associated with the malignant cells. The idea is that the provide clinical information while exposing the patient to minimal radiotracers will accumulate intracellular or close to the targeted radiation. Therefore the desirable properties of radionuclides used cells, for example cancer cells so that the radiation will induce in diagnosis are as follows: cell death in malignant cells and at the same time spare non- malignant cells. • Half-life approximately equal to the duration of the Radionuclide therapy is used when treating tumours of haema- investigation topoietic origin but has not been as successful when treating solid • No particulate radiation tumours. Radionuclide therapy is in most cases considered as a • Emission of a gamma ray of energy high enough to complement to external radiotherapy and surgery. travel through tissue and low enough for efficient detec- A few radionuclide therapies are used clinically on a regular tion by a gamma camera (approx. 100–300 keV) basis; 131I is used for therapy of the differentiated thyroid cancer • Readily available R and 32P-orthophosphate is used for therapy of polycythemia and thrombocythemia. Palliative treatments of skeletal metastases An example of such a radionuclide is 99mTc. using 89Sr-chloride, 153Sm-EDTMP or 186Re-HEDP are routinely Therapy: The aim here is to destroy diseased cells with radia- performed. tion; therefore, it is important to locate the radionuclide precisely Many pre-clinical studies have shown promising results when within the tissue to be treated. The desirable properties of radio- using radionuclide therapy but little is known about the way pre- nuclides used in therapy are as follows: clinical trials can be applied to clinical applications. Further Reading: Carlsson, J., E. F. Aronsson, S.-O. Hietala, • Half-life short enough so that hospital stay is limited T. Stigbrand and J. Tennvall. 2003. Tumour therapy with • Emission of particulate radiation with penetration radionuclides: Assessment of progress and problems. Radiother. similar to the size of the lesion to be treated Oncol. 66: 107–117. • Emission of a gamma ray in order to visualise the radio- nuclide concentration in the target tissue Radionuclide uptake in tumour cells (Nuclear Medicine) The ideal tissue uptake is full radionuclide An example is 131I. uptake in the tumour and no uptake in surrounding tissue. The method used for targeting tumour cells depends on the properties of the targeted tumour. If the targeted tumour is a solid tumour Radionuclides in radiotherapy it is possible to use non-tumour-specific agents or processes, like (Nuclear Medicine) The number of radionuclides that are suit- increased glucose metabolism, to target the tumour. Increased able for radiotherapy are limited to the isotopes with suitable glucose metabolism is also used for 18F-FDG tumour diagnosis physical and chemical characteristics, availability and production but the process also offers possibilities for selective uptake for procedures. tumour therapy radionuclides. Two examples of radionuclides that The physical properties of any radionuclide refer to the radio- target high cells with high metabolism are radioactive strontium nuclide half-life and emission spectra. Radionuclides that emit and samarium. However, increased radionuclide uptake due an electrons or high-LET radiation are best suited for therapy pur- increased metabolism might not be the best method to target dis- poses since the energy is deposited close to the decay point and seminated tumour cells. not in distant organs. The main therapy purpose for radionuclides is to work as a The ideal radionuclide has three favourable chemical charac- complement for conventional therapy methods and target dis- teristics: (1) it should be simple to separate a

carrier-free sample seminated tumour cells that are next to impossible to remove without polluting isotopes; (2) easy to label it to a tracer and (3) with surgical methods and/or are unaffected by external radio- show in vivo stability. therapy. One possibility is to target antigens or receptors that To minimise the patient dose the half-life of the radionuclide are over-expressed, or even better, specific for certain tumours. should be low. Therefore, the ideal production site is an in-house Adenocarcinomas (e.g. breast cancers) and in some other tumours accelerator where the activity losses during transport are low (e.g. gliomas and bladder cancers) are examples of tumour cells compared to longer transport from distant production sites. with over-expression of receptors. Related Article: Carrier-free sample Radiopacity 782 R adiosensitivity Radiopacity Most of the radiopharmaceuticals prepared in a radiophar- (Diagnostic Radiology) Radiopacity is the characteristic of macy are based on the use of Tc-99m. The eluate from the techne- an object or material that impedes or attenuates the passage of tium generator is added to cold kits according to manufacturers’ radiation, typically x-radiation, through it. instructions. The design of a radiopharmacy falls under the regulatory con- Radiopaque markers straints of the country in which it is operated. The most important (Diagnostic Radiology) Radiopaque markers are typically made considerations are manufacturing practice, radiation safety issues of lead (or lead rubber) and placed somewhere in the x-ray beam and asceptic dispensing conditions. so they are visible in the image. Letters of the alphabet, ‘L’ and Related Article: Radionuclide generators ‘R’ are used to mark the left and right sides of a patient, especially in mammography. Markers might be used to show the location Radiosensitivity of specific anatomical features, such as scars on the surface of a (Radiotherapy) Radiosensitivity describes the relative sensitivity patient that could be visible in a radiograph. of different cells or tissues in the body to damage caused by ionising radiation. This relative sensitivity is expressed by using Radiopharmaceuticals tissue weighting factors to convert the equivalent dose to a tissue (Nuclear Medicine) Radiopharmaceuticals are pharmaceuti- or organ to the effective dose to the whole body in order to then cals labelled with different radionuclides, for use in the field of estimate the risk to the person of developing a stochastic effect nuclear medicine. They contain tracer amounts (from roughly (i.e. cancer or hereditary disease). one tenth of a ng to a few hundred μg) of molecules or biologi- A cell’s susceptibility to radiation varies with the phase of the cal agents. The applicability of a radiopharmaceutical is deter- cell cycle: least sensitive to radiation late in the S phase and most mined by the characteristics of the pharmaceutical part, that is sensitive in the M and G2 phases. For cells with long cell-cycle the main localisation and metabolism in a given organ or tissue times and significantly long G1 phases, there is a second peak of and the emission properties of the radionuclide. Nearly 95% of radio-resistance early in G1. the radiopharmaceuticals are used for non-invasive imaging using The work of two French radiobiologists Bergonie and photon-emitting radionuclides, whereas the rest are for therapy Tribondeau, published in 1906, indicated that the more rapidly using beta-emitting radionuclides. a cell is dividing, the more sensitive it is to radiation. This is R The most used radiopharmaceuticals for imaging are based often called the law of Bergonie and Tribondeau and applies on the use of the radionuclide 99mTc. Prefabricated kits contain- well to rapidly dividing cells such as haematopoietic stem cells, ing the nonradioactive chemicals are used to produce a specific which are radiosensitive. Indeed, it has been shown that the most 99mTc-radiopharmaceutical after adding the required activity of sensitive cells are those that are undifferentiated, well nourished, 99mTc-pertechnetate. Some other radionuclides, such as 111In, divide quickly and are highly metabolically active. Amongst may be labelled to chelating agents. These are claw-like mol- the mammalian cells, the most sensitive are spermatogonia and ecules that bind with a metallic atom in the middle of its struc- erythroblasts, epidermal stem cells and gastrointestinal stem ture and label a particular molecule, for example an antibody or cells. The least sensitive are neurons and muscle fibres, that is a blood cell. differentiated cells that do not proliferate. However, the law does Since radiopharmaceuticals are administered to humans, not apply in all cases. For example, oocytes and lymphocytes are they have to be sterile and pyrogen free. For human use of a non-proliferating and yet they are very radiosensitive. radiopharmaceutical it must be approved by the regulatory A cell’s radiosensitivity can be determined from cell survival authority of the country in which it is used. The most impor- curves. These describe the relationship between the radiation tant considerations are manufacturing practice, radiation safety dose and the proportion of cells that survive following irradiation. issues and aseptic dispensing conditions. Standardised speci- The shape of survival curves is commonly described by the fications that define the quality of about 60 radiopharmaceuti- linear-quadratic model with the ratio of its two parameters, α and cals for diagnostic and therapeutic purposes are stated in the β, often used as a measure of a cell’s radiosensitivity. High α/β- European pharmacopeia. values represent tissues with limited potential for recovery from Besides the term ‘radiopharmaceutical’, which is the radiation damage, for example skin and most tumours, whereas most common in use, other terms like tracer, radiotracer, and normal tissues with low α/β-values are characterised by their high radiodiagnostic/-therapeutic agent are used in the literature. It recovery potential, for example kidney and spinal cord. should be noted that the term ‘radiochemical’ should not be mis- Other factors affect how sensitive a cell is to radiation. Based taken for a radiopharmaceutical, since the former is not sterile on the law of Bergonie and Tribondeau, rapidly dividing tumour and pyrogen free and not usable for human administration. cells may be expected to be more sensitive to radiation than nor- Related Articles: Kits, Nuclear medicine mal tissue cells. However, this is not always true: tumour cells can Further Readings: European Pharmacopeia, European be hypoxic and therefore less sensitive to radiation that indirectly Directorate for the Quality of Medicines (EDQM), Council of damages DNA through free radicals produced by the ionisation Europe. http: / /www .edqm .eu /s ite /H omepa ge -6 2 8 .htm l; Kowalsky, of oxygen. Therefore, the presence or absence of oxygen dramati- R. J. and S. W. Falen. 2004. Radiopharmaceuticals in Nuclear cally influences the biologic effect of sparsely ionising radiations Pharmacy and Nuclear Medicine, 2nd edn., American Pharmacists such as x-rays but there is no effect for densely ionising radiations Association, Washington, DC; Saha, G. B. 2004. Fundamentals of such as α–particles. Nuclear Pharmacy, 5th edn., Springer, New York. Over the years, many chemical and pharmacological agents have been discovered that modify the response of mammalian Radiopharmacy cells to radiation and two types of sensitiser, halogenated pyrimi- (Nuclear Medicine) The radiopharmacy is a specialist pharmacy dines and hypoxic-cell sensitisers, have demonstrated a differen- service involved in the production of radiopharmaceuticals for tial effect between tumour and normal cells and therefore have nuclear medicine. found practical use in clinical radiotherapy. A number of clinical Radiosensitisers 783 R adium trials are currently in progress using the agents Gemcitabine and agents protect normal cells from radiation-induced damage by Temozolomide. promoting cell repair. Currently, the only well established drug There is evidence of a genetic involvement in radiosensitivity. is Amifostine which reduces the radiation associated dry mouth Many radiosensitive mutants have been isolated from cell lines effect (xerostomia) observed in patients receiving radiotherapy maintained in the laboratory. In many cases their sensitivity to for head and neck tumours. cell killing by radiation has been related to a reduced ability to Related Articles: Oxygen enhancement ratio, Reoxygenation, repair double-strand DNA breaks (DSB). Some human patients Repair have been observed to show an abnormally severe normal tis- Further Reading: Hall, E. J. and A. J. Giaccia. 2006. sue reaction to radiation therapy and exhibit the traits of spe- Radiobiology for the Radiologist, 6th edn., Lippincott Williams cific inherited syndromes. The most striking example is ataxia & Wilkins, Philadelphia, PA. telangiectasia (AT) where the fibroblasts of affected patients are two to three times as radiosensitive as normal. Radiation therapy Radiosurgery for patients with AT should be given with doses reduced appro- (Radiotherapy) See Stereotactic radiosurgery priately to avoid considerable normal tissue damage. Affected patients also have an elevated incidence of spontaneous cancer. Radiotherapy Abbreviations: AT = Ataxia telangiectasia and DSB = Double- (Radiotherapy) Radiotherapy is the treatment of a disease with strand DNA breaks. ionising radiation. The term ‘radiation therapy’ is also used. Related Articles: Alpha beta ratio, Cell cycle, Oxygen enhance- Depending on the distance between the radiation source and ment ratio, Repair, Repopulation, Redistribution, Reoxygenation, the target volume, that is the tissues to be treated, radiotherapy is 5Rs of radiobiology, Surviving fraction divided into two categories: teletherapy and brachytherapy. Further Readings: Bergonie, J. and L. Tribondeau. 1906. In teletherapy the source is far from the target (Greek word De quelques résultats de la radiotherapie et essai de fixation tele, far away), while the source is placed close to the target in d’une technique rationnelle. Comptes-Rendus des Séances brachytherapy (Greek word brachy, short). de l’Académie des Sciences 143: 983–985; Hall, E. J. and A. In teletherapy historically x-ray tubes operated at higher J. Giaccia. 2006. Radiobiology for the Radiologist, 6th edn., voltages have been used (orthovoltage therapy). Now electron Lippincott Williams & Wilkins, Philadelphia, PA. linear accelerators, proton and light ion cyclotrons or synchro- trons, and more rarely now 60Co- or 137Cs-irradiation devices, Radiosensitisers are employed. R (Radiotherapy) Radiosensitisers are agents that increase the Conventional brachytherapy uses sealed sources, and conven- radiosensitivity of cells and therefore increase the lethal effects of tional brachytherapy is the subject covered under this heading. (It radiation if administered in combination with the radiation. is to be noted, though, that systemic treatments using radiophar- Over the years, many chemical and pharmacological agents maceuticals, microspheres, nanoparticles, etc. are also covered by have been discovered that modify the response of mammalian the definition ‘sources placed close to the target’.) cells to radiation. The simplest of these, and the one that pos- Related Articles: Brachytherapy, Teletherapy sibly produces the most dramatic effect, is oxygen. It is known that tumour cells can be hypoxic and therefore less sensitive to Radiowaves radiation that indirectly damages DNA through free radicals (Magnetic Resonance) In MRI, pulsed radiofrequency (RF) produced by the ionisation of oxygen. Therefore, the presence or energy is deposited into the investigated sample for spin excita- absence of oxygen dramatically influences the biological effect tion. In order to create the desired spin excitation, the frequency of sparsely ionising radiations such as x-rays but there is a much- is calculated using the Larmor relation which relates the so-called reduced effect for densely ionising radiations such as α–particles. resonance frequency to the main magnetic field. For protons, this For more information on the effect of oxygen see the article on resonance frequency is 42.6 MHz/T, placing the RF pulses in the Oxygen enhancement ratio. same frequency range as of the electromagnetic fields used for The lethal effects of radiation can be increased by the admin- generating and detecting radio waves. istration of radiosensitising agents. Many chemical and pharma- See also Radiofrequency. cologic agents that modify the response of mammalian cells have Related Articles: Radiofrequency, Larmor frequency been discovered but most offer no practical gain in radiotherapy since they affect tumours and normal tissues alike. There is no Radium point in employing a drug that increases the sensitivity of tumour (Radiotherapy, Brachytherapy) Radium-226 is a naturally occur- cells if it also increases the sensitivity of normal cells to the same ring radionuclide belonging to the decay series uranium-238 – extent. Two types of sensitisers have demonstrated a differential lead-206. Radium decays (alpha) with a half-life of 1600 years effect between tumour and normal cells and have found practical into radon-222, an inert gas with a half-life of 3.82 days. The use in clinical radiotherapy. These are halogenated pyrimidines decay of radium-226 via radon-222 to stable lead-206 results in (differential effect is based on the premise that tumour cells cycle a number of photons with energies up to 2.45 MeV, beta particles faster than normal

cells and therefore incorporate more of the (maximum energy 3.3 MeV) and alpha particles. A radium source drug than the surrounding normal tissues) and hypoxic-cell sensi- (a radium salt mixed with a filler) with a platinum encapsulation, tisers (differential effect based on the premise that hypoxic cells that is filtered by 0.5 mm platinum, has an average photon energy only occur in tumours and not normal tissues). A number of clini- of 0.83 MeV, when the radium is in equilibrium with its daugh- cal trials are currently in progress using the agents Gemcitabine ter products (the exposure weighted average energy is 1.25 MeV). and Temozolomide. The beta and alpha particles from the decay are absorbed in the Conversely, the differential effect of radiation response encapsulation. between tumour and normal cells may be increased by the admin- Types of Radium-226 Sources: Radium was initially used for istration of radioprotecting substances (radioprotectors). These temporary low dose rate implants, its specific source strength is Radium substitute isotope 784 Radon (a) (b) FIGURE R.12 (a) Lund applicator for cervical cancer treatments, intrauterine probe and box, coupled for secure positioning with the probe perpen- dicular to the box, and (b) a radium applicator box showing the loading pattern for radium needles, one dummy radium needle loaded. low, and even low dose rate sources were large in size. Radium the artificially produced radionuclides, miniature sources, after- sources were available as needles and tubes with different dimen- loading techniques, flexibility in applicator design and flexibility sions and source strengths. in time-dose patterns of applications. This is exactly what has Needles and tubes were often loaded into other types of happened since 1985, with the development of computers, imag- applicators. Figure R.12a shows an applicator for treatment of ing techniques, treatment planning systems, etc. cervical cancer, consisting of an intrauterine probe and a box. The ICRU also emphasises the importance of careful evalua- The box was loaded with a number of needles, see Figure R.12b. tion when the treatment technique is different from a traditional R (This specific coupled applicator had source strengths of 3.3 GBq well-established treatment technique: changes in dose and dose + 4.1 GBq [90 mCi + 110 mCi] apparent activity for the probe rate distributions, applicator design and fractionation schedules. and the box respectively. The dose rate at point A was 2.0 Gy/h. Radium is in principal not used today. It has been replaced Dimensions: probe length 6 cm and diameter 6.5 mm, box size by the ‘substitute’ isotopes mentioned earlier, other higher-energy 4.5 × 4.5 × 0.6 cm3.) photon sources and also lower-energy sources (allowing, e.g. per- A radium source, if damaged, could leak the toxic radium manent seed implants without radiation protection problems). salt with its decay products, including the radioactive radon Abbreviation: ICRU = International Commission on Radiation gas. The high photon energy further constitutes a radiation Units and Measurements. protection problem. Radium sources are in principle not used Related Articles: Brachytherapy sources, Iridium-192 today, but the vast amount of clinical experience gained with Further Reading: ICRU Report 38. 1985. Dose and volume brachytherapy using radium sources still makes an impact on specification for reporting intracavitary therapy in Gynecology, modern brachytherapy practises. ICRU, Washington, DC. Radon-222 Sources: The radioactive gas radon-222 was the first isotope used for permanent implants. The gas from the decay Radon of radium-226 was trapped and encapsulated as a small seed. The (General) radon source has a very short half-life, 3.82 days, and the same photon spectrum as radium-226. Related Articles: Brachytherapy sources, Radium substitute Symbol Rn isotope, Intracavitary brachytherapy Element category Noble gas Mass number A 222 Radium substitute isotope Atomic number Z 80 (Radiotherapy, Brachytherapy) In the ICRU Report 38 (1985), Atomic weight 222 kg/kg-atom Dose and Volume Specification for Reporting Intracavitary Electronic configuration 1s2 2s2 2p6 3s2 3p6 3d10 4s2 4p6 4d10 Therapy in Gynaecology, the following is stated: 4f14 5s2 5p6 5d10 6s2 6p6 ‘The replacement of radium by 137Cs, 192Ir and 60Co may be Melting point 202 K accomplished according to two options. In the first option, the Boiling point 211.3 K new sources (mainly 137Cs) are similar in size and have an output Density near room temperature 9.73 kg/m3 similar to radium sources. The same technique of application can then be applied and the clinical experience gained with radium remains fully relevant. The principal advantages of the replace- History: Radon was discovered in 1900 by Friedrich Dorn ment of the radium concerns radiation protection; these include as one of the products of the radioactive decay of radium. At no contamination from leakage and less shielding in the case of room temperatures it is a dense colourless odourless gas which 137Cs and 192Ir’. is radioactive, undergoing alpha decay. The least unstable isotope Compare also the source strength specification ‘mg Ra eq’. of radon, radon-222, has a half-life of about 3.8 days. Radon is The second option, according to the ICRU, lies in the possibil- produced naturally by successive decays of thorium and uranium ity to take advantage of new technology; high specific activities of found in some igneous rocks. The radon gas seeps to the surface RAID technology 785 Ramp time where, due to its high density, it can become trapped in buildings RAM memory and can, in certain locations, accumulate to levels that represent a (General) The RAM (random access memory) of one computer significant radiological health hazard. system is a memory allowing direct access to any byte on the chip. Medical Applications: In early years, radon (222Rn) seeds pre- Usually the RAM is a volatile type memory, which requires con- pared by capturing the radon gas from radium (a radon genera- stant refreshing (e.g. every several ms), but has very short access tor) were used in cancer therapy. Radon seeds were subsequently time (e.g. every several ns). Usually all data processed by the CPU replaced by other sources such as 198Au and 125I. is temporarily stored into the RAM, this way the amount of RAM Related Articles: Radioactivity, Radioactive materials, is of prime importance for the speed of data processing. There are Radioactive decay, Alpha emission many different types of RAM. RAID technology Ramp converter (Diagnostic Radiology) RAID is an abbreviation that stands for (General) A ramp converter is a type of analogue to digital con- redundant array of independent disks. The RAID technology verter (also called linear ramp converter, ramp-compare, inte- refers to different hardware and/or software methods of data stor- grating, dual-slope, multi-slope or Wilkinson type ADC) used in age which allow increased efficiency, reliability and performance radiation detection systems. Its purpose is to convert the analogue by simultaneous usage of two or more hard disk drives (RAID pulse into a digital signal. It does so by comparing the amplitude array). of an incoming pulse with that of a pulse generated by a linear Each RAID array combines the different hard drives into one ramp. The input pulse is used to charge a capacitor, and discharge logical drive, thus achieving one of the following – mirroring, time, which is proportional to pulse amplitude (proportional to striping, error correction. radiation energy), is measured using a clock oscillator. The num- Most common types of RAID (Figure R.13): ber of the clock pulses counted is proportional to the discharging time, which in turn is proportional to the radiation energy. • RAID 0 – stripping – In RAID 0 the data are distrib- Related Articles: Analogue to digital converter (ADC), uted among two or more disk drives. The capacity of the Wilkinson converter logical drive and the speed are increased. • RAID 1 – mirroring – RAID 1 uses two or more iden- Ramp filter tical disks drives, each one storing the same copy of (Nuclear Medicine) A ramp filter is the basic filter used in the reconstruction technique of filtered back projection. It is used R data. The hard drives in RAID 1 are independent and replaceable in case of failure. The main characteristic with simple back projection to eliminate the 1/r blurring effect. of RAID 1 is enhanced reliability. The shape of the filter in Fourier space is shown in Figure R.14. It • RAID 5 – stripping with parity – A combination of at can be seen from this diagram that the low-frequency components least three disks. Data are distributed between the dif- in the image (e.g. the star artefact) are suppressed whereas higher- ferent disks by mirroring not the disk as a whole, but frequency components are amplified. certain sections. Being a high-pass filter that linearly enhances higher frequen- cies a ramp filter yields the highest resolution possible in a recon- The combination of different types of RAID is known as struction, but also propagates the high frequency noise associated nested RAID: with low count statistics. This propagation of noise often results in images which are difficult to interpret clinically. A second filter or • RAID 01 – stripping and mirroring – RAID 01 com- filtering window is therefore used in combination with the ramp bines two or more RAID 0 arrays (strip) into RAID to modify or roll off the higher frequency noise. 1 array (mirror), thus increasing capacity, speed and Related Article: Image reconstruction reliability. • RAID 10 – mirroring and stripping – RAID 10 com- Ramp time bines two or more RAID 1 arrays (mirror) into RAID (Magnetic Resonance) In MRI, the term ramp time denotes the 0 array (strip), thus increasing capacity, speed and time interval for a magnetic field gradient, frequently measured reliability. in millitesla per meter or mT/m, to build up by driving a current • RAID 50 – combination of stripping and distributed through one or more of the physical gradient coils. parity. • RAID 51 – combination of mirroring and distributed parity. • RAID technology is used in all Hospital Information Systems HIS, PACS, etc. Raid 0 Raid 1 Raid 5 A D A A A B P(AB) B E B B C P(CD) D C F C C P(EF) E F Frequency FIGURE R.13 Most common types of RAID. FIGURE R.14 A ramp filter. Weight factor RANDO phantom 786 Range modulation G(r)/ Bz(r)/T photons in the process. The annihilation photons are emitted back mT/m to back, that is ∼180°. Following event detection in a detector, a number of opposite detectors are scanned in order to detect the t/ms corresponding annihilation photon, that is there is a projection r/m for each detector. Ramp time If both annihilation photons travel through the object without interacting with the surrounding materials before being detected, FIGURE R.15 During the ramp time, the gradient amplitude G(r), mea- the coincidence is referred to as a true coincidence. sured in mT/m, is increased/decreased to the intended value prescribed by A second coincidence type is random coincidences. They occur the pulse sequence code. when photons from two different annihilations are registered in opposite detectors near-simultaneously. The occurrence of random coincidences increases with the square of the administered activ- The ramp time during the build-up phase is also frequently ity, unlike true coincidences that increase linearly. denoted rise time. Ramp times of course also occur when a mag- More information about event types in PET can be found in netic field gradient is turned off (see Figure R.15). the article Event type in PET. The ramp time to maximum gradient amplitude of the physi- Related Articles: PET, Event type PET, True coincidences, cal gradients is, together with the maximum amplitude itself, an Scatter coincidences important characteristic of an MRI system, since this time influ- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. ences, for example minimum achievable echo times in imaging Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, pulse sequences. Philadelphia, PA, pp. 340–342. In a conventional pulse sequence, gradients in the slice, phase and readout (frequency) direction are mandatory and each of Range these gradients is associated with two ramp times (up and down). (General) A charged particle moving through a specific material In modern MRI scanners, minimum ramp times to maxi- loses its kinetic energy through interactions with electrons and mum gradient strength are on the order of a few hundreds of nuclei of the material. The thickness of material that stops a par- microseconds. ticle of given kinetic energy, mass and charge is called range (R) The parameters minimum

ramp time and maximum gradient R of the particle in that material. The range is an average quantity strength are governed by health aspects, since when a magnetic and identical particles with the same kinetic energy moving in the field variation dB/dr is built up or turned off a temporally varying same material and density material will have different pathlength magnetic field dB/dt interacts with the body according to (S). Their range will be equal to or less than the path length. For electrons R<S, while for heavy charged particles R≈S. dB dB L 2 Related Articles: Pathlength, Water equivalent path length = × r / L = SR × r (R.1) dt dr Dt 2 Random noise In the previous equation, dB/dt denotes the magnetic field varia- (Nuclear Medicine) See White noise tion over time, dB/dr is the intended gradient strength in the r direction after ramping up, Lr is the effective length of the gradi- Range compensator ent coil in the r direction and Δt is the ramp time for the gradient (Radiotherapy) In passive scattering, a range-compensator of to reach the value dB/dr. SR denotes slew rate and is defined as the variable thickness (usually made of a low atomic number material ratio between dB/dr and Δt. such as Perspex) is used to shape the proton beam in the depth Related Articles: Gradient, Gradient coils direction. Range compensators are patient and field specific. They compensate for lateral variations in: tissue heterogeneity, RANDO phantom patient surface shape and the distal shape of the tumour. Across (Radiation Protection) RANDO is a popular body phan- a field, they ‘pull back’ the range of incoming protons by vari- tom (anthropomorphic phantom), trade mark of The Phantom able amounts to match that field’s dose-fall off to the tumour’s Laboratory. The phantom is used for assessing dose distribution, distal edge. Alignment errors, internal organ motion and other mainly in radiotherapy and diagnostic radiology. The phantom uncertainties present in proton therapy can be taken into consider- is constructed of tissue-simulating material (with natural human ation in the design of the compensator (Khan and Gibbons, 2014). skeleton) and mimics human body. It is sliced at 2.5 cm sections In modern proton therapy, passive scattering and its associated and each section includes openings for inclusion of dosimeters patient-specific hardware (including the range compensator) has (e.g. TLD tablets). There are two main types of phantoms – largely been superseded by pencil beam scanning. RANDO man and RANDO woman. Related Articles: Pencil beam scattering, Range modulation, Related Article: Anthropomorphic phantom Range shifter Further Reading: http://www .phantomlab .com /rando .html Further Reading: Khan, F. M. and J. P. Gibbons. 2014. Khan’s The Physics of Radiation Therapy, 5th edn., Wolters Kluwer Random coincidence Health. (Nuclear Medicine) A coincidence between two photons originat- ing from different annihilation positions. Range energy relationship In PET imaging the coincidences are localised along a line of (Radiotherapy) See Electron practical range response (LOR). The LOR is drawn between two detectors with a near-simultaneous detection. This is possible because of the Range modulation use of positron emitting nuclides. When a positron is brought to (Radiotherapy) A pristine Bragg peak is too narrow and the high- a near halt, it annihilates together with an electron, emitting two dose region is not sufficient to cover a tumour. Range modulation Range shifter 787 Range straggling is necessary to create a spread-out Bragg peak (SOBP) by com- a range shifter. A range shifter can change the range in steps by bining various pristine Bragg peaks of reduced ranges and energy inserting parallel plates individually in the beam or continuously (Khan and Gibbons, 2014). by using moving wedges (McGowan et al., 2018). In passive scattering, a range modulator wheel is used to In passive beam scattering, a range modulator wheel, rotating spread the beam energy and modulate the range or depth of range shifter wheel or ‘propeller’ is used to spread the beam energy penetration. Two scattering foils (double scattering) are used to and adjust the range or depth of penetration (Khan and Gibbons, spread the beam laterally. Range compensators, usually made of 2014). This ensures that each layer of the target volume is covered. Perspex, and collimators/apertures, usually made of brass, can be As the range shifter wheel rotates, it inserts successively thicker used to shape the beam and achieve a conformal dose distribution layers of plastic into the proton beam. This spreads the beam (Horton and Eaton, 2017). Range compensators shape the beam energy and modulates the range, creating a spread-out Bragg peak in the depth direction. Collimators/apertures shape the beam in (SOBP). Two scattering foils are used to spread the beam laterally. the lateral direction. Field-specific collimators and range compensators are used to help In active scanning, the target is scanned in layers. Each layer is achieve a conformal dose distribution by adjusting the range to positioned at a certain depth and requires a specific proton beam coincide with the distal end of the target volume. energy to reach it. Magnets are used to scan the beam laterally The reader is referred to IPEM Report 75 for schematic dia- and paint the layer. Once the layer has been painted, the depth of grams showing passive scattering and active scanning delivery the Bragg peak can be adjusted by either changing the energy of systems. the proton beam or by using a range shifter. The next layer is then Related Articles: Pencil beam scanning, Range compensator, painted and so on until the whole target has been treated. Active Range modulation scanning does not require field-specific hardware such as colli- Further Readings: Horton, P. and D. Eaton. 2017. Design and mators and range compensators. This reduces neutron production Shielding of Radiotherapy Treatment Facilities, IPEM Report and offers an advantage from a radiation protection perspective. 75, 2nd edn., IOP Publishing; Khan, F. M. and J. P. Gibbons. The reader is referred to IPEM Report 75 for schematic dia- 2014. Khan’s the Physics of Radiation Therapy, 5th edn., Wolters grams showing passive scattering and active scanning delivery Kluwer Health; McGowan, S. E., N. G. Burnet and A. J. Lomax. systems. 2013. Review article: Treatment planning optimisation in proton Related Articles: Bragg peak spreading, Beam modulation, therapy. Br. J. Radiol. 86:20120288. Modulation wheel, Spread-out Bragg peak (SOBP), Pencil beam scanning Range straggling R Further Readings: Horton, P. and D. Eaton. 2017. Design and (Radiotherapy) A singular proton has a finite, energy-dependent Shielding of Radiotherapy Treatment Facilities, IPEM Report range. As the proton slows down, the likelihood of interaction 75, 2nd edn., IOP Publishing; Khan, F. M. and J. P. Gibbons. increases, producing the characteristic Bragg peak. When pro- 2014. Khan’s the Physics of Radiation Therapy, 5th edn., Wolters tons interact with the atoms in a phantom or patient they are scat- Kluwer Health. tered and follow slightly different trajectories. Variable scattering means that proton Bragg peaks do not all coincide at precisely the Range shifter same depth, even for a uniform phantom. The slight variations in (Radiotherapy) In pencil beam scanning (active scanning), the proton trajectory lead to ‘range straggling’, which blurs the Bragg target volume is divided into layers. Each layer is located at a cer- peak. This effect is proportional to the number of scattering tain depth and requires a specific proton beam energy to reach it. events and therefore has a greater impact on higher energy beams. Active scanning uses steering magnets to scan the beam laterally, Figure R.16 Shows (a) a low energy beam with a small amount covering one layer at a time. For each layer, the range or depth of of range straggling and (b) a higher energy beam with more range penetration can be adjusted by either changing the energy of the straggling. proton beam or by dynamically adding material in the beam with Related Article: Pristine Bragg peak FIGURE R.16 (a) A sharp Bragg peak with relatively little range straggling, due to the relatively low proton beam energy: 70 MeV. (b) A Bragg peak exhibiting substantial range straggling due to the higher initial proton beam energy: 240 MeV. Rapid acquisition relaxation enhancement (RARE) 788 Rayleigh distribution Rapid acquisition relaxation enhancement (RARE) Waves (Magnetic Resonance) Rapid acquisition relaxation enhancement Compression (RARE) is a pulse sequence characterised by a 90° pulse followed by a series of rapidly applied 180° rephasing pulses and multiple echoes. See Fast spin echo (FSE) for a more detailed description. Related Articles: Echo train length, Fast spin echo (FSE), Half-acquisition single-shot turbo spin echo (HASTE), Turbo Rarefaction spin echo (TSE) RARE (rare acquisition relaxation enhancement) (Magnetic Resonance) See Rapid acquisition relaxation enhancement Pressure Rare earth metals (Diagnostic Radiology) The rare earth metals are a series of about 17 elements in the periodic table (lanthanides series with atomic number from 58 to 70) that include lanthanum, gadolinium, and ytterbium. The significance in radiology is that compounds of 1 cycle several of these elements are fluorescent materials used in inten- sifying screens. See Rare earth screen for more details. FIGURE R.18 Oscillating movements of the particles result in pressure changes. (Courtesy of EMIT project, www .emerald2 .eu) Rare earth screen (Diagnostic Radiology) Up until the 1970s calcium tungstate those parameters are decreased compared to the steady state, (CaWO4) was the typical fluorescent material in radiographic Figure R.18. intensifying screens. These have been replaced by a variety of Related Article: Longitudinal wave R materials/compounds with a rare earth element (such as lan- thanum, gadolinium, and ytterbium) as the x-ray absorber. A major advantage is a higher x-ray absorption efficiency than Ratemetre tungsten, because the lower atomic numbers (Z), places the K (Nuclear Medicine) Ratemetres give a continuous indication of edge at a more effective energy with respect to the typical x-ray the pulse rate coming from the associated detector. Changes in spectrum. the pulse rate may be monitored by a ratemetre. Ratemetres are Some more typical rare earth phosphors include lanthanum used in survey metres and laboratory contamination monitors. oxybromide and gadolinium oxysulphide. Depending on the acti- There are analogue and digital ratemetres available. vator added to the rare earth screen the colour of the emitted light Further Reading: Orvis, A. L. 1967. Systems for data accu- may change. For example, activators with thulium emit blue light, mulation and presentation, in G. Hine (ed.), Instrumentation in activators with terbium emit green light, etc. yttrium oxysulphide Nuclear Medicine, Academic Press, New York, pp. 119–161. is suitable for blue-sensitive films as it emits blue light. Raw data (Magnetic Resonance) This term is used to describe the array of Rarefaction complex data, generated by the process of data acquisition and (Ultrasound) A sound wave is a mechanical wave in which the sampling, which is the basis of MR image reconstruction. It is a particles in the medium move backwards and forwards in an concrete manifestation of the abstract concept of k-space. oscillating pattern. In liquids and tissue the particles move in MR scanner manufacturers often apply filters and other forms the direction of wave propagation. These are longitudinal waves, of manipulation to raw data, and obtaining true raw data from a Figure R.17. Compression occurs in a region where the particles commercial scanner can be problematic. move towards each other and rarefaction is the opposite where the Related Article: k-space particles move apart. In the region of compression the density of the medium and the pressure increase; in the rarefaction region Rayl (Ultrasound) The Rayl is a unit of acoustic impedance, the ratio of sound pressure to the resulting particle velocity. The unit is Longitudinal waves named after Lord Rayleigh (1842–1919). In SI units, 1 Rayl = 1 N s/m3. Related Article: Acoustic impedance Rayleigh distribution (Ultrasound) The Rayleigh distribution, or more precisely, the Rayleigh probability distribution function, is given by Displacement Direction of wave ( ) B -B2 / 2s = ( 2 ) PDF B e s2 FIGURE R.17 In a longitudinal wave the particles in the medium moves in the same direction as the wave propagates. (Courtesy of EMIT project, where B and σ are parameters. This distribution is encountered www .emerald2 .eu) when the amplitude from a large number of scattering objects Rayleigh scattering 789 RC circuit Im wavelength of the light. In this process the incident photon inter- acts with an absorber atom as a whole and this implies that all orbital electrons

of the absorber atom are tightly bound and con- Btot tribute coherently to the scattering process. The energy of the incident photon is conserved, but its direction of propagation is changed. Related Article: Elastic scattering Bj Rayleigh scattering (Ultrasound) Rayleigh scattering is said to occur when sound (or Re light) is scattered by particles much smaller than a wavelength. ‘Much smaller than’ in practice usually means about one tenth of the wavelength. Typical for Rayleigh scattering is that the inten- sity varies as the fourth power of frequency (up to the point where the scatterer becomes large compared to the wavelength, and it FIGURE R.19 An illustration of a random walk realisation, where a is no longer Rayleigh scattering). The name comes from Lord number of complex phasors Bj with different phase and amplitude add Rayleigh, who showed that the sky appears blue due to sunlight up to the phasor Btot. being scattered in the air, blue representing the shortest wave- lengths, or equivalently, the highest frequencies. Scattering in this regime has important implications in ultrasound imaging. Tissue 4500 is often modelled as subwavelength scatterers, as well as blood cells, used in modelling of Doppler signals. 4000 RBE (relative biological effectiveness) 3500 (Radiotherapy) See Relative biological effectiveness (RBE) 3000 RC circuit 2500 R (General) An RC circuit (resistor–capacitor circuit or RC filter) is 2000 an electric circuit composed of resistor(s) and capacitor(s) driven by a voltage or current source. The simplest example of an RC 1500 circuit is a circuit composed of one resistor and one capacitor in series. The charged capacitor will discharge its energy into 1000 the resistor. This voltage across the capacitor over time could 500 be found through Kirchhoff’s current law, by which the current coming out of the capacitor must equal the current going through 0 the resistor. This results in the linear differential equation (left). 0 0.5 1 1.5 2 2.5 3 When solved, it results in the exponential decay function (right): Amplitude ×10–4 dV V C + = 0 FIGURE R.20 A histogram of occurrences of amplitude values for the dt R received signal from a large number of point sources. The grey line is a fitted Rayleigh distribution. When solved, it results in the exponential decay function: V (t ) = V e-t / RC 0 is studied, see Figure R.19. In a pulse echo system, each scat- terer in a resolution cell will contribute with a signal that can be where described by its phase and amplitude in a complex phasor dia- V0 is the input voltage, and the product of the circuit resistance gram, as in Figure R.20. The sum of all contributions Bj sum up R (in ohms) at the receiver to Btot. If all phases are uniformly distributed over The circuit capacitance C is the time constant τ (in seconds) [−π,π], and the number of scatterers is large, it can be shown that the real and imaginary parts are Gaussian random variables with RC circuits can operate as filters (high-pass or low-pass), inte- the same variances, and zero means. Through a transformation it grators or differentiators (Figure R.21). can further be shown that the phasor magnitude Btot is Rayleigh distributed according to the previous relation, where N 1 s2 2 = lim Bj 2N å j Rayleigh scattering (a) (b) (Radiation Protection) Rayleigh scattering (named after Lord Rayleigh, 1842–1919) is the elastic scattering of light or other FIGURE R.21 (a) RC low-pass filter or ‘integrator’ and (b) RC high- electromagnetic radiation by particles much smaller than the pass filter or ‘differentiator’. Number of occurrences RC time constant 790 Real-time imaging At high frequency, that is when ω » 1/RC, the RC circuit oper- limit the extent to which a particular approach to acceleration can ates as integrator. The capacitor has insufficient time to charge up be applied. and so its voltage is very small. Thus the input voltage approxi- The main approaches to acceleration of data acquisition mately equals the voltage across the resistor. encountered in real-time imaging are as follows: At low frequency, that is when ω « 1/RC, the RC circuit oper- ates as differentiator. The capacitor has time to charge up until its 1. Multiple k-space lines per shot: Acquisition of mul- voltage is almost equal to the source’s voltage. tiple echoes with different phase encoding following a Related Articles: Circuit, Circuit(s), Electrical, Delay circuit single excitation pulse is a well-established technique, exploited in sequences as such as turbo spin echo/fast RC time constant spin echo and a variety of fast gradient echo methods. (General) See RC circuit In the limit (e.g. single shot echo planar imaging [EPI]), it is possible to traverse the whole of k-space following RDSR (Radiation Dose Structured Report) a single excitation, but image quality can be seriously (Radiation Protection) See Radiation Dose Structured Report compromised by poor SNR and marked geometrical (RDSR) distortion. 2. Reduced coverage of k-space: Data from the edges of Reaction time k-space in the phase encoding direction can be omit- (General) Reaction time is the speed at which we are able to pro- ted, reducing the number of k-space lines that need to cess information and make decisions. Reaction time is the time be acquired but also compromising spatial resolution. between the onset of the stimulus and the start of the movement In half-Fourier imaging, data from just under half of in response to it. Mean reaction time for young adults is approxi- k-space are omitted and the missing data points are mately 190 ms to detect a visual stimulus, and approximately 160 reconstructed by exploiting symmetries in k-space. This ms to detect an auditory stimulus. approach compromises SNR. In technology, response time is the time a system or functional 3. Partially parallel imaging: Techniques such as SENSE, unit takes to react to a given input. SMASH and GRAPPA allow reduction in data acqui- Further Reading: Kosinski, R. J. 2008. A literature review on sition, in real space or in k-space, by exploiting the R reaction time, Clemson University. http://biae .clemson .edu /bpc/ redundancies inherent in data acquired using RF coils bp/Lab/110/reaction .h tm composed of multiple elements. Coils with up to 32 elements are now available commercially, and the fea- sibility of larger numbers of elements has been dem- Readout gradient onstrated. As well as coil technology, partially parallel (Magnetic Resonance) ‘Readout gradient’ is an alternative term imaging is limited by SNR and artefact issues. used to refer to the frequency encoding gradient in a conventional 4. Non-Cartesian k-space trajectories: Coverage of MRI pulse sequence. k-space in, for example radial or spiral trajectories Related Articles: Frequency encoding, Readout period can accelerate imaging or reduce the impact on image quality of other approaches that reduce k-space cover- Readout period age. However, these techniques introduce artefacts and (Magnetic Resonance) This term is used to refer to the period of reconstruction is problematic. time when the MRI scanner is acquiring data in the presence of 5. View sharing: In this approach, also known as sliding the readout or frequency encoding gradient. window imaging, data are acquired continuously and Related Articles: Frequency encoding, Readout gradient images are reconstructed more frequently than the whole of k-space is refreshed, so some data are retained Real-time imaging over several sequential images. (Magnetic Resonance) Magnetic resonance imaging (magnetic 6. Temporal redundancy: This is a related class of tech- resonance) is normally regarded as a relatively slow imaging nique to view sharing. In keyhole imaging, the centre modality, because of the need to acquire multiple signals in of k-space is updated in each frame of a dynamic image order to provide adequate k-space data set for image reconstruc- set, but data from the edges of k-space from the first tion. However, techniques have been available for some time to frame are retained throughout the series. This approach accelerate the imaging process, the motivation generally being is suitable in situations where image contrast changes to improve patient acceptability and minimise the likelihood of over time, but structural information remains static (e.g. motion artefacts. Pushing these techniques to their extremes, contrast agent uptake studies). There are more advanced often combining several different approaches, it is possible to col- approaches that update different parts of k-space at dif- lect images at rates of multiple frames per second (fps), allow- ferent intervals of time in an effort to improve handling ing dynamic imaging of physiology (e.g. contrast agent uptake) of high-resolution data. Emerging techniques such as or kinematic processes (e.g. cardiac motion), or the guidance of kt-BLAST take a more sophisticated approach to exploi- interventional procedures. tation of spatiotemporal correlation and redundancy in MRI is always a trade-off between speed (temporal resolu- a way that is mathematically optimised to a particular tion), spatial resolution and signal-to-noise ratio (SNR). The com- application through the use of training data. promises involved in real-time imaging are often detrimental to image quality, introducing a variety of artefacts. The applications Real-time imaging is finding particular applications in cardiac of dynamic imaging often do not require the same image quality MRI, making multiphase imaging over a large volume possible as would be needed for static diagnostic imaging, but neverthe- within a single breath-hold (Figure R.22). less there are application-specific image quality requirements that Related Article: Interventional MRI Real-time portal imaging 791 Receiver FIGURE R.23 Summed real-time portal image of an IMRT beam. FIGURE R.22 Static image/frame from a movie of a volunteer mouth- ing ‘aba-ada-aka’, acquired at 13 fps using radial kt-SENSE. Abbreviations: ICRU = International Commission on Radiation Units, IM = Internal margin, IMRT = Intensity- modulated radiotherapy and ITV = Internal target volume. Real-time portal imaging Related Articles: Portal imaging, Electronic portal imaging, (Radiotherapy) With the availability of digital, electronic por- Electronic portal imaging device tal imaging systems, it is possible to acquire images at rates Further Readings: Partridge, M., P. M. Evans, M. van Herk, of several frames per second. These images may be used to L. S. Ploeger, G. J. Budgell and H. V. James. 2000. Leaf posi- obtain movie loop sequences during the course of radiother- tion verification during dynamic beam delivery: A comparison apy beam delivery. These data provide two types of dynamic of three applications using electronic portal imaging. Med. Phys. R information: 27(7): 1601–1609; Shirato, H., M. Oita, K. Fujita, Y. Watanabe and K. Miyasaka. 2004. Feasibility of synchronization of real- time tumor-tracking radiotherapy and intensity-modulated radio- 1. Patient movement during beam delivery – so-called therapy from viewpoint of excessive dose from fluoroscopy. Int. J. intra-fraction movement Radiat. Oncol. Biol. Phys. 60(1): 335–341. 2. Verification of dynamic treatment such as dynamic intensity-modulated radiotherapy (IMRT) Real-time tomography Intra-Fraction Patient Movement: This is the motion of (Diagnostic Radiology) Real-time tomography is a logical name anatomy whilst the treatment is being delivered. ICRU recom- for very fast CT scanners (with high temporal resolution) neces- mend the use of an internal margin (IM) to define the internal sary for examination of moving anatomical structures. However, target volume (ITV) to account for this. Intra-fraction is particu- ‘Real Time Tomography’ (RTT) is a vendor name of a type of larly important for treatment sites where the effects of breathing high-speed CT scanners, introduced by Rapiscan Systems. are significant, such as the lung or liver. Real-time portal imaging Currently several vendors produce such type high-speed CT scan- provides little soft tissue information and thus implanted fiducial ners. The RTT uses several static x-ray sources (switched sequen- markers are often used to enable determination of the level of tially) with opposite boxes of detectors. intra-fraction motion. High-speed CT scanning is used in industry and medicine Verification of Dynamic Treatment: Examples of this include (specifically for cardiac examinations). With the speed of rota- dynamic IMRT delivery, in which a sequence of images of the tion of classical CT scanners, it is difficult to achieve rotation of changing field shape is acquired and the field shape and intensity the x-ray source and detectors for more than several rotations per in each image is compared with its prescribed values. The distri- second. This requires use of several rotating x-ray sources and bution of the summed delivery is also compared with prescrip- detectors, which effectively act as multiplication of the speed of tion. Figure R.23 illustrates an image of an IMRT beam acquired rotation. Multiple static

sources of radiation (distributed at spe- summing a set of real-time portal imaging frames. cific angles around the scan plane), as well as Swept Electron Beam CT Systems are also used for the purpose of achieving high temporal resolution. Although these CT systems are very fast, at present their high cost makes them suitable mainly for research EXAMPLE purposes. Related Article: Electron beam CT A novel approach to real-time portal imaging has been Further Reading: Lionheart, W. R. B. and W. M. Thompson. developed by Shirato et al. They have developed a method 2017. High-speed X-ray computed tomography, in P. Russo (eds.), of tumour tracking in real time using kV energy portal Handbook of X-ray Imaging: Physics and Technology, CRC Press. imaging with two angled kV x-ray sets and x-ray image intensifiers to image fiducial markers at 25 frames per sec- Receiver ond (see Further Readings). (Ultrasound) The term ‘receiver’ is used to denote the detecting element of an ultrasound transducer. For a pulse-echo system, the Receiver coil 792 R eceiver operating characteristic (ROC) same elements are used for transmit and receive; for continuous Sensitivity (the ability to detect an abnormal finding in the wave (CW) Doppler, separate transducer elements are required. image) – this is the fraction of abnormal findings, that are actu- Related Articles: Pulse echo, CW Doppler ally classified as abnormal. This can be expressed by the formula: Receiver coil Sensitivity = TP/ (TP + FN) (Magnetic Resonance) The receiver coil in MRI detects RF signal from the patient. The origin of the signal is the transverse com- Specificity (the ability to detect a normal finding in the image) – ponent of nuclear spin precession throughout the volume to which this is the fraction of normal findings, that are actually classified the coil is sensitive. The coil detects this signal and feeds it to the as normal. Sometimes specificity is also called selectivity. This system electronics, typically for separation into in-phase and in- can be expressed by the formula: quadrature components, digitisation and ultimately for transcrip- tion to k-space. Specificity = TN/ (TN + FP) Physical realisations of RF coils range from simple single loop coils to multi-element phased arrays. Receiver coils form part of The overall accuracy of the exercise can be expressed by the a tuned circuit designed to resonate at the RF frequency of inter- formula est. The copper strips typically used to construct the coil together with added capacitance form the inductive/capacitive elements in Accuracy = (TP + TN) / (TP + FP + FN + TN) the resonant circuit. In practical application, receive coils are associated with a Plotting the sensitivity as a function of the specificity (for the particular anatomical region of interest, for example head coil, whole exercise) presents a curve useful for assessing image spine coil, spine coil, breast coil, knee coil, endorectal coil, etc. quality. Usually on the y-axis is the sensitivity (also referred to Generally these coils can be connected and disconnected as as true positive fraction, TPF). On the x-axis is (1 − specificity), required by the system operator for a given clinical examination. also referred to as false positive fraction, FPF. The ROC curve In addition an MRI system will have a permanently built in coil received this way is a good quality indicator and can be used to in the system housing called the ‘body coil’ which acts both as compare various imaging methods (or modalities) for the detec- transmit and receive coils. Coils implementing both transmit and tion of some pathological findings. The main areas of the ROC R receive functions are called ‘transceiver’ coils. curve are shown on Figure R.24. A ‘volume coil’ receiver coil is designed to provide good uni- If this assessment is made by medical physicists they can formity of sensitivity to signal throughout the volume imaged, use a test object (phantom) instead of a real medical image with which translates as good image uniformity. Volume coils are pathology. The test object used can be one used for assessment used where a relatively large anatomical volume is imaged, for of image contrast resolution (or noise). During the exercise the example a head coil or body coil. With a ‘surface’ coil, sensitiv- observers mark the visibility of the low contrast inserts of this ity to signal falls off rapidly with distance from the coil. While test object. this reduced field of view gives a poorer uniformity than a vol- Figure R.25 presents three typical ROC curves – line A shows ume coil, SNR values are higher over the anatomy of interest. a case when abnormality can not be distinguished from normality. Surface coil designs may be preferred where the anatomy imaged In this case the SNR of the image is almost zero and the observer is superficial and can be placed close to the receiver coil (e.g. a spine coil). Phased array coil designs combine many surface coil elements together in order to increase potential extent of Optimum anatomy imaged while preserving desirable features of surface coil performance. 1 0.9 Receiver operating characteristic (ROC) (Diagnostic Radiology) The receiver operating characteristic 0.8 (ROC) method is used to analyse data (images). It takes into Better 0.7 0.6 0.5 ss– no disc rim inati on account the skills (or the bias) of the observer (the operator). The method is also related with the SNR of the image. The assessment of the images uses the criterion of finding (seeing) an abnormality in the image (e.g. pathological structure in the lungs), when this abnormality actually exists. The possible results in such an obser- 0.4 gue vation can be as follows: 0.3 dom Ran Worse • Seeing an abnormality when it exists on the image (true 0.2 positive, TP) • Seeing an abnormality, but in fact it does not exist on 0.1 the image (false positive, FP) 0 • Not seeing an abnormality (all looks normal) when it 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 exists on the image (false negative, FN) (1 – specificity) • Not seeing an abnormality and in fact there is no abnor- mality on the image (true negative, TN) FIGURE R.24 Main areas of the ROC. The diagonal marks result from a random guess – true positive is equal to false positive likelihood of a When a number of images are seen by various observers and their binary test. Increasing values show improved performance towards the findings are analysed, one can extract two important measures: optimum with 100% sensitivity and specificity. (Sensitivity) Receptor targeting 793 Reconstruction kernels 1 between positive ions and free electrons or negative ions resulting in neutral atoms (molecules) being formed. The degree of recom- C bination depends on the geometry of the chamber, the polaris- ing voltage and the rate of ionisation produced by the radiation. B Therefore the recombination correction is needed to exposure measurements (dosimetry) made with ionisation chambers. The recombination correction factor ksat used in dosimetric protocols is defined as the inverse of the collection efficiency f: A æ 1 ö ksat = ç ÷ è f ø The recombination correction factor is generally estimated from 0 a two-voltage analysis (TVA) for each beam quality (x-ray and 0 FPF (1 – specificity) 1 high-energy photon beams, electron beams and proton beams). The two-voltage technique proposed by P.R. Almond consists of FIGURE R.25 Different ROC curves (see the text). two measurements of current or charge: 1. I1 (Q1) current (charge) for the normal operating bias is guessing the results. Curve B shows improvement as the abil- voltage V1 ity to detect abnormality (or to see the inserts of the test object) 2. I2 (Q2) current (charge) or the voltage V2 < V1 increases. The best case is when the curve is close to the axes (as curve C). The ion recombination correction factor ksat is then equal to The area under the curve is a measure of detectability. In the worst case (line A) the detectability is 0.5. In the perfect case the é area under the curve (the detectability) will be equal to 1 (the ê( 2 V1 /V2 ) -1ù k ë ûú curve coincides with the axes of the diagram – i.e. above C on sat = é(V V )2 - I ù Figure R.25). ëê 1 / 2 ( 1 /I2 )ûú R Further Readings: Bushberg, J. T., J. A. Seibert, E. M. Leidholdt and J. M. Boone. 2002. The Essential Physics of Recombination correction factors are measured at different dose Medical Imaging, 2nd edn., Lippincott Williams & Wilkins, rates and different polarising voltages during the calibration of Philadelphia, PA; Dowsett, D. J., P. A. Kenny and R. E. Johnston. chambers in diagnostic radiology and in radiotherapy dosimetry. 1998. The Physics of Diagnostic Imaging, Chapmann & Hall Related Articles: Exposure, Ionisation recombination loss, Medical, London, UK. Ionisation chamber Further Readings: Knoll, G. F. 2000. Radiation Detection Receptor targeting and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, (Nuclear Medicine) Receptor targeting refers to use of radiophar- pp. 134–136; Shani, G. 2000. Radiation Dosimetry: Instruments maceuticals which targets specific cell receptors. Tumour cells and Methods, 2nd edn., CRC Press, Boca Raton, FL, pp. 134–136. with a high density of a specific receptor can be targeted by the bio-engineered radiopharmaceuticals. The radiopharmaceuticals Recombination effect will ideally accumulate selectively in malignant cells; hence giv- (Radiation Protection) The charged particles created by ionising ing no unnecessary radiation dose to the surrounding tissue. radiation in gas-filled detectors may collide with neutral atoms One example of a molecular imaging agent is 111I-Octreotide. that are in thermal random motion. There are many types of col- It is an 8-peptide residue that binds to somatostatin subtype-2 lisions, including charge transfer collisions. In the charge trans- peptides. The somatostatin is over-expressed in neuro-endocrine fer collisions a positive ion can take an electron from the neutral tumours, which makes it an attractive choice for both therapy molecule or a free electron can be attached to a neutral atom and imaging. Another example is 99mTc-Depreotide which binds (molecule). During the collision between the positive ion and the to somatostatin subtypes 2, 3 and 5 with high affinity and these free electron or a negative ion a neutral atom (molecule) may be receptors are over-expressed in small cellular lung cancer. formed. This process is called ion recombination. Related Articles: Tracers, Analog tracers, Distribution vol- See Related Articles for further details. ume, Partition coefficient, Tracer flux between compartments, Related Articles: Ion recombination, Ionisation recombina- Tracer kinetic modelling tion loss, Recombination correction, Thimble chamber Recoil electron Recombination factor (General) A recoil electron is the electron that is ejected from an (Radiotherapy) See Recombination correction atom as a result of a Compton interaction with an incident photon. Only some of the energy of the photon is given to the electron, and Reconstruction kernels the photon is scattered. (Diagnostic Radiology) In CT image reconstruction, various Related Articles: Compton effect, Compton scattering image processing algorithms are used to control specific image characteristics, especially detail (sharpness or spatial frequency) Recombination correction and noise. The reconstruction ‘kernel’ is the adjustable fac- (Radiation Protection) Ion recombination in gas-filled radiation tor among the different algorithms that can be selected to give detectors, that is ionisation chambers, occurs during collisions image-specific characteristics. A CT system typically has several TPF (sensitivity) Recovery coefficient (RC) in emission CT 794 Rectification, half-wave phase encoding purposes, and hence a reduction in the overall image acquisition time. As long as acquisition extends as far out in k-space, pixel size and hence spatial resolution will be unaffected. Because less data are collected, use of a rectangular FOV leads to a reduction in signal to noise (SNR) in the image, accord- ing to the square root of the rectangular FOV factor. Thus if the field of view is reduced to 0.75 of its original size along one axis, the resulting SNR will be √0.75 = 87% of its original value. Related Articles: k-space, Phase encoding Rectangular pulse FIGURE R.26 Effect of reconstruction kernel for lung CT: (a) high reso- (Magnetic Resonance) In magnetic resonance imaging, the rect- lution reconstruction kernel, (b) standard reconstruction kernel. angular pulses or hard pulses are achieved by applying a time independent B1 field, resulting in a RECT function–shaped (exci- tation) pulse in the time domain. reconstruction kernels

that can be selected to suit the target organs or diagnostic purposes. Generally, the reconstruction kernel for the abdomen is designated as a standard reconstruction kernel of each CT system. The standard reconstruction kernel has a spatial frequency B characteristic such as smoothing, which suppresses the high 1 spatial frequency noise (fine component of the noise). In con- trast, the high-resolution reconstruction kernel, such as for the bone, has a characteristic for emphasising the high spatial fre- quency, and it provides extremely sharp images. Applying the R standard reconstruction kernel in the lung is not suitable for T diagnosis because higher spatial frequencies are not preserved (Figure R.26b). Rectangular pulse can be designed to be very short and there- Recovery coefficient (RC) in emission CT fore can be used to excite a broad frequency range. If the small (Nuclear Medicine) Due to the partial volume effect, small objects flip angle approximation is fulfilled, the frequency profile of a near the resolution limits of the imaging system appear to have a rectangular pulse is a sinc (i.e. [sin(πx)]/[πx]). lower concentration of radioactivity than is actually present. The The flip angle [θ] of a rectangular pulse is directly propor- recovery coefficient (RC) is defined as the ratio of the apparent tional to the amplitude of the applied RF field [B1] and the pulse concentration to true concentration. The RC for a three-dimen- duration [T]: sional object is the product of the RCs in each dimension. Theoretically, if the size of a given small object and the reso- lution of the emission tomography system are known, a recovery q = gB1T coefficient correction factor can be used to correct for the partial volume effect and calculate the actual radioactivity concentration in the object. where γ is a constant called the gyromagnetic ratio. Related Article: Partial volume effect Pulses are used for the initial magnetisation preparation or for spectroscopy where it is beneficial to excite a wide range of Rectangular FOV frequencies. (Magnetic Resonance) Use of a rectangular field of view (FOV) is Related Articles: B1 field, Gyromagnetic ratio, Magnetic reso- an approach to reducing MRI image acquisition time in situations nance imaging in which the body part being imaged is longer in one direction Further Readings: Bernstein, M. A., K. F. King and X. J. than the other within the desired image plane. Zhou. 2004. Handbook of MRI Pulse Sequences, Elsevier Inc, An example is coronal imaging of the head. In such a situation Amsterdam, the Netherlands; Hacke, E. M., R. W. Brown, M. the FOV need not be square: fewer pixels can be acquired in the R. Thompson and R. Venkatesan. 1999. Magnetic Resonance left–right direction than in the cranio–caudal direction, without Imaging: Physical Principles and Sequence Design, John Wiley the risk of image wrap-around due to aliasing. & Sons, Inc., New York. This can be achieved without loss of spatial resolution by collecting lines along the phase encoding axis in k-space more Rectification sparsely. The interval between lines in k-space is inversely pro- (Diagnostic Radiology) See Rectifier portional to the FOV size: 1 Rectification, full-wave FOV = Dk (Diagnostic Radiology) See Rectifier Reduction of the number of k-space lines equates to reduc- Rectification, half-wave tion in the number of repetitions of the sequence needed for (Diagnostic Radiology) See Rectifier Rectifier 795 R ed, Green, Blue (RGB) Rectifier Red, Green, Blue (RGB) (Diagnostic Radiology) Rectifiers convert alternating current (General) RGB refers to the additive colour model where a wide (AC) to direct current (DC). The conversion process is known as range of colours is constructed from different combinations of the rectification. three primary colours red, green and blue. Any one colour in the A rectifier is a diode (semiconductor or valve) or a combina- RGB model is built up from the three colours; red, green and blue tion of diodes with a specific arrangement that direct the current is referred to as the components of those specific colours. The to pass only in one direction. The output of the rectifier is essen- RGB colour is determined by the individual intensity of compo- tially half-AC current, which is then filtered into DC. nents. When all the components have full intensity, the result- Half-Wave Rectifiers: Half-wave rectification uses either ing colour is white; when all components have zero intensity, the the positive or the negative AC wave. Half-wave rectifiers use colour is black. either one diode (single-phase power supply) or three diodes The RGB colour system is used in devices using emitting light (three-phase power supply). This type of rectification is rel- – i.e. image display on a monitor. Here all colours are formed as atively inefficient, because it blocks half of the input signal the sum of various amounts of the three basic RGB colours. The (Figure R.27). fact that active displays (projecting light, as TV monitors) add Full-Wave Rectifiers: Full-wave rectifiers use both the neg- various proportions of red, green and blue light intensity to form ative and the positive waves of the AC signal, by inverting the a specific colour, led to the naming these colours ‘additive’. polarity of one the waves. In case of single-phase power supply Historically the three primary colours have been accepted full-wave rectifiers use two diodes. One of them conducts during as red, yellow and blue (RYB). This system is still used in art the positive AC wave, while the other one conducts during the (painting). negative AC wave (Figure R.28). Today we know that three primary colours can only create a limited number of colours (a gamut). This number is smaller than Rectilinear scanner the human eye can perceive, but is more than sufficient for repro- (Nuclear Medicine) A scanner with a single collimated PM tube. duction of colour in prints, monitors, etc. The trichromatic theory The scintillation crystal used is sodium iodine (NaI (Tl)). The led to the use of the RGB model for television and John Baird cre- PM tube is scanned from one side of the patient to the other. The ated the first colour TV transmission (1928) using red, green and PM tube is connected, via electric circuitry to a light bulb. The blue emitting phosphors. This established the RGB as the techni- intensity of the light bulb is proportional to the energy deposited. cal ‘primary’ colours. Beneath the light bulb is a very light sensitive film. The light bulb Combinations of two primary colours give the secondary R scans over the film in the same way as the PM tube scans over the colours: magenta (red + blue), cyan (blue + green) and yellow (red patient, thus creating an image that represents tracer distribution + green). in the patient. In contrast to the ‘additive’ colour model with emitting light, The scanner was one of the first equipment used to image in the ‘subtractive’ colour model is with reflecting light. The ‘sub- vivo radionuclide distribution. But the scan is time consuming, tractive’ model is based on pigments which absorb specific RGB one scan takes nearly 30 min and most of the rectilinear scanners colours from the spectrum of the light they reflect. The ‘subtrac- have today been replaced by the scintillation camera invented by tive colours’ of red, green and blue are respectively cyan, magenta H. Anger in 1956. and yellow (CMY). These are sometimes called ‘subtractive pri- Rectilinear scanners are also referred to as radioisotope scan- maries’, or ‘secondaries’, or incorrectly, ‘negatives’ (following the ners or scintigraphs. photographic white-black analogy). Light reflected from maximal Related Article: Radioisotope scanner concentration of CMY pigments will be almost black, as its RGB RL 0 0 FIGURE R.27 An example of half-wave rectifier/rectification. RL 0 0 FIGURE R.28 An example of full-wave rectifier/rectification. Redistribution 796 Reference depth components will be absorbed by the pigments. The CMY sub- Technetium-99m Pharmaceuticals – Preparation and Quality tractive colour model is used for colour printing, but black (K) Control in Nuclear Medicine, Springer, Heidelberg, Germany. is added as a separate pigment (to avoid use of too much colour Related Article: Stannous chloride pigment ink) – the so-called CMYK model. The need for better colour digital printouts (more colour Reference air kerma rate (RAKR) nuances) requires the inclusion of additional subtractive colours (Radiotherapy, Brachytherapy) Calibration of source strength is (e.g. light cyan, light magenta) forming six-colour printing (aka a very important part of a comprehensive brachytherapy quality CMYKLcLm), while the additive colours remained almost system. The instruments, ion-chambers and electrometres, used unchanged. for source strength determinations, should have calibrations that Related Articles: HSL (Hue, Saturation, Luminance), Image are traceable to national and international standards. display Specification of Source Strength for Photon Emitting Further Reading: Tabakov, S. 2013. Introduction to vision, Sources: Source strength for a photon emitting source can be colour models and image compression, Journal Medical Physics given as a quantity describing the radioactivity contained in the International 1:50–55. source or as a quantity describing the output of the source: Redistribution 1. Specification of contained activity (Radiotherapy) Redistribution, also called reassortment, is the a. Mass of radium; mg Ra return towards a more even distribution of cells within the cell b. Contained activity; Ci, Bq cycle following the selective killing of those in the more radiosen- 2. Specification of output sitive phases at the time of irradiation. a. Equivalent mass of radium; mg Ra eq A cell’s susceptibility to radiation varies with the phase of the b. Apparent activity cell cycle: least sensitive to radiation late in the S phase and most c. Reference exposure rate sensitive in the M and G2 phases. Mammalian cell populations d. Reference air kerma rate are asynchronous, that is cells are distributed throughout the e. Air kerma strength cell cycle. The cells in the most sensitive phases will be killed The ICRU Report 38 – Dose and Volume Specification for preferentially by a dose of radiation. Therefore the majority of Reporting Intracavitary Therapy in Gynecology – contains the R the surviving cells will be those in the more resistant phases following statement: resulting in a partial synchronisation of the cell population. If ‘It is recommended that radioactive sources be specified in a fractionated radiotherapy regime is employed, the surviving terms of “reference air kerma rate”. The reference air kerma cells will move through the cycle between fractions and some rate of a source is the kerma rate to air, in air, at a reference of the previously resistant cells will be in a more sensitive phase distance of 1 meter, corrected for air attenuation and scatter- for the next delivery of radiation. This can be considered sensi- ing. For this purpose, the quantity is expressed in μGy*h-1 at tisation resulting from redistribution and may result in a thera- one metre’. peutic gain since sensitisation by this mechanism only occurs in In modern brachytherapy dosimetry, reference air kerma rate rapidly dividing cells and not in late-responding normal tissues. or air kerma strength is the quantity used to calculate absorbed In practice, in human tumours the proportion of cells in the M dose. and G2 phases is normally low, so the effect of redistribution See Source strength for a full description of specification of is expected to be small. Additionally, although redistribution source strength. does not affect late-responding tissues it will occur in acutely Abbreviation: ICRU = International Commission on Radiation responding normal tissues potentially limiting the therapeutic Units and Measurements. gain. Related Articles: Source strength, Mass of radium, Contained Related Articles: Cell cycle, Fractionation, Radiosensitivity, activity, Equivalent mass of radium, Apparent activity, Air kerma Repair, Repopulation, Reoxygenation, 5Rs of radiobiology strength Further Readings: Hall, E. J. and A. J. Giaccia. 2006. Further Readings: ICRU Report 38. 1985. Dose and volume Radiobiology for the Radiologist, 6th edn., Lippincott Williams & specification for reporting intracavitary therapy in Gynecology, Wilkins, Philadelphia, PA; Nias, A. H. W. 1998. An Introduction Washington, DC; ICRU Report 60. 1998. Fundamental quantities to Radiobiology, 2nd edn., John Wiley & Sons Ltd., Chichester, and units for ionizing radiation, Washington, DC. UK. Reference depth Reducing agent (Radiotherapy) The calibration of a radiation beam in exter- (Nuclear Medicine) The 99Tcm-solution obtained from the nal beam radiotherapy consists in establishing the relationship 99Mo/99Tcm generator is in the form of sodium pertechnetate with between the output of the equipment which produces the radia- the oxidation state 7+. This form is rather nonreactive and a

lower tion and the output of the monitor which permits to determine oxidation state is needed to facilitate the labelling procedure. the dose delivered to a patient. Calibration protocols have been Various reducing agents have been used such as stannous chlo- developed for many years to measure the absorbed dose per moni- ride, stannous fluoride or stannous tartrate but stannous chloride tor unit of an accelerator or the absorbed dose rate for a cobalt is the most commonly used reducing agent. Technetium-99m is unit. The formalism applied for the determination of dose using reduced to state 4+ but also states 3+ and 5+ may be formed. the ionisation chamber method is in principle the same in most Further Readings: Kowalsky, R. J. and S. W. Falen. 2004. dosimetry protocols. In addition to the choice of the ionisation Radiopharmaceuticals in Nuclear Pharmacy and Nuclear chamber type, cylindrical or parallel plane, the protocols indicate Medicine, 2nd edn., American Pharmacists Association, the geometry of the measurements and the reference conditions. Washington, DC; Saha, G. B. 2004. Fundamentals of Nuclear The proposed formalisms take into consideration the beam mea- Pharmacy, 5th edn., Springer, New York; Zolle, I. ed. 2007. surements performed at a reference depth which depends on the Reference ionisation chamber 797 Reference isodose type of radiation and its energy. For photon beams the calibration itself should be performed through a measurement at the refer- ence depth which is indicated in most dosimetry protocols as 5 or 10 g/cm2 depending on the beam quality. The dose at the depth of its maximum (dmax) can be calculated by dividing the dose mea- sured at the reference depth by the appropriate percentage depth dose (PDD), tissue phantom ratios (TPR) or tissue maximum ratio (TMR) that are used for the clinical dosimetry. The choice of the protocols to calibrate the dose at a reference depth deeper than dmax is to avoid the influence by electrons scattered in the col- limation system or in any other material in the beam. The dose at dmax is also dependent on the field size at high-energy photons. For electron beams the response of the ionisation chamber per unit absorbed dose in a water phantom varies with the beam qual- ity and the depth of measurement. This variation is determined mainly by the Spencer–Attix water to air stopping-power ratio (L /r)w air FIGURE R.29 Photograph of a Farmer chamber. (Photo courtesy of w æ L ö Kamil Kisielewicz, M. Sklodowska-Curie Memorial Institute, Krakow, ç ÷ Poland.) è r øair at the chosen point of measurement. The beam quality and the reference depth in water must be specified in order to permit an accurate transfer of (L /r)w air . In most protocols the first step is to determine the mean electron energy at the phantom surface E0 E0 R using a given relationship between this energy and the half-value of the depth dose distribution R50. The chamber is then positioned at a specified reference depth dref in water and the value of (L /r)w air at dref is determined from tables which give (L /r)w air as a function of E0 and the depth in water. In most protocols the reference depth is normally chosen as dmax or a specified depth which is deeper than dmax. For incident electron energies above 10 MeV the value of dmax can vary by a large amount from machine to machine for beams of the same R50. It follows that the value of (L /r)w air at dmax will also vary between machines for beams which have the same R50. This leads to the need to express the Spencer–Attix water to FIGURE R.30 Photograph of a Markus chamber. (Photo courtesy of air stopping-power ratio as a function of both R50 and depth. Kamil Kisielewicz, M. Sklodowska-Curie Memorial Institute, Krakow, Related Articles: Percent depth dose (PDD), Tissue phantom Poland.) ratio (TPR), Tissue maximum ratio (TMR), Stopping power ratio Reference ionisation chamber The Markus chamber (Figure R.30) is a parallel plate electron (Radiation Protection) A reference ionisation chamber is ion chamber applied for relative and absolute electron or proton a chamber that is used for the calibration of other chambers dosimetry. The open air chamber, with a plate with a diameter of and it is designed for absolute dosimetry of photon, electron about 30 mm, depth of about 10 mm and nominal volume of 0.2 or proton beams. The specialised laboratory, for example the mm, is used to measure doses in water and solid type phantoms. National Physics Laboratory in the United Kingdom, calibrate Related Articles: Cylindrical ionisation chamber, Ionisation and check dosimeters which are absolute standards. The sec- chamber, Parallel plate ionisation chamber, Thimble chamber ondary standards are the dosimeters used in hospitals and are Further Readings: Graham, D. T. and P. Cloke. 2003. regularly calibrated against the absolute standard. The substan- Principles of Radiological Physics, 4th edn., Elsevier Science dards are the dosimeters calibrated against the secondary stan- Limited, Edinburgh, UK, pp. 331–335; Knoll, G. F. 2000. dard dosimeter. Radiation Detection and Measurement, 3rd edn., John Wiley & In radiotherapy for absolute photon and electron dosimetry, Sons, Inc., New York, pp. 110–144. the Farmer chamber (Figure R.29) is usually used. The Farmer chamber is a cylindrical ionisation chamber with inner electrode Reference isodose made of aluminium and outer electrode of pure graphite. The (Radiotherapy, Brachytherapy) The reference isodose defines the insulation is made of teflon. The nominal volume is about 0.6 cm3. reference volume, see the article Reference volume. The chamber is equipped with covers of different thicknesses to In the Paris dosimetry system, the reference isodose is 85% of cover the outer electrode with the aim of achieving a constant the basal dose. calibration coefficient. Related Article: Reference volume Reference levels 798 Reflection coefficient Reference levels (Radiation Protection) See Diagnostic Reference Levels DRL Reference volume (Radiotherapy, Brachytherapy) Incident Reflected Reporting in Intracavitary Brachytherapy – ICRU Report wave θi θr wave 38: The ICRU Report 38 (1985) defines the reference volume as ‘the volume enclosed by the reference isodose surface. In order pi p Z r to facilitate intercomparisons between radiotherapy centres, it 1 is necessary to agree upon a reference dose level. The treatment dose level defining the treatment volume may be equal to or dif- Z2 ferent from this reference dose level. For reporting intracavitary therapy, it is necessary to determine the dimensions of the refer- pt ence volume’. This means that the reference volume is used to θt compare different dose prescribing systems where different dose Transmitted reference levels are used in centres. wave The ICRU Report 38 recommends a reference dose level of 60 Gy. Reference dose levels on the order of 75–85 Gy are used for the high-risk target volume treated with a combination of external beam radiotherapy and brachytherapy. Reporting in Interstitial Brachytherapy – ICRU Report 58: A reference volume is a volume encompassed by an isodose sur- FIGURE R.31 Reflection and transmission of an acoustic wave. The face defined in relation to the mean central dose. amplitudes are determined by the acoustic impedances of the two media. In the Paris system, the reference dose is 85% of the basal (Courtesy of ImPACT, UK, www .impactscan .org) dose, and the basal dose is identical to the mean central dose in ICRU Report 58 (1997). The reference dose in the Paris system is also the minimum target dose of ICRU Report 58. Reflection back to the transducer plays an important role in R In the Manchester system for interstitial brachytherapy, the the formation of ultrasound images as shown in Figure R.33. minimum target dose is about 90% of the prescribed dose. Reflection from smooth surfaces is described as specular reflec- Abbreviation: ICRU = International Commission on Radiation tion. This is a comparatively strong reflection and is very direc- Units and Measurements. tionally dependent; echoes back to the transducer provide strong Related Articles: Paris system, Dosimetry systems, Reference echoes contributing to the image; echoes in other directions do isodose, ICRU reference point not. Reflections from irregular surfaces provide multi-directional Further Readings: ICRU Report 38. 1985. Dose and volume echoes – those back in the direction of the transducer contribute specification for reporting intracavitary therapy in Gynecology, to the image. In both cases, the onward-transmitted ultrasound Washington, DC; ICRU Report 58. 1997. Dose and volume energy is diminished. specification for reporting interstitial therapy, Washington, DC; The diagram on Figure R.33 depicts an ultrasound pulse meet- Gerbaulet, A., R. Pötter, J.-J. Mazeron and E. van Limbergen, ing a boundary between two tissue types with different acoustic eds. 2002. The GEC ESTRO Handbook of Brachytherapy, avail- impedances. For specular reflection from a smooth surface the echo able at the ESTRO web site: www .estro .be (accessed on 9 July is reflected in one direction. If this is towards the transducer then 2012). the echo is detected and contributes to the image. If it is directed away from the transducer then the echo will not be detected. For Reflection a non-specular reflection, there are echoes in several directions – (Ultrasound) When a propagating ultrasound wave encounters a some are towards the transducer where they contribute to the image. medium with different acoustic impedance (Z) part of it will be Related Articles: Reflection coefficient, Acoustic impedance, reflected from the boundary, Figure R.31. The amplitude of the Speed of sound reflected beam will be higher if there is a large difference between the acoustic impedances. Acoustic impedance Z can be defined as Reflection coefficient Z = (ρk)1/2 where ρ is the tissue density and k the stiffness. (Ultrasound) When a propagating ultrasound wave encounters Consider a material consisting of particles with mass m linked a medium with different acoustic impedance (Z) part of it will together with springs with stiffness k, Figure R.32. If pressure be reflected, Figure R.31. The definition of the pressure reflec- is applied to one end, the particles are displaced. The movement tion coefficient (Rp) is the ratio of the pressure amplitudes of the will be easily transferred if the masses are small and the springs reflected pulse (pr) and the incident pulse (pi). Thus, the reflection are weak. Conversely it is harder to move a material with large coefficient tells us how large echoes we can expect from a bound- masses and stiff springs. As long as all the masses and springs are ary between two media. the same, the wave will be transferred with the same amplitude RP = pr/pi and can be expressed as Rp = (Z2 − Z1)/(Z1 + Z2) and particle velocity (assuming no damping effects). However, at for normal incident and as Rp = (Z2 cos θi − Z1 cos θt)/(Z2 cos the boundary between two materials with different Z (different ρ θi + Z1 cos θt) in the case of oblique incidence. The angles are and/or k) these properties are changed for the transmitted wave related by Snell’s law. The corresponding equations express- and as either pressure or particle velocity can change abruptly this ing the intensity reflection coefficient for plane waves are phenomenon implies that a second wave is formed at the bound- as follows: Ri = It /Ii = R2 p ,Ri = ((Z2 - Z1)/(Z1 + Z 2 2 )) and ary travelling in the opposite direction (pi = pt + pr and vi = vt + R q 2 i = ((Z2cosqi - Z1cosqt )/(Z2cos i + Z1cosqt )) which easily can vr). This is reflection. be derived using the relationship I = p2/2Z (see Intensity). Refocusing 799 Refraction E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E FIGURE R.32 A medium can acoustically be modelled as a matrix of masses connected with springs. (Courtesy of EMIT project, www .emerald2 .eu) z΄ y΄ x΄ (a) (b) Trailing edge 180° flip Leading edge A B y΄ Dispersing Closing ‘fan’ of spins y΄ ‘fan’ of spins Leading edge Trailing edge x΄ x΄ (c) (d) R FIGURE R.34 (a) Longitudinal magnetisation prior to a flip is (b) Incident beam flipped onto the transverse plane on application of an RF pulse (90° in Reflected beam this case) (c) Looking

down on the transverse plane in a rotating frame Transmitted beam of reference, the bulk transverse component is made up of a spread of components covering a range of frequencies, represented here by a ‘fan’ of spins. In a rotating frame the slower spins slip behind in an FIGURE R.33 Specular (L) and non-specular (R) reflection. anti clockwise direction and the fastest spins move ahead in a clock- wise direction. The slowest spin represents the trailing edge of the opening fan and the fastest spin the leading edge. (d) Application of Related Articles: Acoustic impedance, Intensity, Snell’s law, a 180° pulse rotates all spins around the y′-axis. The fan of spins is Transmission coefficient now closing, and is refocused when the leading and trailing edges of the fan meet. Refocusing (Magnetic Resonance) On application of an RF flip to a given volume magnetisation of tissue a detectable RF transverse com- effects fall in this category, and refocusing will eliminate their ponent appears. The transverse component is the vector sum of dephasing effects. Dephasing caused by the intrinsic T2 decay individual transverse component of spins throughout the vol- characteristics of the tissue is not stable in time and will not ume. While all spin components remain in step or ‘in phase’, be rephased. For this reason the peak of an echo in a spin echo the detected signal will be a relatively strong oscillation at the sequence lies on a T2 decay curve. Larmor frequency of the nuclei concerned. In practice, preces- sion frequencies through any given volume will vary due to local Refraction variations in magnetic field strength, susceptibility effects and (Ultrasound) Refraction is the change in direction of a wave due chemical shift. This range of precession frequencies causes spin to its speed. In medical ultrasound this occurs when the sound transverse components to move apart from one another, dephas- wave crosses the boundary of two tissues with different speeds of ing the spins and causing signal to drop off. sound, at an oblique angle (Figure R.35). Snell’s law defines this Refocusing is the process of reversing this dephasing effect, direction change. causing signal to build up again towards a peak through the appli- In ultrasound imaging, refraction can give rise to image arte- cation of an RF refocusing pulse (see Figure R.34). The most facts displaying objects at a displaced position, Figure R.36. effective refocusing is achieved through application of a 180° The ultrasound scanner presumes speed of sound of 1540 m/s. flip, although any flip angle will cause some degree of refocus- Example of soft tissue differing from this speed is the lens in the ing. Refocusing in a spin-echo type generates the spin echo at eye (1620 m/s), fat (1460 m/s), muscle (1600 m/s) and bone (3300 time TE, with the refocusing pulse applied at time TE/2 after the m/s). In some cases, refraction may cause severe loss of energy excitation pulse. deep to the site of refraction (Figure R.37). Refocusing can only ‘correct’ dephasing that is stable in time Related Articles: Snell’s law, Speed of sound, Reflection for any given spin. Magnet non-uniformities and susceptibility coefficient Refresh rate, monitor 800 Registration Effect of refraction on beam direction Speed of sound c1 θi λ1 Speed of sound c1 Snell’s law Speed Speed of sin θ of sound c2 (> c1) t c = 2 θ sound c2 t sin θi c1 λ2 FIGURE R.35 Principles of refraction in sound. The incident beam has a wavelength λ1. When it meets an oblique interface with a medium with FIGURE R.37 Refraction through bone can lead to large changes in higher speed of sound the wave moves faster with a wavelength λ2. The beam direction with corresponding loss of energy deep to the site of angle of transmission is given by Snell’s law as shown. The blue line is refraction. beam direction, red dotted line is the normal to interface at the point of refraction. compared with CRT monitors. In LCD monitors, usually the backlit may create some visual feeling of a flicker. Although this feeling is linked to less eye fatigue, quality monitors would often R have refresh rates of the order of 200 Hz. Some LCD monitors Speed of sound c use dynamic refresh rate – i.e. they could adapt the current refresh 1 rate, according to the frame rate of the displayed images. Related Articles: Digital display, Frames per second, LCD (Liquid crystal display), Cathode ray tube Region of interest (ROI) Speed of sound c (Nuclear Medicine) The region of interest is a user defined region 2 which is commonly abbreviated as ROI. In nuclear medicine ROIs are often placed around particular regions or organs when True Apparent evaluating the organ functionality, radiocompound kinetics, position position activity quantification, etc. When performing quality controls or when validating the camera parameters of a new emission image system the ROI can be used to measure the spatial resolution, uni- formity, etc. An example of a practical implementation of the ROI is when FIGURE R.36 Effect of refraction on mis-registration of objects. The measurement is made of the spatial resolution in a scintillation scanner cannot determine that there has been a change in beam direction camera. The first step is to image a line source. An ROI placed and the object is assumed to be in line with the original transmitted beam over the line source will produce a source profile (which for a direction. number of reasons is smeared out). A measure of the spatial reso- lution is the full width at half-maximum of the source profile. Refresh rate, monitor (Diagnostic Radiology) The refresh rate of a monitor is the rate Registration per second (frequency) at which the images (frames) are updated, (Nuclear Medicine) Image registration refers to the process of measured in times per second – i.e. hertz. accurately aligning one image with another. This is particularly In CRT monitors, this is related normally to the vertical update useful in nuclear medicine and PET as these images suffer from a of images – i.e. the time for which the scanning beam passed from lack of anatomical detail. top to bottom the screen. The afterglow of the phosphor elements Computer algorithms have been developed to register of the CRT monitor reduces the visual flicker which the human SPECT and PET images with CT or MRI scans acquired on eye would detect when images/frames are updated. This flicker separate imaging devices. Markers may be used to align the leads to eye fatigue and has to be minimised. Usually the refresh images or distinct features within the images themselves. These rate of the order of 60–80 Hz does not produce a perception of techniques have been particularly successful for the brain as flicker. It is accepted the primary class CRT monitors should have its shape is constrained by the skull. Outside the brain, organs a refresh rate above 70 Hz. can shift relative to one another depending on the positioning In digital LCD monitors (where there is not a scanning beam, of the patient. Here image registration is a more complicated but a matrix of active pixels), the problem with flicker is smaller, procedure. Regulatory authority 801 Relative biological effectiveness (RBE) Many of the problems encountered in image registration by calculating the function value of pdf*(x) from the can be alleviated by the use of multi-modality imaging systems sampled x value and then check if R2 < pdf*(x). If this such as SPECT/CT and PET/CT. Here the images are acquired relation is fulfilled, then x is accepted as a proper distrib- either simultaneously or consecutively without re-positioning the uted stochastic value. Otherwise, a new x value needs to patient. be sampled, according to the procedure in step 2. Registered images from different modalities are often dis- played as fused images. These show both the anatomy and the Related Articles: Monte Carlo, Distribution function method physiology on the same image. It is also possible to register images from the same modality Relative anisotropy (RA) which have been performed at different times. An example of this (Magnetic Resonance) The relative anisotropy (RA) is a diffusion is to examine effect of treatment on tumour uptake. anisotropy index, determined from the elements of the diffusion Related Article: Image fusion tensor. Commonly, the scaled relative anisotropy (sRA) is calcu- Further Reading: Hill, D. L. G., P. G. Batchelor, M. Holden lated, with the sRA defined as and D. J. Hawkes. 2001. Medical image registration. Phys. Med. Biol. 46: R1–R45. ( 2 2 2 l1 - l) + (l2 - l) + (l3 - l) sRA = 6l2 Regulatory authority (Radiation Protection) The regulatory authority is the national å(l2 authority designated by the government that regulates the intro- i - l) duction and conduct of any practice involving sources of radia- = 6l2 tion. The regulatory authority should be independent of any governments departments and agencies that are responsible for where the promotion and development of the practices being regulated. λi is the ith eigenvalue of the diffusion tensor The regulatory authority must also be independent of registrants, λ– is the mean of the eigenvalues, that is the mean ADC licensees and the designers and constructors of the radiation sources used in practices. The sRA ranges from 0 (isotropic diffusion) to 1 (anisotropic In some countries regulatory responsibility for different diffusion) and is related to RA as sRA × 21/2 = RA. High RA practices or different aspects of radiation safety may be divided values are found in white matter, which in which the diffusion is R between different authorities. highly dependent on the direction in which the diffusion is stud- Further Reading: IAEA. 1996. International Basic Safety ies. Lower RA values are found in grey matter, where the rate of Standards for Protection against Radiation and for the Safety of diffusion is more similar in all directions. Radiation Sources, Safety Series No. 155, International Atomic The relative anisotropy is a rotationally invariant metric, that Energy Agency, Vienna, Austria. is it is independent of the major diffusion direction. Related Article: Diffusion tensor Reject film analysis (Diagnostic Radiology) Reject film analysis is a procedure used in Relative biological effectiveness (RBE) radiography for quality assurance. The procedure is to collect the (Radiation Protection; General); Relative biological effective- films that have been rejected (and usually repeated), review and ness (RBE) accounts for the varying destructive effects of dif- evaluate them to determine reason for rejection, and then sort and ferent radiations on tissue. RBE is expressed relative to x- or tabulate by causes such as patient positioning, technique error, gamma-ray photons, so the RBE of these radiations is unity. Beta processing, etc. The rejects can also be tabulated with respect to radiation also has an RBE = 1. individual radiographers who performed the examination. Studies have shown that, for equal absorbed doses, alpha, pro- The results of the analysis are used to identify the significant ton, neutron and other ionising radiation is more destructive than causes for rejecting films so that specific actions (training, equip- photon or beta radiation. Proton radiation is known to be around ment calibration, revised procedures, etc.) can be taken to reduce 2 times more destructive than photons, alpha radiation is known rejects in the future. to be between 4 and 16 times more destructive than photons, and neutrons are up to 8 times more destructive. Rejection method Effect on tissue is also dependent on other aspects of the (General) Occasionally the distribution function method is cum- radiation, for example energy and radiation quality (spec- bersome to use due to mathematical difficulties in the calculation tral characteristics). For these reasons, and due to incomplete of the inverse of the cumulative probability distribution function knowledge of the risk levels for all kinds of radiations at all (CPDF). One can use the rejection method described by the fol- energies and different qualities, it was decided to adopt single lowing three steps to obtain a random sample: values of 2 for protons, 20 for alpha particles and other heavy fission products, and use an equation modelling a continuous 1. Let the probability distribution function, pdf(x), be curve for neutrons: bounded in the range [a, b]. Calculate a normalised function pdf*(x) = pdf(x)/ max [pdf(x)] so the maximum ì é 2 ï -(ln E ù value of pdf*

is equal to unity. + ê n ) 2.5 18.2 exp ú for En < 1MeV 2. Sample a uniform distributed value of x within the ï ëê 6 ûú range [a, b] from the relation x = a + R1(b−a) and where wR = ï í 2 ï é R1 is a random number. - 2En 3. A second random number R2 with then decide whether 5.0 +17.0 ê (ln ( )) ù ï exp ú ú for En ³ 1MeV ï ê 6 the sampled x should be accepted. This choice is made î ëê ûú Relative biological effectiveness (RBE) 802 R elative biological effectiveness (RBE) The radiation-weighted or equivalent dose is then calculated as H = wR × D for an absorbed dose D. The adoption of over-cau- tious values of wR allows general risk estimates to incorporate radiation factors. More specific radiation-weighting factors are used in radiotherapy to accurately calculate doses. Related Articles: Equivalent dose, Radiation weighting factor Further Reading: Relative Biological Effectiveness (RBE), RBE Quality Factor (Q), and Radiation Weighting Factor (wR), ICRP Publication 92 Ann. ICRP 33(4), 2003. Relative biological effectiveness (RBE) (Radiotherapy) Relative biological effectiveness (RBE) is the ratio of dose at a reference radiation quality and dose of a test radiation that produces an equal biological effect. The reference 100 radiation is usually 250 kV x-rays, but should always be specified. LET (keV/μm) The response of tumours and normal tissues to radiation is affected by the type of radiation. As the linear energy transfer FIGURE R.39 RBE increases with LET, reaching a maximum at a LET (LET) increases, the radiation produces more cell killing per Gy. of around 100 keV/μm beyond which it falls. Radiation with a LET of 100 Figure R.38 shows a typical survival curve for cells exposed to keV/μm is therefore optimal in terms of producing a biological effect. x-rays and fast neutrons. Note that the LET scale is a log scale. The initial shoulder for fast neutrons is smaller than that for x-rays and the final slope is steeper. This indicates that for high- LET radiation either there is a higher ratio of lethal to potentially lethal lesions or that radiation damage is less likely to be repaired 8 correctly. 7 The RBE depends on the LET of the radiation, the radiation 6 R dose, the number of dose fractions, the dose rate and the biologi- cal system or end-point. 5 Figure R.39 illustrates the RBE dependence on LET. The 4 positions of the maxima have been measured for a range of 3 mammalian cells and found to be at LET values in the region of 100 keV/μm. This maximum value of RBE can be understood 2 in terms of the production of double-strand DNA breaks. For 1 low-LET radiations, for example x-rays, there is a low prob- 0 ability that a single track will cause a double-strand break and –20 –10 0 10 20 30 40 therefore the biological effectiveness of the radiation is low. For Temperature (°C) very high-LET radiations, for example those with LET of 200 keV/μm, there is a high probability that a single track will result FIGURE R.40 Saturated vapour pressure of water as function of in a double-strand break but since the ionising events occur very temperature. close together, more energy is deposited than is necessary. This effect is sometimes referred to as overkill. Since RBE is the ratio of dose producing equal biological effect this very high- The optimal LET radiation is that for which the average sepa- LET radiation has a lower RBE than the optimal LET radiation. ration of ionising events is on the order of the DNA diameter. Examples of optimal LET radiations include low-energy protons and alpha-particles. RBE depends on the radiation dose and the number of frac- tions because the shape of the dose–response relationship varies for radiations with substantially different values of LET as illus- trated in Figure R.40. Dose rate can affect RBE because the dose–response curve for low-LET radiations varies with dose rate (see the article on Dose rate dependence) but there is little effect on that for high-LET X-rays radiation. In general, RBE values are high for biological systems that 15 MeV neutrons accumulate and repair a substantial amount of sub-lethal damage and low for those that do not. A more detailed discussion of RBE can be found in the book by Hall and Giaccia. Abbreviations: LET = Linear energy transfer and RBE = Dose Relative biological effectiveness. Related Articles: Cell survival curve, Dose rate dependence, FIGURE R.38 Typical survival curve for cells exposed to x-rays and Dose–response model, Fractions, Fractionation, Linear energy fast neutrons (pale line). transfer, Neutron therapy, Radiation quality, Repair Surviving fraction Vapor pressure (kPa) Relative electron density 803 Relaxation Further Readings: Hall, E. J. and A. J. Giaccia. 2006. smoking cigarettes), including late effects such as cancer, in an Radiobiology for the Radiologist, 6th edn., Lippincott Williams & exposed population. It can also mean the comparison (the ratio) of Wilkins, Philadelphia, PA; Nias, A. H. W. 1998. An Introduction the risk of harm between two populations where one is exposed to to Radiobiology, 2nd edn., John Wiley & Sons Ltd., Chichester, a factor, and the other group is not. UK; Steel, G. G. 2002. Basic Clinical Radiobiology, 3rd edn., Relative risk should always be analysed together with the total Arnold Publishers, London, UK. or absolute risk. Related Articles: Absolute risk, Excess risk Relative electron density (Radiotherapy) Relative electron density (RED) for some mate- Relaxation rial denotes the electron density (i.e. the number of electrons per (Magnetic Resonance) In magnetic resonance imaging (MRI) the volume unit) for that material divided by the electron density for term ‘relaxation’ refers to different processes by which nuclear water. In the following table, relative electron densities for some magnetisation in a non-equilibrium state (typically created by tissues are reported. radiofrequency excitation) returns to the equilibrium distribution. Different physical processes cause different rates of spin relax- ation in different directions with respect to the static magnetic Material Physical Electron Electron Density field B0. Relaxation is typically divided into two types, that is Density (g/ Density per Relative to spin–lattice relaxation (longitudinal relaxation, T1 relaxation) and cm3) cm3 × 1023 Water (RED) spin–spin relaxation (transverse relaxation, T2 relaxation). The rates of spin–lattice and spin–spin relaxation are described by the Water 1.00 3.340 1.000 time constants T Lung (inhale) 0.20 0.634 0.190 1 and T2, respectively, often referred to as relax- ation times. Note that in an ideal environment, where the nucleus Lung (exhale) 0.50 1.632 0.489 being observed is completely isolated, relaxation would not exist. Adipose 0.96 3.170 0.949 In such an idealised environment, magnetisation which is set into Breast (50/50) 0.99 3.261 0.976 a non-equilibrium state cannot equilibrate. Muscle 1.06 3.483 1.043 Spin–Lattice Relaxation (Longitudinal Relaxation, T1 Liver 1.07 3.516 1.052 Relaxation): Spin–lattice relaxation is the process during which Trabecular bone 1.16 3.730 1.117 the spins dispose of the energy that was obtained from the radio- Dense bone (800 mg/cc) 1.61 5.052 1.512 frequency (RF) pulse during excitation. The energy is transferred R Dense bone (1000 mg/cc) 1.66 5.243 1.570 back to the surrounding physical and chemical environment (i.e. Dense bone (1250 mg/cc) 1.83 5.718 1.712 the lattice) in order to restore the equilibrium state. The lattice Dense bone (1500 mg/cc) 2.00 6.209 1.859 consists of neighbouring nuclei or molecules that show vibra- Dense bone (1750 mg/cc) 2.17 6.698 2.005 tional, rotational or translational motion, forming a local mag- Titanium 4.51 12.475 3.735 netic field called the lattice field (BL). The correlation time τc (i.e. the average time between collisions) is often used to describe the mobility of the molecules. An environment with long τc is Related Article: Electron density characterised by slower movements and lower frequencies, while short τc corresponds to higher mobility and higher frequencies. In biological tissue, water is assumed to exist in a number of bind- Relative humidity ing states, from tightly bound water molecules (long τc and low (General) Relative humidity describes the amount of water frequency) at the surface of macromolecules (such as proteins vapour that exists in a gaseous mixture of air and water. Relative and polysaccharides) via a ‘structured’ component (medium τc), humidity (RH) is defined as the ratio of actual vapour density at an intermediate distance from the large molecule, to more or to saturation vapour density at a given temperature. RH there- less free water (short τc and high-frequency components) farthest fore indicates the maximum amount of vapour that air can hold away from the macromolecule. The complex motion patterns lead at a given temperature. The quantitative expression for relative to fluctuations in BL, corresponding to a spectrum of motional humidity as percentage is given by frequencies (described by a characteristic spectral density func- tion, see Figure R.41). Those particular fluctuations in the lat- %Relative humidity tice field that correspond to the resonance frequency ω0 of the æ spins will enable interaction with excited nuclei, causing them to Vapour density ö = ç ÷ ´100 return to the lower-energy state. The energy that is released by è Saturation vapour density ø the nucleus will increase the lattice vibrations and rotations (i.e. transfer into heat). Note that the excess energy is not emitted as The partial pressure of water vapour is proportional to its concen- radiation. Since the spectral density function of tissue is not uni- tration and so %RH can be expressed in terms of vapour pressure form over all frequencies, the spin–lattice relaxation (and conse- which represents the partial vapour pressure contributed by the quently T1) tends to be dependent on the magnetic field strength. water molecules. Therefore %RH is also given by the ratio of the In tissue, an increased magnetic field strength (implying a higher actual partial pressure to the saturated partial pressure of water resonance frequency) leads to prolonged T1 due to fewer compo- vapour at a prescribed temperature (Figure R.40). nents of BL fluctuations at higher frequencies. Formally, the lon- gitudinal relaxation time T1 is the time constant for the recovery Relative risk of the component of the magnetisation vector M that is parallel to (Radiation Protection) Relative risk is used to compare the risk the main magnetic field B0, denoted Mz. The increase of Mz due to from different factors in causing harm (such as medical x-rays and longitudinal relaxation is thus described by Relaxation rate 804 Relaxation time Jω Different tissues show different T2 values (see Relaxation time), but T2 does not display any pronounced dependence on the magnetic field strength within the range of B0 values typically Long τc (bound water) used in MRI. T * 2 and the Static Magnetic Field Inhomogeneity: Inhomogeneities in the static magnetic field of an MRI unit are Medium τ an external source of phase dispersion similar to the effects of c (structured water) spin–spin relaxation. The phase dispersion caused by static field inhomogeneities is, however, not a true relaxation process. The inhomogeneities are static in time and dependent on the location Short τ of the spin in the magnet, that is the signal component lost due to c (free water) magnetic field inhomogeneities can be recovered by performing a spin echo experiment. T * 2 is the time constant for the decay of Mxy when effects of true T2 relaxation as well as static field inhomoge- ω0 ω neities are included: Increasing τc 1 1 = + gDB0 T * FIGURE R.41 Schematic spectral density function Jω for three water 2 T2 components in different binding states (i.e. with different correlation times τc). where γ represents the gyromagnetic ratio ΔB0 is the static magnetic field inhomogeneity (i.e. the local variation in B0) M /T1 z (t ) = M0 (1 - e-t ) The T * 2 relaxation time is always shorter than the T2 relaxation where time. R M0 is the equilibrium magnetisation (along the z-axis) Related Articles: Relaxation time, Relaxation rate t is time Further Reading: Bloembergen, N., E. M. Purcell and R. V. Pound. 1948. Relaxation effects in nuclear magnetic resonance Different tissues show different T1 values (see Relaxation time). absorption. Phys. Rev. 73: 679–712. Spin–Spin Relaxation (Transverse Relaxation, T2 Relaxation): Spin–spin relaxation refers to interaction between Relaxation rate excited nuclei, leading to dispersion of transverse magnetisa- (Magnetic Resonance) In MRI, the relaxation rate

(in s−1) is tion that is out of equilibrium. The spin system does not lose any defined as the inverse of the relaxation time, for example the lon- energy in the spin–spin relaxation process, but the phase coher- gitudinal relaxation rate R1 is given by R1 = 1/T1 and the transverse ence of spin precession is gradually lost. The phase dispersion is relaxation rate R2 is given by R2 = 1/T2. caused by inhomogeneities in the local static magnetic field, and Related Articles: Relaxation, Relaxation time only the inhomogeneities internal to the proton system contribute to true T2 relaxation. Free water molecules with short τc display Relaxation time rapid movements, and one particular magnetic dipole will thus per- (Magnetic Resonance) In MRI, the term ‘relaxation time’ is used ceive high-frequency fluctuations in the local magnetic field, effec- for the time constants associated with changes in the magnetisa- tively averaging out over a few milliseconds. Due to this so-called tion vector M after excitation of the spin system with one or more motional averaging the spins experience a relatively homogeneous radiofrequency (RF) pulses. local field and limited dephasing. Bound water molecules (close to Spin–Lattice Relaxation Time T1: The spin–lattice or longi- macromolecules), on the other hand, display large low-frequency tudinal relaxation time T1 is the time constant for the recovery of motion components (close to zero frequency) and this corresponds the component of the magnetisation vector M that is parallel to to static-field fluctuations, that is a local magnetic field inhomo- the main magnetic field B0, denoted Mz. The increase of M neity. Obviously, such an environment will lead to substantial z due to ge longitudinal relaxation is described by phase dispersion and an efficient spin–spin relaxation. With regard to the spectral density function, T2 relaxation is governed by the motion components close to zero frequency (corresponding to M (t) = M (1 - e-t /T1 z 0 ) fluctuations in the static magnetic field). The transverse relaxation time T2 is the time constant for the decay of the component of the where magnetisation vector M that is perpendicular to the main magnetic M0 is the equilibrium magnetisation (along the z-axis) field B0 after RF excitation, denoted Mxy. The decrease of Mxy due t is time to transverse relaxation is thus described by T1 displays a dependence on the magnetic field strength, and M t) = M e-t /T2 xy( xy(0) × different tissues show different T1 values (see Table R.1). Spin–Spin Relaxation Time T2: The spin–spin or transverse where relaxation time T2 is the time constant for the decay of the com- Mxy(0) is the transverse magnetisation immediately after RF ponent of the magnetisation vector M that is perpendicular to excitation the main magnetic field B0 after RF excitation, denoted Mxy. The t is time decrease of Mxy due to transverse relaxation is described by Relaxivity 805 Relay such as magnetic field strength, temperature and physiological TABLE R.1 environment. Approximate Relaxation Times In Vivo for Human Related Articles: Relaxation, Relaxation rate, Relaxation time Tissues Further Reading: Rohrer, M., H. Bauer, J. Mintorovitch, M. Requardt and H.-J. Weinmann. 2005. Comparison of magnetic T1 (ms) properties of MRI contrast media solutions at different magnetic Tissue 0.5T 1.5T T2 (ms) field strengths. Invest. Radiol. 40: 715–724. Grey matter 800 1000 100 Relaxometry White matter 500 600 90 (Magnetic Resonance) The most common magnetic resonance Muscle 550 900 40 imaging (MRI) techniques for a quantitative diagnosis are Fat 200 250 100 relaxometry (R), magnetisation transfer (MT), diffusion imag- ing (DWI), volumetry and spectroscopy (MRS). Relaxometry Note: Relaxation-time data vary considerable between different literature refers to the study and/or the measurement of the relaxation sources. More detailed information is provided in standard MRI variables in MRI and their dependence on physical param- textbooks (e.g. McRobbie et al. and references therein). eters. One can create a map, based on the relaxation time itself. Generally, applications for the characterisation of tis- sues involve T2 and T2*, making use of spin-echo sequences Mxy(t) = M 0 t /T xy( ) × e- 2 with two or more different echo times (TE) and a long repeti- tion time (TR). T2 and T2* relaxometry maps may be gener- where ated either by spin-echo or by gradient echo sequences, with M the latter being used to measure T2*. Consequently the results xy(0) is the transverse magnetisation immediately after RF excitation may be noisier because the influence of inhomogeneities in the t is time magnetising field is greater. At least two images are needed to generate a map of relaxation rate or relaxation time using spin Different tissues show different T2 values (see Table R.1). echo sequences. The sensitivity of the technique depends on T * the sequence, the repetition time (TR), the echo time (TE), the 2 Relaxation Time and the Static Magnetic Field R Inhomogeneity: T * 2 is the time constant for the decay of Mxy when number of images acquired with different TE and the model effects of true T2 relaxation as well as static field inhomogeneities adopted for fitting the experimental data. are included: Relay 1 1 (General) A relay is a device (switch) that opens or closes a con- = + gDB0 T * 2 T tact when energised under the control of another electrical circuit. 2 When the voltage or current in a relay input exceeds the specified where ‘pickup’ value, the relay contact changes its position and causes γ represents the gyromagnetic ratio an action in the output circuit. In the original form, the switch is ΔB0 is the static magnetic field inhomogeneity (i.e. the local operated by an electromagnet to open or close one or many sets of variation in B0) contacts (relays used in older type of x-ray equipment are shown as illustrated in Figure R.42). A relay is able to control an output The T * 2 relaxation time is always shorter than the T2 relaxation circuit of much higher power than the input circuit. time. The conventional relay type of electromagnetic (induction Related Articles: Relaxation, Relaxation rate type) relays of older equipment models have been replaced by Further Reading: McRobbie, D. W., E. A. Moore, M. J. static relays (solid state relays) that essentially consist of electronic Graves and M. R. Prince. 2003. MRI: From Picture to Proton, Cambridge University Press, Cambridge, UK. Relaxivity (Magnetic Resonance) The relaxivity r describes the ability of a chemical compound (often a contrast agent) to increase the relax- ation rate R of the surrounding proton spins. Relaxation is divided into longitudinal relaxation (T1 relaxation) and transverse relax- ation (T2 relaxation), and the corresponding T1 and T2 relaxivities are denoted r1 and r2, respectively. The relaxivity ri (i = 1, 2) is defined as the change in relaxation rate Ri per unit concentration of contrast agent and is typically expressed in units of m/M/s. The relaxivity is normally defined as the slope of the linear-regression equation obtained from a plot of measured relaxation rate versus the concentration c of the contrast agent: Ri = Ri,0 + ric where Ri,0 is the relaxation rate of the solvent without contrast agent. Relaxivities generally depend on a number of factors, FIGURE R.42 Electromagnetic relays in older type of x-ray equipment. Remote afterloading 806 Reorientation circuitry to include all those characteristics that are achieved by i. Start and control of the treatment (interrupt, moving parts in an electromagnetic relay. For example, in a static emergency stop) relay the operating time can be adjusted by adjusting the value ii. Indicators for the treatment of the resistance in the RC time delay circuit. By analogy with b. Computer with printer the functions of the original electromagnetic device, a solid state i. Treatment control software/source movements relay is made with a thyristor or other solid-state switching device. ii. Documentation of treatment To achieve good electrical isolation and high-speed switching an iii. Connection to trolley optocoupler (a light-emitting diode [LED] coupled with a photo iv. Connection to treatment planning system transistor) can be used. 3. More pieces of equipment Related Articles: Photoelectric relay, Delay relay a. Patient monitoring systems i. Two-way audio system (intercom) Remote afterloading ii. Television system (Radiotherapy, Brachytherapy) See Remote afterloading unit b. Door interlock system c. Emergency container in the treatment room Remote afterloading unit d. Independent radiation level monitor in the treat- (Radiotherapy, Brachytherapy) ment room with separate battery back-up; radiation Source Handling and Loading: The brachytherapy source/s level ‘high’ indicated both by sound and light must be handled and loaded into the applicators for treatment, and e. Radiation survey meter many methods have been used over the time. These methods have f. Treatment planning system been developed primarily to reduce the dose to the personnel but i. Computer – can be the same as the control also to improve the quality of the treatment itself. computer Remote afterloading: ii. Connection to treatment console iii. Printer (plotter) 1. Applicators, needles, catheters, etc. are inserted. iv. Connection to CT-scanner, digitiser, etc. 2. Correct positions are verified using dummy sources. (Figures R.43 and R.44) 3. The source is loaded into the applicator/s using a remote controlled afterloading unit. Relevant personnel operate Related Articles: Brachytherapy, Source loading in brachyther- R the afterloading unit from the operator’s room close to apy, Afterloading, Manual loading, Manual afterloading, Remote the treatment room (compare to ‘linac’ treatments). afterloading A typical installation of a remote afterloading unit for high dose Reorientation rate brachytherapy: (Nuclear Medicine) Initially reconstructed myocardial perfu- sion images are aligned with patient coordinates as transver- 1. In the treatment room: The treatment unit, the trolley, sal images. Due to the orientation of the heart, it is common to with a shielded safe reorient the transaxial images to images perpendicularly to the a. One single source patient-specific long axis of left ventricle (LV). Knowledge of this i. Geometrically small accurate LV orientation is required in order to obtain a correct ii. High specific source strength (e.g. Iridium 192, ‘10 Ci’) b. Stepping source movement, requiring precision drive motor c. Computer controlled source movement of i. Dwell time ii. Step size d. Several treatment channels – computer controlled indexing between channels e. Applicators and source guide tubes, transfer tubes, to connect applicators to the trolley head (applica- tors must be closed – the source must not come into contact with body fluids) f. Built in safety systems i. Dummy source (not radioactive) for verifica- tion of applicator connections (identical to the active source) ii. Detector for radiation level measurements (indication on control panel) iii. Hand crank to rewind the source cable manually iv. I nterlock systems v. Battery back-up, etc. g. Connection to the treatment console and the treat- ment planning system 2. I n the treatment control room: The treatment console FIGURE R.43 Treatment unit (GammaMed Plus, Varian) with 24 treat- and computer ment channels, emergency container, source guide tubes handing on a a. Console rack. Reoxygenation 807 Repair imaging. Hypoxia has been demonstrated as a common feature of human solid tumours that can influence both the malignant pro- gression and the response of tumours to radiotherapy. However, it can result from two quite different mechanisms. Chronic hypoxia results from the limited diffusion distance of oxygen through tis- sue that is respiring. The distance to which oxygen can diffuse is largely limited by the rapid rate at which it is metabolised by respiring tumour cells. In contrast, acute hypoxia is the result of the temporary closing of a tumour blood vessel owing to the mal- formed vasculature of the tumour. There is good evidence that tumour blood vessels open and close in a random fashion and so different regions of the tumour become hypoxic intermittently. At the moment of irradiation, a proportion of cells may be hypoxic but if the radiation is delayed to another point in time, a different group of cells may be hypoxic. The oxygen status of cells in a tumour is dynamic and constantly changing. In the 1960s van Putten and Kallman performed experiments using a transplantable sarcoma in a mouse and demonstrated that during a course of treatment hypoxic cells become oxygenated. FIGURE R.44 Treatment consol and computer, printer, patient moni- They found that the fraction of hypoxic cells in a tumour is about toring systems, emergency instructions, main unit independent radiation the same at the end

of a fractionated regime of radiotherapy as in level monitor. the untreated tumour. If the hypoxic cells were not reoxygenated during the course of treatment, the proportion of hypoxic cells would increase since the radiation depopulates the aerated-cell population more than the hypoxic-cell population. 10 32 Apical 32 Anterior In animal experiments, it has been found that some tumours Anterior take several days to reoxygenate, in others the process is complete Apical within just a few hours, and in some tumours both fast and slow R components of reoxygenation are evident. The type of hypoxia Inferior Inferior being reversed, chronic or acute, is reflected in the differences in time scale. The slow component stems from the reoxygenation of chronically hypoxic cells that occurs as the tumour shrinks: cells FIGURE R.45 Reorientated images. that were beyond the range of oxygen diffusion are closer to a blood supply. The fast component stems from the acutely hypoxic cells reoxygenating as tumour blood vessels open and close: cells 3-D myocardial perfusion distribution. Automatic techniques that were hypoxic at the time of irradiation because they were in for reorientation of the left ventricle are now used routinely. The regions where a blood vessel was temporarily closed reoxygenate reoriented images are usually viewed as slices that coincide with quickly when that vessel is opened. the long and short axes of the LV and called horizontal long-axis, Clearly, reoxygenation of tumours has important implications vertical long-axis and short-axis. With interactively manipulation for radiotherapy. If human tumours do reoxygenate as rapidly and of the slices, comparison between rest and stress studies can be efficiently as most of the animal tumours that have been stud- made (Figure R.45). ied, then the use of a multi-fraction course of radiotherapy may be all that is required to deal effectively with any hypoxic cells. Reoxygenation Unfortunately, knowledge of the time course of reoxygenation in (Radiotherapy) Reoxygenation is the process by which hypoxic human tumours is not known and in fact it is not known with clonogenic cells become better oxygenated during the period after certainty whether any do actually reoxygenate. However, support- irradiation. ing evidence exists from the standard clinical use of fractionated Over the years, many chemical and pharmacological agents treatments where many tumours are eradicated with doses on the have been discovered that modify the response of mammalian order of 60 Gy given in 30 fractions which would be unlikely in cells to radiation. The simplest of these, and the one that possi- the presence of a very small proportion of hypoxic cells. bly produces the most dramatic effect, is oxygen. It is known that Related Articles: Alpha beta ratio, Fractionation, Interruption tumour cells can be hypoxic and therefore less sensitive to radia- of treatment, Radiosensitivity, Repair, Redistribution, tion that indirectly damage DNA through free radicals produced Repopulation, 5Rs of radiobiology by the ionisation of oxygen. Therefore, the presence or absence of Further Reading: van Putten, L. M. and R. F. Kallman. 1968. oxygen dramatically influences the biological effect of sparsely Oxygenation status of a transplantable tumor during fractionated ionising radiations such as x-rays but there is no effect for densely radiotherapy. J. Natl. Cancer Inst. 40: 441–451. ionising radiations such as α-particles. The oxygenation status of human tumours has been deter- Repair mined with a variety of techniques including measuring the dis- (Radiotherapy) Repair refers to the process by which the function tance between tumour cells and vessels in histological sections, of macromolecules is restored following radiation damage. As a determining the oxygen saturation of haemoglobin, monitoring result there is an increase in cell survival or a reduction in the changes in tumour metabolism through to the newer techniques of extent of radiation damage to a tissue when time is allowed for oxygen probes, hypoxic markers, the comet assay and non-invasive repair to occur. Septal Lateral Septal Lateral Repair of radiation damage 808 Repetition time (TR) DNA is considered the principal target for the biological 1 effects of radiation, including cell killing, carcinogenesis and mutation. DNA is a large molecule with a double helix structure Split dose consisting of two strands. If cells are irradiated with a modest dose of x-rays, many breaks of a single strand (SSB) may occur. Such lesions are of little biological consequence for cell killing since they are repaired easily using the opposite strand as a tem- plate. Likewise, if both strands are broken but the breaks are far apart, repair again can occur readily since the two breaks are Single dose handled separately. However, if the break in the two strands are opposite one another (or within a few base pairs), a double strand break (DSB) may result. DSBs are regarded as the most important 0.1 D/2 D radiation-induced lesions. Dose Radiation damage to mammalian cells can be categorised as FIGURE R.46 If the dose is delivered in two fractions separated by a 1. Lethal damage (LD): Irreversible and irreparable, lead- time interval, there is an increase in cell survival because the shoulder of ing irrevocably to cell death. the curve must be expressed with each fraction. 2. Potential lethal damage (PLD): Radiation damage that under normal circumstances causes cell death but can be modified by post-irradiation environmental treatment since tumours generally have high values of the α/β- conditions. ratio while late responding normal tissues have low values of the 3. Sublethal damage (SLD): Damage which under nor- α/β-ratio. Therefore a fractionated regime may enhance the thera- mal circumstances can be repaired within hours unless peutic effect. additional sublethal damage is added, for example a Abbreviations: DNA = Deoxyribonucleic acid, DSB = Double second dose of radiation, with which it can interact to strand break, LD = Lethal damage, NLD = Non-lethal damage, form lethal damage. PLD = Potential lethal damage, SLD = Sublethal damage and R 4. Nonlethal damage (NLD): It results in cells with heri- SSB = Single strand break. table lesions (sometimes called lethal mutations) which Related Articles: Alpha beta ratio, Fractionation, Linear do not prevent proliferation but may affect the rate of quadratic (LQ) model, Linear-quadratic dose–response curve, proliferation. Radiosensitivity, Repopulation, Redistribution, Reoxygenation, 5Rs of radiobiology, Surviving fraction, Therapeutic effect. Repair is possible for PLD and SLD. Further Readings: Hall, E. J. and A. J. Giaccia. 2006. PLD is repaired if post-irradiation conditions are sub-optimal Radiobiology for the Radiologist, 6th edn., Lippincott Williams & for growth. In these circumstances, mitosis is delayed so there is Wilkins, Philadelphia, PA; Nias, A. H. W. 1998. An Introduction the potential for any damage to the chromosomes to be repaired to Radiobiology, 2nd edn., John Wiley & Sons Ltd., Chichester, before it is attempted. Repair of PLD has been observed in trans- UK. plantable animal tumours so it is reasonable to suppose that it also occurs in human tumours. Indeed, it has been suggested that Repair of radiation damage radioresistant human tumours may have more efficient mecha- (Radiation Protection) When cells are exposed to ionising radia- nisms to repair PLD than radiosensitive tumours. This hypothesis tion, damage can occur, both in the DNA itself, and in other bio- has yet to be proven. logical molecules within the nucleus and cytoplasm. This damage The effect of a given dose of radiation is less if it is split into does not necessarily lead to adverse effects, or bioeffects. Cells two fractions delivered a few hours apart. This is termed repair have proteins and enzymes whose main function is to act as part of SLD. Considerable repair occurs within 15–60 min with com- of the mechanisms to repair damage to DNA. Ordinarily these plete recovery usually by around 4–6 h, although repair seems mechanisms work very well to repair radiation damage. to be slower in some normal tissues such as the spinal cord. The For more information, see the articles on Bioeffects, Radiation effect of the repair of SLD can be seen in the cell survival curve damage and Radiobiological models. of a cell population receiving a dose split into two equal doses Related Articles: Bioeffects, Adverse effects, Radiation dam- separated by a time interval compared with that of a cell popula- age, Radiobiological models tion receiving the whole dose in a single fraction, Figure R.46. More cells survive in the population that receive the fraction- Repetition time (TR) ated treatment than for the single fraction population because (Magnetic Resonance) Repetition time (TR) is the time between the shoulder of the curve must be repeated with each fraction. successive pulse sequence executions, applied to the same slice or In general, there is a good correlation between the extent of volume of interest. In standard MR imaging, a pulse sequence is sublethal damage repair and the size of the shoulder of the cell repeated in order to achieve sufficient k-space coverage and the survival curve. phase encoding gradient is changed one (or several) times dur- In terms of the linear quadratic model description of the sur- ing each execution. In single-shot sequences (acquiring the whole vival curve, tissues with small values of the α/β ratio have high k-space information during one repetition), the term ‘TR’ is still sublethal damage repair potential while repair potential is lim- used if several images of the same volume are acquired in a time ited for tissues with large values of α/β. Hence, tissues with low series, for example in fMRI applications. α/β-ratios are sensitive to the dose fractionation scheme (fraction- TR, echo time (TE) and excitation pulse flip angle are impor- ation effect) while those with high α/β-ratios are hardly affected tant parameters in the build-up of image contrast. As an example, by dose fractionation. This has implications for radiotherapy a very long TR (approximately five T1 relaxation times) allows full Surviving fraction Repopulation 809 Repopulation TR of spinal cord in rats show that the effects of repopulation occur TE much later in a fractionated regime (40 days compared with the 90° 180° 12 days observed for skin in mice) with a slower increase in the 90° extra dose required to compensate for proliferation with time. Comparable data for humans are not available and the times- RF cales are expected to be very much longer than those observed in rodents. However, if the experimental data on rodents were to be GS translated to clinical radiotherapy for humans, it would be reason- able to assume that for conventional radiotherapy protocols, early- responding tissues are triggered to proliferate within a few weeks GP of the start of fractionated treatment but overall treatment time (typically 5–7 weeks) is not long enough to allow the triggering of GF proliferation in late-responding tissues. Therefore early reactions, such as reactions of the skin or mucosa, can be easily dealt with by simply prolonging the overall treatment time. However, such ADC a strategy has no effect on the late reactions and, as we will now discuss, may actually have a detrimental effect on the therapeutic FIGURE R.47 Example of a spin-echo pulse sequence. RF and ADC efficacy of the treatment if proliferation of tumour cells becomes are the radio frequency pulse and the signal received from the slice. GS, significant. GP and GF are the slice selective, phase and frequency encoding gradi- Irradiation of tumour cells can trigger the surviving cells to ents. TR is in this basic sequence type the time between two successive divide faster than before. This is known as accelerated repopula- 90 pulses. tion. This phenomenon has been observed in animal studies and there is evidence for it in some human tumours. Withers et al. (1988) performed a review of tumour control data for head and build-up of the longitudinal magnetisation to the order of its ther- neck cancer and demonstrated a ‘dog-leg’ variation in total dose mal equilibrium value after excitation, while a shorter TR does for a given tumour control response with treatment time of the not. In the latter case, different T1 values in different tissue types form illustrated in Figure R.48. If the treatment is completed give different build-up of longitudinal magnetisation during TR, before the initiation of accelerated repopulation (denoted by the R and when the magnetisation is repeatedly flipped into the trans- discontinuity at time Tdelay), the dose required is fixed and inde- verse plane, the detected signal will contain T1 contrast informa- pendent of treatment time. However, if the treatment duration is tion (Figure R.47).

greater than Tdelay, the dose required to produce the same tumour Related Articles: Echo time (TE), Encoding gradients, Flip response increases and is time-dependent. This effect is attributed angle, Relaxation time to tumour repopulation and should be considered in biological Further Readings: Brown, M. A. and R. C. Semelka. 2003. effective dose (BED) calculations for schedules where the overall MRI: Basic Principles and Applications, 3rd edn., Wiley-Liss, treatment time is extended compared with that prescribed, such Hoboken, NJ; Haacke, E. M., R. W. Brown, M. R. Thomson and as may occur when there is an interruption in treatment. Recent R. Venkatesan. 1999. Magnetic Resonance Imaging. Physical data have identified that prolongation of overall treatment time Principles and Sequence Design, Wiley-Liss, New York. detrimentally affects local control rates for tumour types includ- ing squamous cell carcinomas (SCC) of the head and neck region, Repopulation cervix, lung, oesophagus, skin, and possibly vagina. It is also (Radiotherapy) Repopulation refers to the process following irra- thought to affect medulloblastoma and carcinoma of the bladder. diation whereby the number of cells in a normal tissue or tumour is restored by the proliferation of surviving cells. For fractionated radiotherapy treatment, ideally the normal tissue would be completely repopulated following irradiation while the tumour would show no growth between fractions. If this were the case, the tumour would be progressively depopulated while the surrounding normal tissue would maintain a steady state. There is in general no difference in cell cycle times between tumour cells and those of normal tissues but their effective dou- bling times may show considerable differences due to the effects of cell loss and other factors influencing cell population kinetics. In some normal tissues there may be a large increase in the proliferation rate following the initial radiation injury until the normal tissue is fully reconstituted. This is seen, for example in haemopoietic stem cells which show an immediate response with both an increase in the growth fraction and a decrease in the maturation rate. Studies of skin damage in mice show that the effects of repopulation are seen at around 12 days into a fraction- Tdelay Treatment time ated regime after which the extra dose required to compensate for proliferation increases very rapidly with time. The behaviour FIGURE R.48 Schematic of effect of overall treatment time on the total of such early-responding normal tissues is considerably different dose required for a given tumour control probability (TCP). The dose from that of late-responding tissues such as spinal cord. Studies rises steadily after an initial delay period, Tdelay. Total dose required for given TCP Resistance, electrical 810 Resistance index Rapid proliferation does not occur in carcinoma of the breast and where prostate and hence overall treatment time is not so critical for ℓ is the length (in m) of the object these tumour types. A is the cross sectional area (in m2) of the object A number of clinical trials have been performed using alter- ρ is the resistivity (in Ω·m) of the material native fractionation schedules to the standard (one fraction per day of around 2 Gy given 5 days a week for up to 7 weeks) to The resistance of a resistive object determines the amount of try to reduce the effect of tumour repopulation on local con- current through the object for a given potential difference across trol. Continuous hyperfractionated accelerated radiation therapy the object, in accordance with Ohm’s law: (CHART), which reduces overall treatment from 6–7 weeks to 12 V days and gives 36 small fractions, has been tested in multicentre I = randomised controlled clinical trials. The trial in non-small-cell R lung cancer showed improvement in survival and this regime is where now the government recommended standard of care for eligible R is the resistance of the object (measured in ohms) patients in the United Kingdom. In the head and neck CHART V is the potential difference across the object (measured in trial, there was only a small, non-significant improvement in the volts) disease-free interval. However, the DAHANCA Trial 6 and 7 I is the current through the object (measured in amperes) for SCC of the head and neck, initiated to examine whether a reduction of the overall treatment time by increasing the number The electrical resistivity of a metallic conductor decreases of weekly radiotherapy fractions from 5 to 6 (while maintain- gradually as the temperature is lowered. Near room tempera- ing same total dose and fraction number) improved the tumour ture, electric resistance of a typical metal increases linearly with response (and were acceptable with regard to early and late mor- increase of temperature. Even near absolute zero copper shows bidity), showed an improvement in 5 year loco-regional control a non-zero resistance. In comparison, the electric resistance of a for the 6 fraction per week arm compared with 5 fraction per typical intrinsic (non-doped) semiconductor decreases exponen- week arm. This accelerated schedule has become the standard tially with increase of temperature. baseline treatment in Denmark. In certain materials a phenomenon of zero electrical resistance Abbreviations: BED = Biological effective dose, CHART occurs, called superconductivity, found to be present generally at = Continuous hyperfractionated accelerated radiation therapy, R very low temperatures. The resistance of a superconductor drops DAHANCA = Danish head and neck cancer and SCC = Squamous abruptly to zero when the material is cooled below its critical cell carcinoma. temperature. The superconductors used in superconducting coils Related Articles: Alpha beta ratio, Biological effective dose, of high magnetic field MRI scanners become superconducting Cell cycle, Cell proliferation, Fractionation, Interruption of treat- only at temperatures below 10 K and special cooling with liquid ment, Radiosensitivity, Repair, Redistribution, Reoxygenation, helium is needed for this purpose. 5Rs of radiobiology, Therapeutic efficacy, Tumour control Related Articles: Insulation resistance, Magnet(s), Super- probability conducting, Superconducting magnet, Superconducting material Further Readings: Dale, R. G., J. H. Hendry, B. Jones et al. 2002. Practical methods for compensating for missed treat- Resistance index ment days in radiotherapy, with particular reference to head (Ultrasound; Clinical; General) The resistance or resistive index and neck schedules. Clin. Oncol. 14: 382–393; Hall, E. J. and is a simple, commonly used measure of arterial waveform shape. A. J. Giaccia. 2006. Radiobiology for the Radiologist, 6th edn., It was designed to reflect distal resistance in the cerebral circula- Lippincott Williams & Wilkins, Philadelphia, PA; Nias, A. H. tion (1). The resistive index (RI) is non-dimensional and indepen- W. 1998. An Introduction to Radiobiology, 2nd edn., John Wiley dent of the ultrasound beam/blood flow direction angle. The RI is & Sons Ltd., Chichester, UK; Overgaard, J., H. S. Hansen, L. calculated from peak systolic and end diastolic velocities in a flow Specht et al. 2003. Five compared with six fractions per week waveform as shown in Figure R.49 and is calculated by of conventional radiotherapy of squamous-cell carcinoma of head and neck: DAHANCA 6 and 7 randomised controlled trial. Lancet 362: 933–940; Saunders, M., S. Dische, A. Barrett et al. æ V RI = max - Vmin ö ç ÷ 1999. Continuous, hyperfractionated, accelerated radiotherapy è Vmin ø (CHART) versus conventional radiotherapy in non-small cell lung cancer: Mature data from the randomised multicentre trial. Radiother. Oncol. 52: 137–148; Withers, H. R., J. M. G. Taylor and B. Maciejewski. 1988. The hazard of accelerated tumour clo- nogen repopulation during radiotherapy. Acta Oncol. 27: 131–146. Vmax PSV Resistance, electrical Peak to (General) Electrical resistance is a ratio of the potential applied peak velocity to a given conductor to the current intensity value. The SI unit of EDV electrical resistance is the ohm, symbol Ω. Assuming a uniform current density, electrical resistance of a Vmin resistive object is a function of both its physical geometry and the resistivity of the material it is made from: × r R = FIGURE R.49 The variables used in the resistance index. The white A broken line represents the time-averaged maximum velocity. Resistive magnet 811 Resolution Abbreviations: EDV = End diastolic velocity, PSV = Peak systolic Resolution velocity, Vmax = Maximum velocity and Vmin = Minimum velocity. (Magnetic Resonance) Spatial resolution is a measure of the abil- Related Articles: Pulsatility index, Peak systolic velocity, End ity to distinguish closely spaced objects in an image. diastolic velocity Ideally the spatial resolution of MR images should be limited Further Reading: Planiol, T. and L. Pourcelot. 1973. Doppler by the size of the voxels in the image matrix. For example, if a effect study of the carotid circulation in Second World Congress 240 mm field of view is imaged using a 256 × 256 acquisition on Ultrasonics in Medicine, Rotterdam, the Netherlands, matrix, then the spatial resolution should be 0.9375 mm in both pp. 104–111. the frequency and phase encoding directions. In practice, whilst frequency encoding is performed over a Resistive magnet single step, phase encoding is performed over a series of repeti- (Magnetic Resonance) In an electromagnet with resistive conduc- tion times (256 steps for a 256 × 256 matrix), each with a differ- tors power is lost as coulomb heat and the achievable magnetic ent gradient amplitude. As a result, the phase encoding resolution field strength is limited. Basic designs for open MRI systems will be more affected by any inconsistencies in gradient switching (Figure R.50) are the ‘double doughnut’ for horizontal B0 field and rate, gradient linearity and/or eddy current compensation. For this the ‘C-shape’ for vertical B0 field. In the latter, the field strength is reason the resolution in the phase encoding direction is generally enhanced by an iron-core. The maximum static magnetic field is of lesser quality than the resolution in the frequency encoding 0.6 T for the iron-cored design and 0.2 T for the air-cored design. direction. Related Articles: Electromagnet, Fringe-field, Magnet, Measuring Spatial Resolution: Spatial resolution may Permanent magnet, Superconductive magnets be assessed visually using images of bar patterns/line pairs. Figure R.51 shows an image of the specially designed EuroSpin Resistor Test Object 4. (General) See Resistance, electrical Here the size of the smallest bar set for which resolution is achieved is taken as the resolution. Further Reading: Lerski, R. et al. 1998. Quality control Resolution in magnetic resonance imaging, IPEM Report 80, Institute of (Diagnostic Radiology) Resolution describes the ability to Physics and Engineering in Medicine, York, UK. ‘resolve’ or see the separation between two objects. The ability of an imaging system to resolve two relatively small R and close objects is spatial resolution. It is reduced by image blur- Resolution ring. Spatial resolution test objects consisting of adjacent lines (Ultrasound) In imaging, resolution describes the ability of a separated by spaces (line pairs) are used to evaluate the effect system to separate objects. In an ultrasound scanner resolution is of blurring in imaging procedures. Often the term ‘resolution’ is usually described and measured by used as synonym of spatial resolution (measured in line pairs per millimetre, Lp/mm). • Spatial resolution, the ability to separate similar objects The ability of an imaging system to resolve small differ- in the axial, lateral and slice thickness planes ences in contrast (low contrast) of objects is contrast resolution. • Contrast resolution, the ability of the display to separate It is reduced by image noise. contrast resolution test objects are the acoustic properties from different tissues mainly related to contrast detail measurements. • Temporal resolution (link), how quickly images are In digital imaging an image is represented by three main com- updated ponents – matrix height, width and depth (bits, greyscale). The first two are related to the spatial resolution; the depth is related to The resolution of ultrasound scanners is dependent on their design contrast resolution (noise dependent). and operation. For example, poor control of beam width leads to a Related Articles: Spatial resolution, Contrast resolution, Line reduced lateral resolution. Scanning deep structures may lead to a pairs, Modulation transfer function, SNR reduction in temporal resolution. The performance of ultrasound scanners is limited by the pulse-echo time required (1/PRF) and the number of pulses required to form images. The best use of this I I I B 1.0 mm Bars 0 MTF blocks 0.3 mm Bars B I 0 0.5 mm Bars 2.0 mm Bars (a) (b) FIGURE R.50 To the left, an air core double doughnut magnet with a 1.5 mm Bars relatively strong fringe field. To the right, an iron core C-shaped magnet, with a relatively

low fringe field, since magnetic field lines are mostly confined to the iron core. Resistive magnets need a power supply and means of dissipating the Coulomb heat in order to maintain a constant temperature and to reduce temperature-related changes in resistance. FIGURE R.51 Image of EuroSpin Resolution Test Object (TO4). Resonant frequency 812 Respiratory gating time is one of the main considerations in scanner design; users are 2.68 ´108 /s/T often faced with choices to optimise temporal resolution or spatial n = ´1.5T resolution of an image. 2p n = 63.98 MHz Resonant frequency (Magnetic Resonance) Classically, under the influence of a static In practice if an MR scanner is used off-resonance, it will produce external magnetic field B0, a magnetic dipole with an orthogonal images with reduced SNR; thus, the resonance should be checked component γh¯ precesses with angular frequency ω0: prior to clinical acquisition by QA measurements. Related Articles: RF pulse, Larmor frequency w0 = gB0 Respiratory gating where γ is the gyromagnetic ratio for a specific nucleus, Hence, (Radiotherapy) Respiratory gating is important in two areas: the resonant frequency ω0 is commonly known as the Larmor fre- three-dimensional imaging of mobile anatomy and treatment of quency (see Larmor frequency article). mobile anatomy using external beams. Quantum mechanically, the energy difference between the Image Acquisition: Tomographic imaging of mobile organs two levels of a spin-1/2 nucleus in an external magnetic field may be subject to geometric distortion due to interplay between (Zeeman splitting, Figure R.52) may be associated with a reso- the scanning time and the breathing or cardiac cycle. A conse- nant frequency according to: E = hν This article will focus upon quence of this is that a spherical object may appear distorted or resonance effects from a quantum mechanical point of view; for a even as two or more objects in a multislice CT scan (Reitzel et classical approach see the RF pulse article. al.). Respiratory gating is used to ameliorate these effects. Two The energy difference between the two levels on Figure R.52 approaches have been used: has an associated frequency g 1. To acquire the scan while the patient is holding their n = B0 2p breath. This is a simple solution but has the disadvantage which is the resonant frequency. for radiotherapy treatment that the technique cannot R Resonant Frequency Effects: Transitions between the two be used during treatment (owing to the long beam-on spin states shown in Figure R.52 are electric dipole forbidden but time) and thus is not likely to be representative of the can be induced by magnetic fields. This can be achieved by apply- treatment-time position. ing a strong static field, but is more easily achieved using a field 2. To acquire the raw scan data with a measurement of oscillating at the resonant frequency of the system, allowing a position in the breathing or cardiac phase. much lower strength field to be used. Quantum mechanically, a spin with N valid energy states can also exist in any linear superposition of these N states. In this case the data may be used to form an image of a particu- Oscillating magnetic fields at resonance with the spin drive it lar phase, or to create a time sequence of images throughout the back and forth between the energy levels, passing through all breathing or cardiac process, in so-called four-dimensional CT the superposition states along the way. RF pulses are described (4DCT). In the case of the cardiac cycle, an electrocardiograph by quantum statistics of a spin ensemble. Thus, by choosing the (ECG) is generally used. To measure the breathing cycle, markers duration of the RF pulse it is possible to generate any desired on the patient surface, spirometry, or measurement of the tem- superposition. perature of air at the patient’s nostrils have been used. In MRI the classically termed 180° pulse corresponds quan- Radiotherapy Treatment: In radiotherapy of treatment sites tum mechanically to a direct interconversion between the two with organs that are mobile due to respiration (including lung and states shown in Figure R.52. The classical 90° pulse causes an liver), it is desirable to control the effects of this motion. This has equally weighted superposition of the two spin states in the quan- the benefit of reducing the risk of geometric miss and allowing tum mechanical picture. margin reduction, with consequent sparing of normal tissues. Two Quantifying the Resonant Frequency: From Figure R.52, the basic approaches to gating are used: frequency of the transition between the energy levels of a spin-1/2 nucleus is given by ν = (γ/2π)B0. 1. To gate the treatment delivery. For a proton, γ = 2.68 × 108 rad/s/T. Therefore, at a field This involves measuring the motion and turning the strength of 1.5 T, the resonant frequency of the proton is treatment beam on and off at the appropriate phase of m 1 I = + 2 γ γћB0 = hν v = B 2π 0 mI = – 1 2 FIGURE R.52 The energy of a spin-1/2 magnetic moment μ split by a B-field as seen in quantum mechanics. Restricted area 813 Retina the motion. Several signals may be used to control this The special name of the unit of restricted CEMA is gray (Gy) process. External markers placed on the patient sur- expressed in JKg−1. face and x-ray imaging of implanted markers are two Related Articles: CEMA, Stopping power, Restricted stop- examples. ping power 2. To gate the patient’s breathing. Restricted collisional mass stopping power This involves either using a device to hold the patient’s breath at a (Radiation Protection) The restricted collisional mass stop- particular phase of the breathing cycle (usually inhale or exhale) ping power describes how charged particle ionising radiation is or asking the patient to hold their breath at inhale or exhale in absorbed in collisions with atomic nuclei in an absorbing medium, a voluntary breath-hold procedure. The first of these approaches but only those interactions that lead to the total energy of the inci- involves using a spirometer to measure the patient’s breathing and dent particle being absorbed either at the site of the interaction, a valve to hold the breath at the desired phase. Active breathing or very close by, such that there is no radiative loss out. This con- control (ABC) is the system that implements this. The beam-on cept is the most useful at the microscopic level in describing how time of a radiotherapy treatment is generally longer than a patient energy is deposited and leads to absorbed dose. can comfortably perform a single breath hold and thus the treat- For further information, see the articles on Collisional mass ment is often delivered in several breath holds with a pause in stopping power and Absorbed dose. treatment in between. Related Articles: Collisional mass stopping power, Absorbed Abbreviations: ABC = Active breathing control, CT = dose Computed tomography, ECG = Electrocardiograph and 4DCT = Four-dimensional computed tomography. Restricted stopping power Related Articles: Radiotherapy, Intrafraction movement, (Radiotherapy) The stopping power indicates the average rate of Margins, Deep inspiration breath hold, Gating – respiratory energy loss by a charged particle in a medium and this is not always Further Reading: Rietzel, E., T. S. Pan and G. T. Y. Chen. equal to the absorbed dose to the medium particularly if the range of 2005b. Four-dimensional computed tomography: Image forma- the secondary electrons produced is large. In fact a secondary elec- tion and clinical protocol. Med. Phys. 32: 874–889. tron with enough energy, called delta ray, can leave the immediate vicinity of the primary particle path and produce its own track. The Restricted area concept of restricted stopping power has been introduced to associ- (Radiation Protection) Restricted area is a general term used R ate an energy loss in a medium more closely to absorbed dose to the to describe an area or room where access to the area must be medium. The restricted stopping power is defined as the linear rate controlled to avoid persons being harmed. Only persons autho- of energy loss due only to collisions in which the energy transfer rised to do so may enter a restricted area, either following does not exceed a specific value Δ. The extent of the localisation is written procedures, or wearing personal protective equipment determined by the size of the cut-off energy Δ. (PPE). In the United Kingdom, the Ionising Radiation Regulations (IRR99) describes a two-tier system of designation given to Retention fraction restricted areas based on the level of radiation hazard and risk (Nuclear Medicine) Retention fraction is the fraction of a radio- present: pharmaceutical delivered to an organ or tissue that is extracted Controlled Area: ‘Any area in which it is necessary to follow into and retained by the tissue. This fraction is the residual after special procedures to limit exposure, or any area where exposure clearance of the vascular component and the portion of radiophar- to > 3/10 of any radiation dose limit is possible’. maceutical that rapidly diffuses back from the organ or tissue into Supervised Area: ‘Any area in which it is necessary to review blood. the need for future control, or any area where it is possible for an individual to receive an effective dose of 1 mSv y−1’. In other Retina words an area should be supervised if a worker is liable to receive (General) The retina forms the internal, light-sensitive layer of more than the public dose limit. the eye, onto which light is focused by the cornea and lens. The All restricted areas must be appropriately demarcated. electrical signals generated by the interaction of light with the ret- Ordinarily inside a building, a controlled or supervised area will ina’s photoreceptor cells are transmitted to the brain via the optic be defined by the walls of a room which act to shield persons nerve. The point at which this connects to the retina is commonly outside the room from the hazard. However, it may be necessary known as the ‘blind spot’. The retina is approximately 42 mm in to define an area either within a room, or outside the room, or diameter, and 0.5 mm thick. The greatest resolution is achieved indeed the building as a restricted area, and to demarcate the area when light is focussed onto a specialised area of the retina, called to warn persons of its existence. fovea (meaning ‘pit’). The diameter of the fovea is about 1.2 mm. Restricted areas (controlled or supervised) must have warning The retina contains two types of photoreceptor cells: rods and signs to describe the hazard and risk, and to state that only autho- cones. rised persons are allowed to enter. There are around 125 million rod cells, each having a diameter Related Articles: Controlled area, Supervised area of 0.002 mm. They are found mostly in the peripheral retina and are more sensitive to low-level light and hence are responsible for Restricted CEMA peripheral and night vision. (Radiation Protection) Restricted CEMA only includes energy There are approximately 6 million cones with a diameter of lost in electronic interaction up to an energy ∆ so excluding inter- 0.006 mm, the majority of which are found in the fovea. They are actions of electrons with energy greater than ∆. In other terms, less sensitive to light compared with rods. However, the cones are the energy loss per unit mass excludes energy losses above a cut- responsible for colour vision. There are three types of cones – red- off ∆. sensitive, green-sensitive and blue-sensitive. Retinal hazard region 814 Reynolds number Related Articles: Cones, Rods, RGB, Visual acuity, Visual perception Further Readings: Hecht, E. 1987. Optics, 2nd edn., Addison A B C D E Wesley; Rods and cones, http: / /hyp erphy sics. phy -a str .g su .ed u/ %E2 %80 %8 Chbas e /vis i on /r odcon e .htm l (accessed 28 Feb 2013); http://en .wikipedia .org /wiki /Retina; Tabakov, S. 2013. Introduction to vision, colour models and image compression. J. x Med. Phys. Inter 1:50–55. Retinal hazard region (Non-Ionising Radiation) The spectral range between 380 and A Pulse transmitted 1400 nm (visible and infrared A). These optical wavelengths may B Interface – some sound reflected travel to the back of the eye and therefore have the potential to C Reflected pulse is itself partially reflected from

transducer/tissue cause damage to the retina. interface D Re-reflected pulse now reflected back from interface Retrospective ECG gating E Image shows echoes from interface and an additional echo (x) from the re-reflected pulse (Diagnostic Radiology) Retrospective ECG gating is used dur- ing 4D scanning with multidetector computed tomography (MDCT). 4D CT (aka dynamic imaging) allows visualisation of FIGURE R.53 Reverberations are caused by echoes from re-reflection, fast dynamic processes such as blood flow. 4D CT reconstruc- often from the tissue/transducer interface. The example shows the ori- gin of a single reverberation from a reflector in tissue and the transducer tion requires the acquisition of scans to be synchronised with the surface. cardiac cycle. One method for this synchronisation is the ‘prospective ECG gating’, when the scanning data is acquired only at specific phases in the cardiac cycle (i.e. the x-ray tube produces radiation only at specific ECG phases). This requires very fast scanning (currently R the assembly of x-ray tube with multidetectors completes over five rotations per second, in continuous rotation). The image of the heart is reconstructed at only one cardiac cycle. During 4D CT, the scanning data is obtained continuously. Simultaneously, the ECG data is obtained continuously, matching each scanning data with the respective ECG phase (ECG gating). After the end of the examination, the scanning data is sorted in relation to respective ECG phases, and the reconstruction is made only of scanning data during identical ECG phases. This pro- cess – ‘retrospective ECG gating’ – makes possible reconstructed images of moving objects – e.g. valvular motion. Related Article: Four-dimensional (4D) computed tomogra- phy (CT) data Further Reading: Mehndiratta, A., S. Bartling and R. Gupta. 2017. 4-D X-ray computed tomography, in P. Russo (eds.), Handbook of X-ray Imaging: Physics and Technology, CRC Press. Reverberation FIGURE R.54 Reverberation echo (arrowed) evident in a jugular vein (Ultrasound) Reverberations are artefacts caused by re-reflection from strong echoes on the anterior wall of the vein and the transducer. of reflected ultrasound echoes. They occur most commonly in the near field if strong reflecting interfaces are present normal to the beam direction, parallel to the beam. For flow in a tube, Reynolds number is defined as Re-reflection often occurs at the tissue-transducer inter- face due to the mismatch of acoustic impedance at this surface rVD (Figure R.53). Reverberation artefacts are most obvious if there Re = m are weak echoes from direct imaging of the tissue, for example in veins, cysts and amniotic fluid (Figure R.54) but they can also where obscure detail in greyscale imaging (Figure R.55) by superim- ρ is the density of the fluid posing extraneous echoes to those arising from the target tis- V is the velocity sue. Closely spaced reverberations from small structures can be D is the diameter of the tube described as comet-tail artefacts. μ is the fluid viscosity Related Articles: Acoustic impedance, Reflection, Comet tail The number is dimensionless and the higher the number the Reynolds number greater the tendency towards turbulence. In a tube with steady (Ultrasound; General) The Reynolds number describes ratio of laminar flow the transition to turbulent flow occurs at a Reynolds inertial forces to viscous forces in fluid flow and so indicates their number of about 2300 although the precise value is dependent on relative dominance in given flow conditions. local conditions. Around this value, flow is transitional, neither RF (radiofrequency) 815 R F pulse The body coil encloses the cylindrical bore internally in the scan- ner and surrounds the patient. A coil that acts as both a transmitter and receiver is called a transceiver. An RF coil could be as simple as a loop of wire with a tuning capacitance added to set the appropriate resonant peak. In prac- tice coil designs adopt a range of conductor geometries in order to maximise SNR or uniformity for a given application. Standard geometries used in practice include coils based on solenoid, bird- cage and saddle designs. RF echo (Magnetic Resonance) A radiofrequency (RF) echo is a rarely used term for ‘spin echo’ to distinguish it from a gradient echo. A few authors use RF echo for a spin echo where the flip angle of the refocusing pulse is less than 180° and greater than 90°. Related Articles: Radiofrequency, Spin echo FIGURE R.55 Reverberation echoes evident in the superficial liver Further Reading: Hahn, E. L. 1948. Spin echoes. Phys. Rev. (ringed) from the liver surface and other subcutaneous interfaces. 80, 580 (1950). Hennig, J. 1991. Echoes: How to generate, recog- nize, use or avoid them in MR-imaging sequences Part I, Concepts in Magnetic Resonance, 1991, 3, 125–143. Part II, Concepts in laminar nor turbulent. Flows with Reynolds numbers <2000 are Magnetic Resonance, 1991, 3, 179–192. laminar and >3000 are turbulent. In the case of pulsatile flow the peak velocity may be used and the number is described as the peak Reynolds number. RF pulse Related Articles: Laminar flow, Turbulent flow (Magnetic Resonance) A radiofrequency pulse is a controlled emission of electromagnetic radiation in the radio frequency part of the spectrum. It is an important feature of MRI because the RF (radiofrequency) resonant frequencies of the nuclei being used lie in this part of (Magnetic Resonance) See Radiofrequency (RF) the spectrum. R This article describes basic effects of ‘hard’ (rectangular) RF RF coil pulses from a classical point of view: for a quantum mechanical (Magnetic Resonance) An RF coil in MRI receives and/or trans- explanation see the Resonant frequency article. mits radiofrequency energy into the patient. A standard MRI sys- Classically, under the influence of an external magnetic field tem will have many RF coils, each appropriate to a specific range B0, nuclear spins precess with the Larmor frequency ω0: of imaging tasks. An RF coil forms a tuned circuit with a resonant peak. Coils w0 = gB0 are designed to resonate at the Larmor frequency for the MRI field strength in question, that is for a 1.5 T MRI, coils will have where γ is the gyromagnetic ratio for a specific nucleus. a resonant peak for protons at approximately 64 MHz and lower In NMR only the magnetic component B1 of the RF pulse is for other nuclei. considered in the rotating frame. When B1 is applied on the res- For a receiver coil it is desirable that the coil displays high onant frequency and perpendicular to B0 the net magnetisation SNR, good uniformity and adequate coverage. Good uniformity Mz will experience a torque which causes it to rotate towards the equates to a uniform gain or sensitivity to signal throughout a tis- transverse plane. sue volume. A volume coil is designed to enclose the anatomy of The angular frequency of nutation by the additional field B1 is interest and provide high uniformity. A surface coil is designed to be placed in proximity to more superficial anatomy (e.g. the w1 = gB1 spine). A surface coil provides better SNR for the anatomy image but a reduced field of view and reduced uniformity. from which the following expression can be obtained: In a transmit coil, the coil generates a magnetic field at right angles to the main static B0 field when the RF system applies q a pulse to the coil. This is called the B1 field. A component of t = gB this magnetic field rotates in the same sense as the direction of 1 precession and causes spins to flip through an angle. In order to which relates the duration of the RF pulse (t), to the angle through generate uniform flip angles throughout a volume, the energy which Mz is rotated (θ in rads). deposited by a coil should be uniform throughout the tissue For maximum MR signal, the transverse component of Mz is volume. maximised, such that it is rotated through 90° from the z-axis into In coil design, coil transmission and reception characteristics the x–y plane. Using the equation above the exact duration of the are linked by the principle of reciprocity. If a coil demonstrates RF pulse to achieve this rotation can be calculated by setting θ = good uniformity of sensitivity while acting as a receiver, the same π/2. Such a pulse is known as a 90° RF pulse. coil will also demonstrate good uniformity in generating flip Another key pulse is the 180° pulse which is either used to angles if acting as a transmitter. invert longitudinal magnetisation, or for refocusing transverse Because surface coils demonstrate poor uniformity, they are magnetisation to correct for dephasing caused by the imperfect not normally used as transmit coils. The body coil is generally B0 field. used as the transmit coil for surface coils and other volume coils. Related Article: Resonant frequency RF uniformity 816 Rhenium-186-hydroxy-ethylidene diphosphonate RF uniformity Atomic weight 186.207 kg/kg-atom (Magnetic Resonance) Radiofrequency uniformity can be divided Electronic configuration 1s2 2s2 2p6 3s2 3p6 3d10 4s2 4p6 4d10 into RF transmission uniformity and RF reception uniformity. 4f14 5s2 5p6 5d5 6s2 The RF uniformity is governed by which RF coil is used. The Melting point 3459 K best RF homogeneity is obtained by using the most appropriate Boiling point 5869 K coil and ensuring that the coil is positioned correctly with respect to the patient and perpendicular to the static magnetic field. Density near room temperature 20,800 kg/m3 Uniformity is important, especially for transmit and receive coils as poor uniformity will affect RF flip angles and therefore History: Rhenium is a high-density, high-melting-point sil- image contrast. Large transmission coils give a very uniform very grey transition metal. It was first discovered in 1925 by spec- transmission field but are not very sensitive for receiving the sig- troscopic analysis of certain mineral ores in which it is present in nal. Receive only coils are often non-uniform in design to obtain trace amounts. In 1928 the discoverers succeeded in extracting 1 the best SNR. Surface coils are the most prone to non-uniformity, g of the metal from 660 kg of molybdenite ore. Rhenium is still providing high SNR close to the surface at the expense of non- obtained only as a by-product of other metal refining processes uniformity. Surface coils are thus typically used to image ana- and is one of the world’s most expensive metals. tomical structures close to the surface of the patient. Medical Applications: X-ray tube target – In a rotating anode To measure the effects of RF uniformity, a uniform phantom is x-ray tube, the focal track for electrons is usually made of a mix- imaged. The phantom should be loaded to simulate the presence ture of tungsten and rhenium. Both of these materials have a high of a patient. An ROI covering 75% of the slice is then analysed. atomic number, a high specific heat capacity and a high melting Integral uniformity equation is given by point. However, as tungsten has low linear expansivity it is prone to crazing with repeated expansion and contraction. Rhenium has (M + m) a higher linear expansivity and thus slows the rate at which craz- I = 1- ( ´100% ( M - m) R.2) ing occurs. Radionuclide therapy – In nuclear medicine therapy rhenium where M is the maximum pixel value and m is the minimum pixel is used in two isotope forms: 186Re and 188Re. value within ROI of 75% of the phantom area. 100% is perfect R uniformity. Abbreviations: RF = Radiofrequency, ROI = Region of inter- Isotope of rhenium 186Re est and SNR = Signal-to-noise ratio. Half-life 89 h Related Articles: B1 homogeneity, B1 inhomogeneity Maximum decay energy, Emax β: 1.07 MeV, γ: 137 keV Isotope of rhenium 188Re R Half-life 17 h f value (Nuclear Medicine) Rf value is the retention factor, a value used Maximum decay energy, Emax β: 2.10 MeV, γ: 155 keV in thin-layer chromatography for identification of components in radiopharmaceuticals. The Rf value is defined as The beta particles emitted from 186Re have a range of 5 mm in Migrationdistanceof thecomponent water whilst those emitted from 188Re have a range of 11 mm, Rf = Migrationdistanceof thesolvent thus 186Re is typically used for small tumours whilst 188Re is more appropriate for larger masses. Whilst 186Re is generated via neutron radiation of 185Re, 188Re The values range from 0 to 1 and are characteristic for a given is more readily produced in a generator from tungsten 188W,

thus compound provided that the same stationary and mobile phases providing an efficient and cost-effective source of radioisotope. are used. Related Articles: X-ray tube, X-ray tube assembly, Further Readings: Kowalsky, R. J. and S. W. Falen. 2004. Radionuclides in therapy, Generators, Radionuclide Radiopharmaceuticals in Nuclear Pharmacy and Nuclear Medicine, 2nd edn., American Pharmacists Association, Washington, DC; Saha, G. B. 2004. Fundamentals of Nuclear Rhenium-186-hydroxy-ethylidene diphosphonate Pharmacy, 5th edn., Springer, New York; Zolle, I. ed. 2007. (Nuclear Medicine) Rhenium-186-hydroxy-ethylidene diphos- Technetium-99m Pharmaceuticals – Preparation and Quality phonate, (186Re-HEDP) has similar uptake mechanisms to 99Tcm- Control in Nuclear Medicine, Springer, Heidelberg, Germany. HEDP and 153Sm-EDTMP. It is used for pain palliation of bone metastases, although it is not as common as 89Sr-Metastron or 153Sm-EDTMP. 186Re emits a beta particle with maximum energy RGB (Red, Green, Blue) 1.07 MeV, mean energy 0.349 MeV, average soft-tissue range 1.1 (General) See Red, green, blue (RGB) mm and a gamma photon of energy 137 keV (9%). The physical half-life is 3.72 days. Whereas the beta radiation is used to irradi- Rhenium ate the cancer, the emitted gamma radiation can be used simulta- (General) neously for imaging and quantification of the 186Re uptake. The absorbed dose to bone surfaces and red bone marrow are Symbol Re 1.4 and 1.3 mGy/MBq, respectively. Other organs with significant absorbed doses are the walls of LLI and urinary bladder, 0.57 Element category Transition metal and 0.54 mGy/MBq respectively. Most other organs receive about Mass number A 186 0.02 mGy/MBq. The effective dose for 186Re-HEDP is approxi- Atomic number Z 75 mately 0.3–0.4 mSv/MBq. Rheostat 817 R ing artefact Further Readings: De Klerk, J. M. H., B. A. Zonnenberg and Related Articles: Artefact, Beam hardening, Computed G. H. Blijham et al. 1997. Treatment of metastatic bone pain using tomography, Cone beam artefact, Helical artefact, Image artefact, the bone seeking radiopharmaceutical Re-186-HEDP. Anticancer Metal artefact, Motion artefact, Partial volume effect (artefact) Res. 17: 1773–1778; Firestone, R. B. 1999. Table of Isotopes, 8th edn., John Wiley & Sons, Inc., New York. Update with CD-ROM. http://ie .lbl .gov /toi .html; Kowalsky, R. J. and S. W. Ring artefact Falen. 2004. Radiopharmaceuticals in Nuclear Pharmacy and (Magnetic Resonance) The ring artefact is similar to Gibb’s arte- Nuclear Medicine, 2nd edn., American Pharmacists Association, fact. Bands of high- and low-intensity pixels form parallel to a Washington, DC; Palmedo, H., J. K. Rockstroh, M. Bangard, K. high contrast edge due to undersampling of the data in the phase- Schliefer, J. Risse, C. Menzel and H.-J. Biersack. 2001. Painful encoding direction. multifocal arthritis: Therapy with rhenium 186 hydroxyethyl- The high contrast edge in question is caused by the introduc- idenedi-phosphonate (186Re HEDP) after failed treatment with tion of gadolinium into the arteries during contrast enhanced medication – Initial results of a prospective study. Radiology MRA. If the centre of k-space is being sampled when the signal 221:256–260. intensity changes rapidly, the line in k-space will have a different Related Articles: Tc-99m-diphosphonates, Sm-153-EDTMP contrast weighting, causing the artefact to occur. [Lexidrom], Sr-89-chloride [Metastron] Abbreviation: MRA = Magnetic resonance angiography. Related Article: Gibb’s artefact Rheostat (General) See Resistance, electrical Rigid stem chamber (Radiation Protection) The rigid stem chamber is a cylindrical ionisation chamber to be used for absolute dose measurement in radiotherapy with high-energy photon and electron beams. The chamber is not waterproof and can be used in solid state phantoms for absorbed dose to water, air kerma and exposure measurements. Related Article: Cylindrical ionisation chamber R Ring artefact (Diagnostic Radiology) Ring or band artefacts in a CT image are a result of an incorrect signal from a detector channel or a group of detector channels (Figure R.56). This could be due to faulty detectors, a problem in the data acquisition system electronics, or incorrect calibration. The artefact is mainly seen on third genera- tion CT scanners. Figure R.57 illustrates the origin of these artefacts. The path Detector out of an x-ray from the source to a particular detector is at a fixed of calibration distance from the isocentre for all angular positions. A defective detector will therefore form a ring pattern during the reconstruc- FIGURE R.57 Illustration of origin of ring artefacts on third generation tion process. CT scanners. (Courtesy of EMERALD project, www .emerald2 .eu.) (a) (b) FIGURE R.56 Appearance of ring artefacts (a) in a water phantom, (b) in air. (Courtesy of ImPACT, UK, www .impactscan .org) Ring artefact 818 Risk communication Ring artefact reach maximum amplitude, in which case this definition may be (Nuclear Medicine) A common image artefact. As the name misleading. indicates the phenomenon is one of a series of rings in the image Rise times are typically on the order of a few hundred micro- caused by a systematic error. The source of error differs between seconds on modern clinical MRI systems. camera systems and some image reconstruction methods can Related Article: B0 gradients cause ring artefacts. For example, using window filters with sharp edges when reconstructing can lead to ring artefacts (also called Risk assessment Gibbs phenomena). (Radiation Protection) Risk is defined as the chance, probability Further Reading: Cherry, S. R., J. A. Sorenson and M. E. or likelihood that a person may experience an adverse effect from Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, exposure to a hazard. Risk can be quantified as a risk coefficient Philadelphia, PA, p. 490. in several ways – for example the risk may be quoted as 1 in 10, 10%, 0.1 or 1 × 10−1. Ring down A hazard, or health hazard, is defined as anything that has (Ultrasound) Ring down is a reverberation artefact where a pair the capacity to cause harm to a human being. Such harm can be of closely spaced reflectors, most commonly the edges of a gas through physical injury, biological damage or other health det- bubble, cause a continuous echo in the image from the reflectors riment (adverse effect) in the form of cancer or other disease. in the direction of the beam. Examples of hazards include fire, chemicals, electricity (electro- Related Articles: Comet tail, Reverberation cution), water (drowning), and exposure to ionising or non-ionis- ing radiation. Ripple Risk assessment is generally accepted to be a five-stage (Diagnostic Radiology) See Voltage waveform process: • Identify the hazard – what is the source of the hazard Ripple factor (e.g. fire, ionising radiation exposure, etc.). (Diagnostic Radiology) The ripple factor is maximal change • Identify those who might be harmed – there will be (drop) of kV, compared with the maximal kV (kVp). The ripple ‘critical groups’ of individuals who may be most at risk factor is expressed as a percentage of the kVmin from kVmax (kVp). R of suffering harm from exposure to the hazard. It is The factor is directly related with the assessment of the kV wave- therefore most important to consider reducing the expo- form of an x-ray generator. For more information see the article sure of these individuals first. on Voltage waveform. From there one can see that a two pulse kV • Quantify the risk – what are the chances that the criti- waveform will have 100% ripple factor (the kVmin drops to zero), cal groups of individuals identified earlier may come while a six pulse kV waveform will have 14% ripple factor (14% to harm. pulsations). A high-frequency (medium frequency) x-ray genera- • Identify control measures – decide on an action plan to tor will have ripple factor below 3%. reduce the risks of harm to acceptable levels. Related Article: Voltage waveform • Review the efficacy of actions taken – analyse the effect of the control measures and revise the action plan if Rise time necessary. (Magnetic Resonance) When a magnetic field gradient is switched on during an MRI pulse sequence, it takes a finite For example, in radiation protection, control measures can be length of time for the current passing through the gradient coils considered as a hierarchy of systems in place to reduce radiation to rise from zero and hence for the gradient field itself to reach dose to individuals. One can consider the boundaries of a room the required value. This period of time is known as the rise time, containing the radiation hazard, the walls, ceiling, floor, and the and is dependent on the resistance and inductance of the gradient doors. Shielding the boundaries or through restriction of access and coils (Figure R.58). the use of interlocks to prevent a person from being exposed to ion- Shortening the rise time allows for shorter gradient pulses, ising radiation may be an ideal solution to reduce radiation dose. which increases imaging speed and can be advantageous in other If however it is necessary for a person to be in the room during ways for specific MRI applications. Thus the minimum rise time the exposure to carry out specific procedures, then it will be nec- to maximum gradient amplitude is an important system perfor- essary to establish a system of work for the procedures. Finally, if mance parameter. the system of work is not able to reduce doses to acceptable levels, The related parameter of gradient slew rate is often defined then personal protective equipment must be used (lead aprons, as the maximum gradient amplitude divided by the minimum etc.). rise time. However, the maximum slew rate that can be delivered Related Articles: Absolute risk, Adverse effect, Hazard, by a system may not be sustainable over the rise time needed to Radiation risk, Risk coefficient Risk coefficient (Radiation Protection) Risk coefficient is the quantification of the chance or probability that a hazard may occur, for example the risk may be quoted as 1 in 10, 10%, 0.1 or 1 × 10−1. Related Article: Risk assessment Rise time Risk communication (Radiation Protection) Risk communication plays a key role in FIGURE R.58 Gradient rise time. helping organisations apply the concepts of risk management to Risk Group 1, 2, 3 (Photobiological optical source) 819 Robotic linac their daily operations. It refers to a two-way exchange of infor- • Changing control measures (e.g. safety by design, mation, advice and opinions between experts and people facing alarms) when required, based on the assessment of pro- threats. In the radiation medicine field it refers to the exchange duction and post-market information. of information, advice and opinions about examinations, therapy and their radiation doses between medical staff and patients and Related Articles: Standards, Life cycle of equipment medi- their families/relatives. cal equipment management, Maintenance, Clinical engineering, The purpose of risk communication is to enable people to Medical equipment management make informed decisions to protect themselves and their loved Further Readings: Medical Device Regulations 2017/745; ones. In other words, the experts must tell people about the risks In-Vitro Diagnostics Regulations 2017/746; ISO 14971; Brosseau, related to a specific hazard. People must also provide feedback B. 2018. Risk Management Under the New EU Medical Device so that they can come up with appropriate mitigation measures. Regulation; Miniati, R., E. Iadanza and F. Dori. 2015. Clinical Risk communication requires an understanding of people’s per- Engineering: From Devices to Systems. Academic Press. ceptions, concerns and beliefs as well as their knowledge and practices. It also includes the early identification and management of rumours, misinformation, etc. Risk perception (Radiation Protection) Risk perception is the subjective assess- ment of the probability of a specified type of accident happening Risk Group 1, 2, 3 (Photobiological optical source) and how concerned we are with the consequences. (Non-Ionising Radiation) See Photobiological lamp safety Road mapping Risk management (Diagnostic Radiology) Road mapping (RM) can be seen as (General) This is the process of managing risks aiming at iden- advanced last image hold (LIH). RM is used in angiography with tifying, monitoring and, thus managing the possible risky events digital fluoroscopy systems. Initially an image of the vessels is which might negatively affect a certain system and at ensuring the made by injecting small amount of contrast media in the vessels maintenance of the conditions such that the system is preserved. It where the catheter will be introduced. This ‘vessels map’ is memo- comprises three phases: risk identification, risk analysis (e.g. haz- rised (as LIH) and is later superimposed over the live fluoroscopy ard analysis, fault tree analysis, failure mode and effects analysis during the introduction of the catheter. This way the radiologist/ etc.) and risk response. angiologist

has good guidance through the patient vessels anat- R In the healthcare field, risk management is essential to ensure omy. The RM method is primarily based on software. patients and staff’s safety and the correct, safe, efficient and Related Articles: Last image hold, Digital fluoroscopy effective operation of a medical device. The devices, in fact, must never compromise safety and the individual and cumulative risks Roam and zoom should always be prevailed by the clinical benefits. (Diagnostic Radiology) High-resolution digital images (as digital In the medical device field, the risk management is stan- mammography) require special display screens. However even dardised by the ISO 14971, which specifies ‘a process for a manu- the best contemporary monitors are not large enough to display facturer to identify the hazards associated with medical devices, at once a full digital mammogram with its best resolution. Due to including in vitro diagnostic (IVD) medical devices, to estimate this reason the image has to be first zoomed (to get the best resolu- and evaluate the associated risks, to control these risks, and to tion) and after this displayed partially through the roam function monitor the effectiveness of the controls’. of the equipment. The new European medical device regulations 2017/745 Further Reading: Beutel, J., H. Kundel and R. Van Metter, are in line with the above-mentioned standard. They contain eds. 2000. Handbook of Medical Imaging, v.1 Physics and more explicit requirements for the manufacturers to have in Psychophysics, SPIE Press, Washington, DC. place and follow a risk management programme and life- cycle risk management. Risk management is thus an iterative Robotic linac process that follows the life-cycle of the medical device and (Radiotherapy) Various technologies have been developed to comprises: improve both the dose distribution and positional accuracy of radiotherapy. The robotic linac is a concept which attempts to • A risk management plan for each medical device; achieve both goals. Such systems are often used for radiosurgery, • Identification, evaluation and analysis of possible in which the treatment is delivered in a single fraction. Examples threats associated with each device; of robotic linacs include the Accuray Cyberknife. • Estimation of the risk associated with the intended use Robotic Delivery System: The delivery system consists of an and/or misuse of the device. In this stage, we often industrial robot with a small, x-band linear accelerator. The accel- quantify the risk as the product of the probability of an erator may be orientated at a variety of angles in three dimen- event and its potential damage; sions to enable the target to be irradiated over a large solid angle • Risk response and elimination. The aim of this stage (typically 2.5p steradians) with the total dose delivered in many is to remove the risks that can be eliminated and man- beamlets with small dose irradiations over numerous positions. age those that cannot, adopting appropriate actions (e.g. Verification System: The position of the patient’s anatomy for training, liming the access, generating warnings, adopt- each beamlet is often verified with an imaging system. This is ing relevant visual codes); usually a combination of projection x-rays and fiducial markers • Assessing the risks associated with the production and for each beam position coupled to more continuous imaging of improving the process for the post-market information the external anatomy using markers placed on the patient’s skin. and relevant data to be collected in the documented risk Related Articles: Image-guided radiotherapy, Intensity modu- assessment; lated radiotherapy, Radiosurgery ROC (receiver operating characteristic) 820 Rose model Further Reading: Adler, J. R., Jr., S. D. Chang, M. J. Murphy, Related Articles: Absorbed dose, Parallel plate ionisation J. Doty, P. Geis and S. L. Hancock. 1997. The Cyberknife: A fra- chamber meless robotic system for radiosurgery. Stereot. Funct. Neurosurg. Further Reading: http://www .ptw .de 69: 124–128. Root mean square (RMS) voltage ROC (receiver operating characteristic) (General) The voltage value that has the same effect and gives the (Diagnostic Radiology) See Receiver operating characteristic correct result on a power calculation as does a DC voltage of the (ROC) same value. RMS voltage is equal to the square root of the mean value of the squares of the magnitudes of an AC voltage measured Rods, retina at each instant over a defined period of time, usually one cycle. (General) The retina contains two types of photoreceptor cells The RMS voltage is also known as the effective voltage (see the – cones and rods. The cones are associated with colour vision. eponymous article). The rods are very sensitive to light (about 100 times more sensi- Related Article: Effective voltage value tive than cones) and provide vision at low intensity levels (twilight and night vision). The rod cells have sensitivity between blue and Rose model green (but closer to green) with a peak at about 498 nm. The max- (Diagnostic Radiology) The Rose model is a classical model imum sensitivity of the human eye is in the region of the green- describing image quality based upon quantum noise and its effect yellow colour. on human perception. It is used in medical imaging as a simplistic The rod cells in the human retina are about 120 million (the model to describe the quantum noise of a diagnostic image and three types of cone cells are about 6 million). The diameter of the how it affects the detection of a signal. The model is of limited rods is c. 0.002 mm, while the diameter of cones is c. 0.006 mm. use in its original form as it is only valid for low contrast, uncor- The rod cells are distributed throughout most of the surface of related Poisson distributed noisy backgrounds and simple objects. the retina. Their density drops sharply in the region of the fovea, However, this model forms the basis for more advanced image where the cone cells dominate. metrics based upon Fourier analysis such as noise power spec- Related Articles: Retina, Cones, Visual acuity, Visual trum (NPS) noise equivalent quanta (NEQ) and detective quan- perception tum efficiency (DEQ). R Further Readings: Hecht, E. 1987. Optics, 2nd edn., Addison The model was first proposed by Albert Rose in 1948, and Wesley; Rods and cones, http: / /hyp erphy sics. phy -a str .g su .ed u/ describes the minimum SNR needed to visually detect a uniform %E2 %80 %8 Chbas e /vis i on /r odcon e .htm l (accessed 28 February circular object in a uniform background. Historically, Rose’s 2013); http://en .wikipedia .org /wiki /Retina; Tabakov, S. 2013. model was important as it was the first to evaluate the perfor- Introduction to vision, colour models and image compression. J. mance of imaging devices by quantum efficiency using an abso- Med. Phys. Intern. 1:50–55. lute scale. Rose investigated the detection, by human observers, of cir- Roentgen (R) cular objects of differing sizes and contrast within a Poisson (General) This was the first quantitative measure of radiation distributed noisy background. The observers were required to initially defined by the ICRU (International Commission on indicate whether the objects were visible or not, while already Radiation Units) in 1928 as the quantity of x-rays when the sec- knowing the presence of the object. This is similar to the Leeds ondary electrons are fully absorbed and any wall effects avoided test phantom used in diagnostic radiology quality control today, would produce 1 esu of charge in 1 cc of air at 0°C and 76 cm see Figure R.59. mercury of pressure. The Rose model assumes an uncorrelated Poisson distributed The unit was later redefined (1937) as the quantity of x or noise throughout. If an image had a uniform background with gamma radiation producing 1 esu of charge per 0.001293 g air. (The mass of air is that which is associated with 1 cc of air.) This unit was subsequently replaced by the SI unit of exposure in C/kg: • 1R = 8.69 mGy • 1R = 0.258 mC/kg The unit is named after Wilhelm Rontgen German physicist, dis- coverer of x-rays (1845–1923). Related Article: Exposure ROI (region of interest) (Nuclear Medicine) See Region of interest (ROI) Roos chamber (Radiation Protection) The Roos chamber is a waterproof plane parallel ionisation chamber used for absolute dosimetry in high- energy electron and proton beams as well as for measuring photon depth dose. This chamber can be used in water or in solid state FIGURE R.59 Leeds test phantom. (Courtesy of EMERALD project, phantoms for absorbed dose to water measurements. www .emerald2 .eu) Rotating anode 821 R otating anode mean quanta per area qB, and an object with mean quanta per area of qO , then the contrast is defined as (qB - qO ) C = qB The signal S of the object is then defined as the difference between the mean quanta of the background and the mean quanta of the object, integrated over the area of the object, such as 1 S = A(qB - qO ) 2 The noise in the signal is defined as the standard deviation of 1 the number of quanta in a section of uniform background of area 3 A, σB. For an uncorrelated background this noise is described by Poisson statistics so that FIGURE R.60 Rotating anode: (1) Real focus (momentary focus); (2) thermal focus (thermal path or thermal track); (3) effective (optical) focal sB = AqB spot. Thus, the SNR is given by A(qB - qO ) SNR = = C AqB AqB Rose deduced that a SNR of 5 or above is needed to detect an object under these conditions. In similar circumstances, such as that using R the test phantom, with low contrast and uncorrelated Poisson dis- tributed noise, this relationship has been shown to be correct. Before this model could be used clinically all limitations must be considered. For, one of the main practical limitations of the model is the definition of noise, as clinical images rarely have Poisson distributed noise. For example, electronic noise is usually additive and secondary quanta such as that produced by a phos- phor screen shows statistical correlations. Also, anatomical noise can inhibit the detection of signal (pathology) within an image and this has been shown to increase with contrast, whereas the detection of the Rose model object should decrease. Related Articles: Anatomical noise, Detective quantum effi- ciency (DQE), Noise equivalent quanta (NEQ), Noise power spectrum (NPS), Signal to noise ratio (SNR) Further Readings: Burgess, A. E. 1999. The rose model, revisited. J. Opt. Soc. Am. A16: 663–646; Rose, A. 1948. The sensitivity performance of the human eye on an absolute scale. J. Opt. Soc. Am. 38: 196–208. FIGURE R.61 Model of the stator of x-ray tube with rotating anode (medium power diagnostic tube). (Courtesy of EMERALD project, www Rotating anode .emerald2 .eu) (Diagnostic Radiology) The first x-ray tube with rotating anode has been produced by Philips in 1929. The main idea of this tube is that the electrons hit different parts of the rotating target, thus The rotating anode is mounted on a stem, which is attached to increasing many times the size of the thermal focus (distributing the rotor of an electrical motor (see the article Anode). The stem the heat in a long target track). At present these are the most- and the base of the rotating anode are made from tungsten–- often-used x-ray tubes (with many different designs). Figure R.60 -molybdenum alloy, which is light and decreases the undesirable and R.63 show schematically a rotating anode, with indication of inertia of the spinning anode. Thin tungsten–rhenium alloy (1–2 its main parts and the three main foci of a rotating anode. mm thick) is coated over the molybdenum alloy, thus forming the The maximal permissible power of the rotating anode (Pmax) rotating target (the target track). In these x-ray tubes the rotating depends on the effective focal spot size ( f), the diameter of the anode and the rotor are within the glass envelope (supported by target track (D), the angle of the anode (α) and the speed of rota- ball bearings), and the stator producing the rotating magnetic field tion (n − rpm): is fitted outside the glass envelope (Figure R.61). The diameter of the rotating disk varies but often is 100–130 mm. f 3/2D1/2n1/2 Increasing the speed of rotation (most often tripling the normal Pmax ~ sina ∼3000–9000 rpm) leads to increase of the

tube power because the Rotating frame 822 Rotating frame bombarded area of the track passes quickly over the fixed elec- etc. Other rotating anodes have radial slits on the tungsten surface tron beam area and has time to cool during the remaining rotation. that diminishes the effect of thermal expansion and contraction of Further demand for long sequences of powerful exposures (espe- the metal (thermal stress). Figure R.63 shows only four such slits, cially in angiography and CT) lead to constructing rotating anodes while usually there are eight slits or more. with bigger heat storage capabilities. For this purpose the rotating Usually rotating anode x-ray tubes have two focal spots (double disk is often constructed by molybdenum with embedded graphite focus, or dual filament tubes) and in this case special care is taken compound, which is used as thermal accumulator, thus increasing for the positioning of the two cathode filament wires. Figure R.63 the heat absorbing volume. This compound has the advantage to be presents the case when the two filaments overlap, this way creat- cheaper and lighter than the solid molybdenum, thus diminishing the ing an area of the anode target, which is heated by both foci. This inertia and decreasing the load on the anode ball bearings (Figure area is most likely to crack quickly (due to thermal stress), which R.62). The thickness of the molybdenum base of the anode disk var- will decrease the life of the tube. This could be avoided by differ- ies normally from 5 to 15 mm, and the graphite compound could add ent positioning of the filament coils. similar thickness to the anode disk base. This way the overall weight Related Articles: Anode, Stationary anode, Target, X-ray tube of the anode of a powerful x-ray tube could reach 2 kg. housing, Anode angle, Dual filament tube, Bearing Although most of the rotating anodes use the basic design described earlier, there are many different designs of diagnos- Rotating frame tic x-ray tubes. Some improvements of the design of x-ray tubes (Magnetic Resonance) It is often helpful to consider the nucleus in include use of titanium, zirconium or molybdenum (in mammog- a different frame from the laboratory frame, especially to under- raphy) as target, special technologies for building of the target, stand the action of RF pulses. (a) (b) R FIGURE R.62 X-ray tube with rotational anode with graphite compound below the molybdenum body (a), compared with the same anode without graphite (b), but with less thermal capacity. in radial cut Anode disk W-Re coating ermal path (LF+SF) ermal path (LF) Mo anode stem α-Anode angle α LF SF Actual focal spot Cathode Graphite layer FF W-Re target coating BF Mo anode body Effective focal spot Central x-ray beam towards the patient FIGURE R.63 Rotating anode with its main parts and focal spots. Note the superimposed actual focal spots from both filaments (large LF and small SF), which create two effective focal spots – broad focus (BF) and fine focus (FF). Rotational 3-D scanning 823 R adiation Protection of Patients (RPOP) website z z z΄ ω0 y x΄ ω y 0 y΄ B1 x x x΄ (a) (b) y΄ FIGURE R.64 On the left hand side is a representation of the laboratory frame where the spin precesses around the z-axis (external B field) with frequency ω0. On the right hand side is a representation of the rotating FIGURE R.65 The RF pulse described by a non-rotating field vector, frame (shown in dark grey) where the whole frame rotates around the B1, in the transverse plane. z-axis with frequency ω0. In the rotating frame the spin is stationary. Rotational 3-D scanning Considering the main magnetic field, B0 which is aligned along (Nuclear Medicine) Rotational 3-D scanning is another way of the z-axis, and causes spins to precess at the Larmor frequency: describing single photon emission computed tomography or SPECT. A gamma camera system consisting of one or more heads w0 = gB0 (R.3) is rotated around the patient, and a number of images are obtained at equal angular intervals over 360° or 180° (depending on the If a frame is chosen which rotates at the same angular frequency area of the body being scanned). From these images a series of as the spin (ω0), then the spin appears not to precess, and thus count profiles for each transaxial slice are extracted and recon- according to (R.2) the axial magnetic field in this frame must be structed using filtered back projection or iterative reconstruction. zero. This frame is known as the rotating frame, and by choosing it, These transaxial slices can then be combined using computer R the main, static magnetic field is effectively cancelled out. Its axes software to provide a three-dimensional image. are usually marked by a prime. The rotating frame corresponds to Related Article: Single photon emission computed tomogra- the interaction picture in quantum statistics (Figure R.64). phy (SPECT). The second field to be considered is the magnetic component of the radio frequency control pulse, which is a linearly pola- Rotor rised field in the plane transverse to the static field. A linearly (General) A rotor is the rotating part of an induction electrical polarised field may be considered as a pair of counter rotating motor. In x-ray tubes with rotation anodes the (usually copper) circularly polarised fields, and it is easier to apply the frame rotor is connected to a (molybdenum) stem rotating the anode (see transformation to these components separately rather than both the article Rotating anode). The rotating anode and the rotor are together. In the rotating frame, the result is two magnetic fields within the glass envelope (supported by ball bearings), and the rotating at angular frequencies given by the sum and the differ- stator producing the rotating magnetic field is fitted outside the ence between the spin angular frequency (ω0) and the radio field glass envelope. angular frequency (ν): The mechanical speed of rotation of the rotor can be measured either in SI units in radians per second (rad/s) or in practical units • RF Field 1, rotating with frequency: ω0 + ν of revolutions per minute (rpm). The speed of rotation of rotor in • RF Field 2, rotating with frequency: ω0 − ν x-ray tubes can be usually varied between about 3000 rpm (nor- mal speed) and 9000 rpm (rapid speed). If an ensemble of spin is exposed to a magnetic field where its Related Articles: Rotating anode, Anode, Stator rotation period is short compared to the application time of the field, the time-averaged effect of the field will be almost zero RPA because any rotation due to the field will be cancelled out by a (Radiation Protection) See Radiation protection advisor (RPA) subsequent rotation when the field is in the opposite direction a short time later. Thus the sum frequency component (RF field 1) can be safely ignored. Radiation Protection of Patients (RPOP) website Note, however, that if the radio field angular frequency is (General) The RPOP is provided by the IAEA as a resource for matched to the spin angular frequency, the difference compo- health professionals, patients and the public on the safe and effec- nent (RF Field 2) will have zero angular frequency, and so in this tive use of radiation in medicine. It provides comprehensive infor- frame the only magnetic field is a static field in the transverse mation on both imaging and therapeutic methods that use ionising plane, producing precession about that axis and allowing 90° and radiation. It is divided into two major categories. The section for 180° pulses to be produced (Figure R.65). patients, caregivers, and the public provides information on what A frequency offset, that is a residual precession in the rotating to expect during medical procedures that involve ionising radia- frame, can be described by a hypothetical z-component yielding tion. The section for health professionals provides answers to fre- the effective B1: quently asked questions about different medical procedures and the safe use of ionising radiation in medicine. This provides a resource for researching specific radiation safety issues, guide- B1x (n - w0 ) = (B1x (n - w0 ) B1y (n - w0 ),(n - w0 ) / g) lines for managing risk in clinical procedures and materials that RPS 824 Rural hospital can be used in education and training activities for medical pro- programme. The rural centres should have the infrastructure fessionals. It is organised by each specific imaging or therapeutic needed to transport patients to urban hospitals when they need method and provides detailed guidance for conditions including more complex healthcare. patient and staff pregnancy. Rural health centre encompasses a variety of analogous terms Hyperlink: www .iaea .org /resources /rpop that describe an establishment of ambulatory primary health care attention, placed in a rural location and generally distant from RPS urban areas. Depending on the place and the complexity of the (Radiation Protection) See Radiation protection supervisor services offered, the staff may be comprised essentially of a med- (RPS) ical doctor, a nurse and a technician. Related Article: Rural hospital Rubidium-82 Further Readings: Borrás, C. ed. Defining the Medical (Nuclear Medicine) A radionuclide used for in vivo PET imaging. Imaging Requirements for a Rural Health Center. www .springer .com /gb /book /9789811016110; WHO 3rd Global Forum on Medical Devices, Geneva, Switzerland, May 2017, Workshop on Defining medical imaging requirements for rural health cen- Half-life Positron Maximum Photon Common ter – Part 1, www .w ho .in t /med ical_ devic es /gl obal_ forum /Defi fraction positron energy emission application ningr equir ement srura lheal thcen tre1. pdf ?u a =1; WHO 3rd Global 78 seconds 0.95 3.35 MeV 511 keV Cardiac Forum on Medical Devices, Geneva, Switzerland, May 2017, perfusion Workshop on Defining medical imaging requirements for rural imaging health center – Part 2, www .w ho .in t /med ical_ devic es /gl obal_ forum /Defi ningr equir ement srura lheal thcen tre2. pdf ?u a=1 82Rb is a potassium analogue, and the most widely used radio- Rural hospital nuclide for the assessment of myocardial perfusion with PET in (General) Rural hospitals are an integral part of a rural health cen- routine clinical practice. 82Rb is produced from an 82Sr / 82Rb gen- tre. They are established in distant (non-urban) locations mainly erator, and is used to distinguish normal from abnormal myocar- to provide primary health care attention and prevent patients from dium in patients suspected of a myocardial infarction. Since the R being forced to travel to distant urban medical facilities. Rural agent has a very short half-life, typical ‘rest-stress’ cardiac imag- hospitals are of key importance for community healthcare ser- ing can be performed within an hour. vices and health. Related Articles: Positron emission tomography, Radionuclide To manage patients properly, rural health centres should be imaging part of regional and more complete systems of medical health Further Readings: Cherry, Sorenson and Phelps. 2012. care installations in the country on the basis of a referral and Physics in Nuclear Medicine, 4th edn., Elsevier; Mettler and counter-referral programme. The rural centres should have the Guiberteau. 2012. Essentials of Nuclear Medicine Imaging, 6th infrastructure needed to transport patients to urban hospitals edn., Elsevier; Saraste, Kajander, Han, Nesterov and Knuuti. 2012. when they need more complex healthcare. PET: Is myocardial flow quantification a clinical reality? J. Nucl. Rural health centre encompasses a variety of analogous terms Cardiol.; Zeissman, O’Malley, Thrall and Fahey. 2014. Nuclear that describe an establishment of ambulatory primary health care Medicine, 4th edn., Elsevier. attention, placed in a rural location and generally distant from urban areas. Depending on the place and the complexity of the Rural centre services offered, the staff may be comprised essentially of a med- (General) Rural health centres are established to prevent patients ical doctor, a nurse, and a technician. from being forced to travel to distant urban medical facilities. To Related Article: Rural centre manage patients properly, rural health centres should be part of Further Reading: Borrás, C. ed. Defining the Medical regional and more complete systems of medical health care instal- Imaging Requirements for a Rural Health Center. www .springer lations in the country on the basis of a referral and counter-referral .com /gb /book /9789811016110 S SABR (Stereotactic ablative radiotherapy) by the government of a State as having legal authority for con-

(Radiotherapy) See Stereotactic ablative radiotherapy (SABR) ducting the regulatory process, including issuing authorisations, and thereby regulating nuclear, radiation, radioactive waste and Safelight filter transport safety’. (Diagnostic Radiology) See Darkroom Further Readings: WHO, Radiation safety culture in medicine, www .w ho .in t /ion izing _radi ation /medi cal _r adiat ion _e xposu re / Safety culture (Radiation safety culture) cu lture /en/; IRPA, Radiation Protection Culture, www .irpa (Radiation Protection) Safety culture is a set of regulations, .net /page .asp ?id =179; IAEA, Safety Culture Continuous values and activities directed towards risk management within Improvement Process (SCCIP), www .i aea .o rg /se rvice s /rev iew an organisation/institution/informal group. Safety culture may -m issio ns /sa fety- cultu re -co ntinu ous -i mprov ement -proc ess -s have various applications in different areas, depending on the ccip; IOMP, Radiation safety culture, www .iomp .org /radiation safety demands and regulations. Historically, the ‘safety cul- -safety -culture/ ture’ was introduced in the 1980s following the Chernobyl disaster. Safety in Radiation Oncology (SAFRON) system Safety culture in healthcare is a prerequisite for establishing (General) The Safety in Radiation Oncology (SAFRON) sys- well-functioning institutions and delivering reliable and risk-free tem was introduced by the IAEA in 2012. This is an integrated services. A strong safety culture contributes to risk and error voluntary reporting and learning system in radiotherapy and reduction for staff, patients and the general public, especially in radionuclide therapy incidents and near misses. The main goal complex environments such as DR/IR, NM and RT units. of SAFRON is to improve the safe planning and delivery of radio- Radiation safety culture in healthcare is a common interna- therapy and radionuclide therapy by sharing safety-related events tional initiative of the WHO, IRPA, IAEA and IOMP with the and safety analysis around the world. Information submitted is involvement of national and regional organisations. This initiative dependent on facilities registering and sharing incidents that is directed towards the establishment of a radiation safety culture occur in their institutions. S in healthcare facilities worldwide through the development of a SAFRON has over 50 registered medical facilities and hos- framework based on the existing regulations and the best profes- pitals all over the world. The system has over 1,300 incident sional practices. reports covering various types of incidents including errors and The initiative comprises a number of jointly organised near misses. SAFRON is available free through the IAEA’s workshops: RPOP website. SAFRON users have to be registered with IAEA’s NUCLEUS information resource portal. • 1st Workshop on Radiation Protection Culture in Related Articles: IAEA, RPOP website Medicine for Latin American Countries, Buenos Aires, Further Reading: Gilley, D., O. Holmberg and P. Dunscombe. Argentina, April 2015 2017. Improving quality and safety in radiotherapy using web- • 2nd European Workshop on Radiation Protection based learning. J. Med. Phys. Inter. 6(2):171–175. Culture in Medicine, Geneva, Switzerland, December Hyperlink: www .i aea .o rg /re sourc es /rp op /re sourc es /da tabas es 2015 -an d -lea rning -syst ems /s afron • 3rd African Workshop on Radiation Protection Culture in Health Care, Stellenbosch, South Africa, 3 November Sagittal plane 2016 (General) To describe anatomical planes imagine a person stand- • 4th Workshop on Radiation Protection Culture in ing in an upright position and dividing this person with imaginary Health Care for Middle East Countries, Doha, Qatar, vertical and horizontal planes. Anatomical planes can be used to 6–7 February 2017 describe a body part or an entire body. • 5th Workshop on Radiation Protection Culture in Picture a vertical plane that runs through the body from front Health Care for Asian and Pacific Countries, Putrajaya, to back or back to front. This plane divides the body into right and Malaysia, 8–10 November 2017 left regions. This plane will pass approximately through the sagit- • 6th Workshop on Radiation Protect Culture in tal suture of the skull, and hence any plane parallel to it is termed Healthcare, part of the Health Physics Society 52nd a sagittal plane. Midyear Meeting, San Diego, USA, 20–21 February See Anatomical body planes 2019 Samarium-153 Under the IAEA Basic Safety Standards, the regulatory authority (Nuclear Medicine) A beta-emitting radionuclide used for nuclear is defined as ‘an authority or a system of authorities designated medicine therapy. It is indicated for palliative relief of bone pain 825 Sample volume 826 S ample volume effects in well-counter detectors in patients with bone metastases. Since it also emits gammas, it duplex scanner. In conventional ultrasound scanners, the sample can be used for post-therapy imaging. volume is usually displayed as a pair of parallel lines along the beam (Figure S.1). The position in the image can be altered with a trackball or trackpad. This changes its position on the screen and Maximum Photon in turn allocates a specific depth and range from which Doppler Half-Life Beta Energy Emission Common Application sampling will be made. The indicated minimum length of the sample volume is usually around 1–1.5 mm, a practical limitation 1.95 days 640, 710, 103 keV Treatment of bone of the narrowband pulses used for Doppler. Larger sample vol- 810 keV metastases (beta); umes can be used; some systems offer sample volumes exceeding post-therapy imaging 15 mm. The sample volume is more complex than portrayed. No (gamma) information is shown about its lateral extent which is dependent on the beam width. There is no information displayed about its Related Articles: Radionuclides in therapy, Radionuclide dimensions in the elevation plane which is dependent on the slice imaging thickness and which varies depending on the depth. Control of Further Readings: Cherry, Sorenson and Phelps. 2012. the sample volume is important for certain applications. A small Physics in Nuclear Medicine, 4th edn., Elsevier; Zeissman, O’ sample volume allows the users greater precision as to where the Malley, Fahey and Thrall. 2014. Nuclear Medicine, 4th edn., sample volume is placed (Figure S.2). A large sample volume Elsevier; Mettler and Guiberteau. 2012. Essentials of Nuclear insonating the entire vessel is required if intensity-weighted mean Medicine Imaging, 6th edn., Elsevier. velocity is to be measured in order to obtain Doppler signals from the range of velocities in the vessel. The spectral display on Figure S.2 shows low velocities near Sample volume the wall. In the left image, the sample volume is in the centre of (Ultrasound) The sample volume is the volume from which the artery and low velocities are not detected or displayed. pulsed wave spectral Doppler is investigated in the image in a Related Article: Pulsed wave Doppler Sample volume effects in well-counter detectors (Nuclear Medicine) This article refers to changes in well-counter geometric efficiency when counting different sample volumes. The geometric efficiency changes with source position within the well as seen in Figure S.3. It is highest at the bottom of the test S tube and declines as the source approaches the well surface. Consider two cases: (1) a sample with the radioactive sub- stance diluted by water. The geometric efficiency in this situation decreases with increased sample volume. In case (2), a solution with constant activity concentration is poured into the test tube. In this case, the geometric efficiency initially increases linearly with sample volume but the proportionality is lost when the source vol- ume approaches the well surface. The sample volume significantly affects the geometric effi- ciency in the well counter, thus two samples should have the same FIGURE S.1 The axial length of the sample volume is displayed as two sample volume in a comparative study. One approach, used when parallel lines in the centre of the artery (see arrow). The size of the sample adequate sample volumes are available, is to fill the entire test volume is shown in the data fields (ringed) as 1.5 mm. tube because differences in total volume concentration will not FIGURE S.2 The two images show the effect of moving the sample volume near to the arterial wall (R). Sampling theorem 827 Saturation activity Spatially selective saturation bands, for example may be used to avoid flow artefacts. In this case a 90° pulse is applied to tis- sue just outside the field of view and then the imaging sequence is applied immediately after. Blood flowing from the saturation Activity band into the FOV has insufficient time to recover its longitudinal source magnetisation, and so produces no MR signal. Frequency-selective saturation bands operate slightly differ- Crystal ently: relying on chemical shift. In order to apply a saturation band either to fat or water, the chemical shift between the two must be sufficiently large that an RF pulse can be used to selec- tively excite one and not the other and the main magnetic field, FIGURE S.3 The geometric efficiency depends on the source ‘depth B0, must be sufficiently homogeneous. Thus frequency-selective in the detector well’. As the source approaches the top of the well the saturation bands are easier to implement at higher field strengths fraction of photons escaping through the well hole increases. At the well as the chemical shift is greater in absolute terms. The situation at surface the geometric efficiency is 0.5. 1.5 T is shown in Figure S.4. Practically, a fat saturation band might be implemented as follows: affect the geometric efficiency to the same extent as differences in partially filled tubes. Sample self-absorption and absorption in the test tube does not 1. A 90° pre-saturation pulse tuned to the resonant fre- greatly affect the count rate unless the γ-ray energy is low (e.g. for quency of fat is applied to the whole FOV, such that the 125I; 27–35 keV). To avoid any major sample volume effects, one bulk magnetisation vector of fat is flipped into the trans- should use identical test tube and sample volumes. verse plane. Related Article: Well-counter detector 2. A series of spoiler gradients are applied, to destroy the Further Reading: Cherry, S. R., J. A. Sorenson, and M. E. phase coherence of the fat signal. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, 3. The desired imaging sequence is applied immediately Philadelphia, PA, pp. 188–190. after the spoiler gradients, such that the fat protons have no time to recover their longitudinal magnetisation. The Sampling theorem resulting images show signal from water nuclei only (Magnetic Resonance) Sampling is the process by which a signal (Figures S.5 and S.6). is converted from a continuous space of time or space into dis- crete space, for example an array of numbers. The sampling theo- One limitation of the frequency-selective saturation band tech- rem states that if the maximal frequency in the continuous signal S nique is that uniform saturation throughout the field of view is f hertz (cps, cycles per second), then it can be completely deter- requires excellent magnetic field homogeneity. mined from a discrete signal sampled with a spacing of (2f)−1. Furthermore, due to the extra RF pulse/gradients applied dur- The sampling theorem is also known as the Nyquist–Shannon ing each slice, sequences with either type of saturation band have sampling theorem. increased SAR and increased acquisition time. Yet as saturation In MRI, the sampling theorem is important, for example when bands modify only the signal of the saturated tissue component reconstructing an image from sampled k-space data. The read- (they do not alter other contrast characteristics) they can be out gradient and the size of the object determine the maximal applied to virtually any pulse sequence. frequency. The sampling distance in the k-space, Δk, then deter- Further Reading: Westbrook, C., and C. Kaut. 1993. MRI in mines the FOV as FOV = 1/Δk. Foldover artefacts are due to vio- Practice, Blackwell Science, Hoboken, NJ. lation of the sampling theorem. Related Articles: Field of view, Nyquist criterion, Nyquist fre- quency, Nyquist theorem Saturation activity (Nuclear Medicine) The creation of a radioactive source is a SAR (scatter air ratio) result of competition between the addition of new radioactivity (Radiotherapy) See Scatter air ratio (SAR) SAR (specific absorption rate) Water (Magnetic Resonance) See Specific absorption rate (SAR) Saturation (Magnetic Resonance) For repetition times which are short rela- Pre-saturation pulse tive to the T1 of tissue there is little time for the net magnetisation vector to return to the z-direction between RF excitation pulses. Fat For this reason, rapidly successive excitations lead to large reduc- tions in longitudinal magnetisation and hence subsequently the MR signal. As

TR reduces eventually excitation leads to almost no signal: this is saturation. ‘Saturation bands’ are often used in clinical MR to reduce sig- Resonant frequency shift: nal (and thus reduce artefacts) from tissues adjacent to anatomy of 220 Hz interest. Saturation bands may be spatially selective or frequency selective. FIGURE S.4 Chemical shift between fat and water at 1.5 T. Saturation activity 828 Saturation activity Pre-saturation pulse Z: slice select Spoiler gradients X: phase Spoiler gradients Y: frequency FIGURE S.5 An example fat saturation gradient echo pulse sequence. 100% 80% 60% S 40% 20% 0% 0 2 4 6 8 10 Time (# of half-lives) FIGURE S.7 Saturation curve over the number of half-lives. (a) (b) FIGURE S.6 (a) A coronal T1-weighted image of the ankle with no satu- dN ration band applied. (b) A coronal corresponding slice with fat saturation I = 0 dt applied. N0 is the total number of particles impinging on the irradiated area ρ is the density by means of a given nuclear reaction and decay of the existing NA is Avogadro’s number radioactivity. The competition will lead to an equilibrium state, σ is the cross section of each individual atom in a target with establishing the saturation activity, when the rate of decay equals thickness x the rate of production. M is the atomic mass of the target material. (Please see article The process is described by an exponentially increasing curve entitled Activation formula thin target.) that will approach a constant value if the neutron flux or beam current is kept constant (Figure S.7). When the irradiation times goes to infinity or at least is large Consider the thin and thick activation formulas: compared to the half-life of the radioactive product, then the fac- tor (1 − eλt) approaches 1. This results in thin target and thick I × r × N × s × x target saturated activities of A(t) = A (1- elt ) and M × r × N × s × x Estart I I × r × N A(t) = A and A(t) = A (1 - e-lt ) ò s(E) ( dE M M ¶E / ¶l ) Estart Ethreshold I × r × N E A(t) = A E respectively where M ò s( ) (¶E / ¶l) d Ethreshold Saturation Saturation curve 829 Scanned beams Note that isotopes with shorter half-lives will achieve saturation rapid linear fall (conventionally). It is named a sawtooth based on earlier than those with longer half-lives. its similarity to the teeth on the blade of a saw. However, there are For a given radionuclide and after an irradiation time of one also sawtooth voltages in which the voltage ramps downward and half-life, 50% of the saturated activity will be achieved, two half- then sharply rises. The latter type of sawtooth voltage is called a lives of irradiation will give rise to 75% of the saturated activity, reverse (or inverse) sawtooth voltage. three half-lives of irradiation gives 87.5% and so on. The sawtooth wave is the form of the vertical and horizontal Related Articles: Activation formula thin target, Activation deflection signals used to generate a raster on CRT-based televi- formula thick target sion or monitor screens. Further Readings: Helus, F. 1983. Radionuclides Production, Related Article: Cathode ray tube Vol. 1, CRC Press, Boca Raton, FL, pp. 93–94; Krane, K. S. 1988. Introductory Nuclear Physics, 2nd edn., John Wiley & Sons, SBRT (Stereotactic body radiotherapy) Inc., Hoboken, NJ; Nordling, C., and J. Österman. 1999. Physics (Radiotherapy) See Stereotactic body radiotherapy (SBRT) Handbook for Science and Engineering, 6th edn., Studentlitteratur (Student literature), Lund, Sweden. Scan converter (Ultrasound) When displaying images the monitor may use a dif- Saturation curve ferent format than the data acquired during ultrasound scanning. (Radiation Protection) Saturation curve is the resulting graph A scan converter is a digital memory used to convert image scan when the increase of one variable no longer produces an increase data to that suitable for the monitor. of the dependent variable. Usually the saturation curve has the The line data will not normally fit in with the number of pix- shape of a hyperbola as shown in the Figure S.8. els in the memory. This is obvious for phased array and curvi- An example is the current–voltage characteristics of an ionisa- linear array image formats where the spacing between the lines tion chamber. increases with depth (Figure S.9). The pixels between the line data will be filled in by new values using interpolation between Saturation voltage the values in adjacent ‘line-data pixels’. (Radiation Protection) Saturation voltage is the voltage at which A number of other imaging processes are performed in the an observed effect does not increase further with increasing volt- scan converter memory, for example, averaging, freeze mode, age. For example, in a gas-filled ion chamber and in the semicon- read and write zoom and cine-loop. ductor diode detector the electric field should be high enough for assuring complete charge collection. The point at which there is Scan line no further increase in the collected charge is called the saturation (Ultrasound) The scan line refers to the direction of the beam voltage or saturation region. from which echoes are received and processed (Figure S.10). The Related Articles: Gas-filled radiation detectors, Germanium term can be used for A-mode, B-mode and colour flow imaging; S detectors, Ionisation chamber it tends not to be used for Doppler imaging where the term ‘ultra- Further Readings: Knoll, G. F. 2000. Radiation Detection sound beam’ is more commonly used. and Measurement, 3rd edn., John Wiley & Sons, Inc, New York, Related Articles: B-mode, A-mode, Pulse echo Chichester, Weinheim, Brisbane, Toronto, Singapore, pp. 129–155, 384; Lutz, G. 2007. Semiconductor Radiation Detectors, Device Scanned beams Physics. Springer – Verlag, Berlin, Heidelberg, pp. 162, 169, 172. (Radiotherapy) Scanned beams (as opposed to scattered beams) are used in pencil beam scanning (PBS) proton/ion therapy. A Sawtooth voltage thin pencil beam of protons/ions is delivered to the gantry from (General) The sawtooth voltage (or sawtooth wave) is a voltage with non-sinusoidal waveform showing a slow linear rise and 100 Scan converters 90 80 Beam direction dictated by construction and 70 performance of the Saturation point scanner 60 50 Scan matrix dictated by screen (e.g. 625 lines) 40 30 20 10 10 20 30 40 50 60 70 80 90 100 Beam direction Scan matrix Variable A FIGURE S.9 The scan converter changes the data obtained from the FIGURE S.8 A typical saturation curve. transducer and image processing into a format suitable for the screen. Variable B Scan projection radiograph (SPR) 830 Scatter air ratio (SAR) FIGURE S.11 Anterior–posterior (A–P) scan projection radiograph with planning lines showing the extent of the intended CT scan. FIGURE S.10 The arrows show the direction of scan lines for phased array (L) and linear array (R) transducers. such a situation a photon could scatter either in the patient or in some of the detector parts (collimator). The event is assumed to the accelerator via a series of magnets that direct the beam whilst have occurred somewhere along a perpendicular line to the detec- keeping the cross-section as small and circular as possible. tor face and the point of detection. This beam of protons/ions is then deflected by two sets of pow- Because of the degenerative nature of scatter, several erful magnets in the treatment head to scan the beam across the approaches to deal with it have been suggested, e.g. multiple treatment volume. energy window technique for SPECT and scatter estimation Abbreviations: PBS = Pencil beam scanning. derived from a transmission image in PET. Related Article: Pencil beam scanning (PBS) Further Reading: Cherry, S. R., J. A. Sorenson, and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Scan projection radiograph (SPR) Philadelphia, PA, p. 355. (Diagnostic Radiology) Scan projection radiograph (SPR) is the generic name used in CT scanning to describe the planning scan Scatter S used prior to the CT acquisition. It is more commonly referred to (Ultrasound) Scattering is the re-radiation of acoustic waves with by the trademark name used by particular CT manufactures, such properties different from those of the incident wave. Opposed to as Scanogram, Scout view or Topogram. the process of absorption, the energy is conserved in this process. The SPR is acquired with the tube stationary, and the couch Mathematically, scattering may be expressed as a boundary value moving through the gantry. The usual position of the tube is either problem, where the scattered wave field is obtained using wave 0° or 180° (A-P/P-A SPR) (Figure S.11), or 90° or 270° (lateral equations and matching boundary conditions at the surface of the SPR). The image obtained is similar in appearance to a conven- ‘scatterer’. A scatterer, or scattering object, is one where the den- tional radiograph. sity and/or the compressibility of the medium are different from The positions of the start and end of the volume to be scanned the surrounding medium. can be selected on the SPR view (Figure S.11). The positions of To solve for the scattered pressure field, there are two principal the slices to be reconstructed are also displayed. methods. The first is to assume that the homogeneous wave equa- On modern CT scanners the SPR is used to obtain patient tion is valid both inside and outside the scattering volume. The attenuation data to be used in automatic tube current modulation solution can be obtained by matching boundary conditions at the for dose optimisation. surface of the scatterer. In this case the geometry of the object Related Article: Automatic tube current modulation must be known, and for instance, in this way the scattering from a sphere can be solved exactly. Scatter For more complex problems, such as for scattering from tis- (Nuclear Medicine) Scattering refers to photons that undergo sues, a so-called Green’s function approach is employed. Here, one or more Compton interactions prior to detection. Scattered a source term is added to the homogeneous wave equation. This photons add an uneven distribution of counts to the final image. leads to an integral equation in which the Green’s function is These counts degenerate contrast and make the images unsuit- integrated over a volume including the inhomogeneities. The able for activity quantification. The fraction of scattered events Green’s function can be seen as the wave arising from an ele- can in some cases (e.g. abdominal imaging) be as high as mentary source. Basically, the terms of the integral equation then 60%–70%. describes the effect of the source (transmitter) and the effect of The probability for photon interactions (photoelectric, the scattering objects. Compton and pair production) is proportional to the distance travelled in an absorbing medium. It also depends on the photon Scatter air ratio (SAR) energy and absorbing medium characteristics (effective atomic (Radiotherapy) The scatter air ratio (SAR) separates out the pri- number, density, etc.). When a photon undergoes a Compton inter- mary component of TAR from the total TAR to get the scatter action there is a change in direction. contribution. It was defined by ICRU 24 as: Consider a source inside a patient at typical organ depth (∼10 The ratio of that part of the TAR that is due to radiation scat- cm) while acquiring an image using a scintillation camera. In tered within the phantom to that part of the TAR that is due to Scatter coincidences 831 Scatter correction primary radiation alone. The primary component cannot be mea- sured directly and is usually obtained by extrapolating to zero Line of field size. response And can be described as in Equation S.1, the calculation of Detector -scatter air ratio from tissue air ratios ring Point of SAR(d, A,E) = TAR(d, A,E) - TAR(d,0,E) (S.1) annihilation TAR(d, 0, E) is known as the zero field tissue air ratio, which Registered represents the attenuation of the primary beam in tissue with photon no scatter contribution. It can be measured using a column of water in narrow beam geometry, but more often is derived FIGURE S.12 A photon is scattered in an absorbing medium before reg- from extrapolation of the TAR values for finite field sizes (see istration in an opposite detector. The photon path is represented by the Equation S.2) black lines and the LOR by the dotted line. The calculation of zero field TAR:

TAR(d,0,E) = exp-meff (z-zmax ) (S.2) Scatter correction (Diagnostic Radiology) Scatter causes severe distortion and where μeff is the effective attenuation coefficient of the photon contrast loss in the acquired/reconstructed image. In diagnostic beam energy E. imaging, scatter corrections can be applied to various techniques Abbreviations: TAR = Tissue air ratio and SAR = Scatter air and applications in order to correct for dose or improve the image ratio. quality. It has been used across many modalities including x-ray, Related Articles: Peak scatter factor (PSF), Percentage depth breast imaging and CT. dose (PDD) Physical techniques have existed to reduce the scatter contri- bution in imaging for years but over the last few decades there Scatter coincidences has been a drive do develop software-based methods to provide (Nuclear Medicine) A scatter coincidence is a PET event type a scatter correction in imaging. As a result, there are now vari- where one or both of the photons are scattered in an attenuating ous correction methods including software, hardware and hybrid material before being registered. If both the photons are regis- approaches. This article broadly discusses some of the methods in tered, this will lead to a misplaced line of response (LOR), as development or used in practice. seen in Figure S.12. The degree of misplacement depends on the Physical Techniques: Anti-scatter grids are an example of scattering angle and the location of the scatter event. Scattered hardware solutions to reduce scatter. These grids reduce the SPR. S photons add to the background noise and will degenerate contrast (See Grid Bucky.) and cause a source of error for quantitative measures of activity. Having a large air gap between the point of scatter and the The fraction of scattered events can in some cases (e.g. abdominal detector is a geometric technique that can be used to reduce the imaging) be as high as 60%–70%. There are three reasons that contribution of scatter in the image. can explain this high fraction. The first reason is that only one of Analytical/Software Methods: MonteCarlo (MC) calculation the two scattering photons must scatter to produce a misplaced can approximate the scatter in a projection and then be applied line of response (LOR). to the reconstructed image. The scatter estimated from previous The second reason is that Compton scatter is the dominating reconstructions can be fed back for scatter compensation on the interaction in scintillators at 511 keV and some incident pho- projected data (Ruehrnschopf and Klingenbeck, 2011). tons might only deposit a small amount of energy. As an exam- Kernel methods deconstruct a measured projection into the ple; imagine an event where one of the annihilation photons scatter and primary components and measure this against an is registered in a detector and the second photon is Compton experiment-calibrated empirical relationship between them (Xu scattered in the opposite detector crystal, thus depositing less et al., 2015). than 511 keV, before escaping the crystal. Such an event would Primary modulation method was developed to extract the low- represent a true coincidence and a true LOR, but if the energy frequency signal from the Fourier Domain (Xu et al., 2015). In deposited does not exceed the lower limit of a potential energy this method, a high-frequency attenuation sheet (primary modu- discrimination window it will be ignored. As a result the pulse lator) is placed between the scanned object and the x-ray source. height analysis window must be widened to include these This modulator then applies a high-frequency modulation onto coincidences. the primary signal. This leads to a strong separation between the The final reason is the low-energy resolution in the LSO and primary and scattered distribution in the Fourier domain (Zhu et BGO scintillators because of their low light output. The latter two al., 2006), and with this setup, it is possible to remove the scat- reasons justify the use of a wide pulse height analyser window ter frequencies from the signal in Fourier space. Processing tech- which at the same time provides an inadequate protection against niques of linear fitting and demodulation are implemented after to scatter events. The energy discrimination window must be opti- effectively correct for the scatter (Gao et al., 2010). mised in order to provide protection to scattered events but at the This method has a key hypothesis that low-frequency compo- same time have a high sensitivity (Figure S.12). nents dominate the scatter distributions even if high-frequency Related Articles: Scatter correction in PET: Line of response, components are present in the incident x-ray intensity distribu- Annihilation coincidence detection tion, meaning that the scatter component does not include high- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. frequency components (Zhu et al., 2006). Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Further Readings: Gao, H. et al. 2010. Scatter correction Philadelphia, PA, p. 342. method for X-ray CT using primary modulation: Phantom Studies. Scatter correction in PET 832 Scatter factor Inter. J. Med. Phys. Res. Practice 37(2):934–946; Ruehrnschopf, reconstruction. This method also accounts for activity outside the E. P. and K. Klingenbeck. 2011. A general framework and review FOV. However, when scatter distribution is complex or when the of scatter correction methods in x-ray cone-beam computerized object covers the entire FOV, this method can cause significant tomography. Part 1: Scatter approaches. Intern. J. Med. Phys. errors (e.g. 5% for brain imaging and more than 10% at the heart Res. Practice 38:4296–4311; Xu, Y. et al. 2015. A practical and lung interface). cone-beam CT scatter correction method with optimized Monte Related Articles: PET, Compton scatter, Annihilation Carlo Simulations for Image-Guided Radiotherapy. Phys. Med. Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Biol. 60(9):3567–3587; Zhu, L., N. Bennett, and R. Fahrig. 2006. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Scatter correction method for X-ray CT using primary modula- Philadelphia, PA, p. 355. tion: theory and preliminary results. IEEE Trans. Med. Imaging 25(12):1573–1587. Scatter correction in SPECT (Nuclear Medicine) In an idealised model all events registered in a Scatter correction in PET projection element originate along the line of response, that is true (Nuclear Medicine) Scattered photons add to the background events. In practice this assumption is most untrue. Scattered events noise and will degrade contrast and cause a source of error for occur when a photon is Compton scattered inside the patient before quantitative measures of activity. The fraction of scattered events being registered in the detector. In some typical patient studies, the can in some cases (e.g. abdominal imaging) be as high as 60%– ratio between scattered events and true events could be as large as 70%. There are three reasons that can explain this high fraction. 40%. In such cases it is important to use a proper scatter correction The first reason is that only one of the two scattering photons must in order to attain a good image contrast and a correct relationship scatter to produce a misplaced line of response (LOR). between reconstructed image intensity and source activity. The second reason is that Compton scatter is the dominating A common approach to correct for scattered photons has been interaction in scintillators at 511 keV and some incident photons to collect projection profiles in a scatter window simultaneously might only deposit a small amount of energy. As an example, with the photopeak window where the scatter window is placed imagine an event where one of the annihilation photons is reg- below (in terms of energy) the photopeak window. The scatter istered in a detector and the second photon is Compton scattered projection profiles are multiplied with a weighting factor before in the opposite detector crystal, thus depositing less than 511 keV, subtracting them from the photopeak profiles. Which weighting before escaping the crystal. Such an event would represent a true factor to use depends on the size of the source, the energy window coincidence and a true LOR, but if the energy deposited does not settings and the scintillation camera energy resolution, and this exceed the lower limit of a potential energy discrimination win- has to be determined experimentally. A limitation to this correc- dow it will be ignored. As a result the pulse height analysis win- tion method is the fact that the scattered photons registered in the dow must be widened to include these coincidences. lower scatter window are more likely to have multiple scatter and S The final reason is the low-energy resolution in the LSO and as a consequence of this the spatial distribution of the events in BGO scintillators because of their low light output. The latter two the scatter window may differ from the scatter in the photopeak reasons justify the use of a wide pulse height analyser window window. which at the same time provides an inadequate protection against A more useful method is based on two narrow energy win- scatter events. The energy discrimination window must be opti- dows placed adjacent to the photo peak window. The scatter in mised in order to provide protection from scattered events but at the photo peak energy window is estimated by the average of the the same time have a high sensitivity. acquired images in each of the narrow energy windows after cor- It is impossible to separate scattered events that originate from recting for differences in window width. the body from the ones scattered inside the crystal, hence simple If the attenuation correction is performed prior to the scatter correction schemes such as dual-energy windows are not as suc- correction, then all events including scattered events will be cessful for PET as for SPECT. There are two main approaches corrected for. It is therefore important that scatter correction today for scatter correction. precedes the attenuation correction. Alternatively, one needs to In the first approach the correction is derived from the original reduce the magnitude of the attenuation coefficient to compensate scatter-contaminated image and a transmission image (represent- for the extra counts in the image created by the scatter. The actual ing the attenuation coefficient for different tissues). The transmis- value of such an effective attenuation coefficient depends on the sion image is acquired using a transmission source or data from patient size, source location and tissue composition. a CT scan. With data from these two images and some simple Iterative reconstruction methods sometimes may use the assumptions, a computer model can derive an estimation of the transmission images from the attenuation correction (see separate scatter distribution for each projection. The contribution of this articles on Attenuation correction in SPECT) to analytically scatter is then subtracted from each projection and the recon- calculate the probability of scatter from a location in the patient struction is performed with scatter-corrected data. This method being detected in a specific detector element. works well when the activation is contained inside the field of view Related Article: Attenuation correction in SPECT (3 articles) (FOV). A source outside the FOV can lower the scatter correction Further Reading: Cherry, S. R., J. A. Sorenson and M. E. accuracy. The method is computationally intensive. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, The second approach involves examination of projection pro- Philadelphia, PA, pp. 315–317. files just outside the object. After correcting for random coinci- dences it is assumed that all counts in such a profile are scatter Scatter factor induced. Scatter is assumed to be a low-frequency phenomenon (Radiotherapy) The scatter factor SF for a field A is defined as with little structure and the scatter distribution can be extrapo- the ratio lated using a cosine or Gaussian function so that the scatter esti- mation covers the entire projection. The scatter distribution is PSF(A,E) SF(A,E) = then subtracted from each individual projection prior to image PSF(10,E) Scatter phantom ratio (SPR) 833 Scatterer Hence it is the peak scatter factor normalised to 1 for a S 10 × 10 cm2 field, and quantifies the variation of scatter con- SPR = (S.4) P tribution with field size to the dose in tissue at the reference depth, normalised to the reference field. Note that SF(10,E) = 1 S = Scattered Photons, P = Primary Photons by definition. If the SPR = 1 then this means that there are just as many scat- This is also known as the phantom scatter factor, Sp, or the tered photons detected as there are

primary photons. An SPR of normalised peak scatter factor, NPSF. 3 means that there are three times more scattered photons then Related Article: Peak scatter factor (PSF) there are primary photons, meaning that 75% of the photons detected by the detector carry little or no useful information con- Scatter phantom ratio (SPR) cerning the anatomy. (Radiotherapy) The scatter phantom ratio (SPR) bears the same The SPR increases as the volume of tissue irradiated by the relationship to tissue phantom ratio (TPR) as the scatter air ratio x-ray beam increases. It also increases as the field size increases (SAR) does to the tissue air ratio (TAR). It separates out the pri- (Bushberg, 2012). Scatter grids can be used to reduce the scatter mary component of TPR from the total TPR to get the scatter in diagnostic imaging and hence reduce the SPR. contribution. Primary radiation retains the information regarding the atten- The scatter phantom ratio (SPR) can be described as in uation characteristics of anatomy in contention and delivers max- Equation S.3. The calculation of scatter phantom ratio from tissue imum subject contrast. Scattered radiation degrades the subject phantom ratios: contrast. (Bushberg, 2012) Further Readings: Bushberg, J. T., Seibert, E. M. Leidholdt SPR(d, A,E) = TPR(d, A,E) - TPR(d,0,E) (S.3) Jr. and J. M. Boone. 2012. The Essential Physics of Medical Imaging, 3rd edn. Lippincott Williams & Wilkins, Philadelphia; See Related Articles for more detail. Health Physics Society, 2016. HPS Specialists in Radiation Abbreviations: TPR = Tissue phantom ratio, SAR = Scatter Protection. [Online] Available at: https :/ /hp s .org /publ icinf ormat air ratio and TAR = Tissue air ratio. ion /a te /q1 1028. html (Access date 16 July 2019). Related Articles: Tissue phantom ratio (TPR), Scatter air ratio Hyperlink: Health Physics Society – https :/ /hp s .org /publ icinf (SAR), Tissue air ratio (TAR) ormat ion /a te /q1 1028. html Scatter subtraction (Nuclear Medicine) Scatter in a nuclear medicine image comes Scattered radiation from the fact that the energy resolution of NaI(T l) crystal (Diagnostic Radiology) Scattered radiation refers to all pho- generally is relatively poor and in order of 10% for 99m-Tc tons and charged particles resultant from scattering interaction photons (140 keV). This means that in order to maintain rea- between an incident photon or particle of ionising radiation and S sonably large counting statistics, a large energy window (dis- the medium. criminator) is needed. Because a photon in a scattering event Scattered radiation may travel in any direction away from can lose only a small fraction of the energy and still change the site of interaction. Therefore further ionisations and energy the direction significantly, there will be a chance that these deposition in the medium may occur outside the primary radia- photons will be registered within the energy window. These tion beam. Interactions between scattered radiation and the unwanted events will add to the total number of true events and medium are important for radiation protection purposes – they will thereby reduce the image contrast because they appear to are important in radiotherapy treatment planning when attempt- be coming from another location than the decay location. It is ing to deliver the prescribed dose to the target volume (tumour) therefore desirable to correct for this scatter contribution. A while minimising the peripheral dose to surrounding healthy tis- scatter correction can be made by subtracting an estimate of sues due to such scattered radiation. the scatter in the main energy window by an estimate obtained Scattered radiation may also exit the medium altogether. Such either from an analytical procedure or from a measurement in transmitted scattered radiation is important in diagnostic radiology a second energy window. The drawback with scatter subtrac- where it may be detected by the imaging receptor, contributing to tion is that if the main data is noisy and the scatter is noisy the ‘fogging’ of the x-ray image. Scattered radiation is minimised then the noise in the result from a subtraction will be even in diagnostic radiology by the introduction of anti-scatter bucky larger. Furthermore, in the case of inaccurate estimates and grids between the patient and the image receptor. noise, a negative number can appear in the result that needs to Related Articles: Secondary radiation, Secondary electron, be accounted for. Peripheral dose, Bucky grids Scatter to primary ratio (SPR) Scatterer (Diagnostic Radiology) Scatter radiation detected by the detector (Ultrasound) Scattering is the re-radiation of acoustic waves with can significantly degrade the image quality. For diagnostic imag- properties different from those of the incident wave. As opposed ing, scattered photons are one of the major contributors of noise to the process of absorption, the energy is conserved in this pro- and if there is a significant amount of noise in the image then the cess. Mathematically, scattering may be expressed as a boundary signal-to-noise ratio will reduce, leading to a reduction in image value problem, where the scattered wave field is obtained using quality. wave equations and matching boundary conditions at the surface One parameter used to characterise the amount of scatter of the ‘scatterer’. A scatterer, or scattering object, is one where detected in the image is using the scatter to primary ratio (SPR). the density and/or the compressibility of the medium are differ- The SPR is a ratio between the number of scattered photons ent from the surrounding medium. Scattering back to the trans- there are for every primary photon. Equation S.4 describes this ducer plays a large role in the formation of B-mode and Doppler relationship (Bushberg, 2012). images. Scattering, classic (coherent) 834 Scintillation camera To solve for the scattered pressure field, there are two principal methods. The first is to assume that the homogeneous wave equa- tion is valid both inside and outside the scattering volume. The solution can be obtained by matching boundary conditions at the surface of the scatterer. In this case the geometry of the object must be known, and for instance, in this way the scattering from a sphere can be solved exactly. For more complex problems, such as for scattering from tis- sues, a so-called Green’s function approach is employed. Here, a source term is added to the homogeneous wave equation. This leads to an integral equation in which the Green’s function is inte- grated over a volume including the inhomogeneities. The Green’s function can be seen as the wave arising from an elementary source. Basically, the terms of the integral equation then describe the effect of the source (transmitter) and the effect of the scatter- FIGURE S.13 Schlieren image of the field from a single crystal ultra- sound transducer. ing objects. Related Article: Green’s function will generate a variable pressure field which gives rise to vari- Scattering, classic (coherent) able refractive indexes in the field. Diffraction of light passing (Radiation Protection) See Coherent scattering perpendicular through the field can then be used to visualise the ultrasonic field and to calculate the spatial pressure distribution Scattering, coherent or Rayleigh (Figure S.13). (Radiation Protection) See Coherent scattering Scintillation camera Scattering cross section (Nuclear Medicine) The scintillation camera is the most common (Ultrasound) The scattering cross section is defined as the time- imaging system used within the field of radionuclide imaging. It is averaged total scattered power divided by the time-averaged inci- also referred to as the Anger scintillation camera which is named dent intensity. The unit is in square metres, as implied by the name after the inventor, Hal Anger. The basic principle of the scintilla- cross section. Physically this cross section corresponds to the area tion camera is to depict the distribution of administered radioac- of the incident wave that contains the amount of power that is tive substances by ‘collecting’ emitted radiation. scattered by the scattering object. This means that the scatter- The localisation and bio-distribution of the activity pro- S ing cross section divided by the geometrical cross section of the vide functional information about organ systems and individual object is a measure of how effectively the object scatters sound. organs, that is a high regional iodine uptake in the thyroid gland Related Articles: Absorption cross section, Extinction cross could sometimes be an indication of a pathological condition. section, Differential scattering cross section The four major components in a scintillation camera are a collimator, a NaI (T l) crystal (typically), a light guide and an Scattering, incoherent array of PM tubes. The collimator, which is a lead plate with (Radiation Protection) See Incoherent scattering holes, is a necessity in order to gain spatial resolution in the images. Photons with an oblique angle of incidence will be Scattering, Rayleigh attenuated in the collimator whereas photons with a near paral- (Radiation Protection) See Rayleigh scattering lel direction will pass through the detector and reach the crystal. Photons reaching the crystal will interact and through a number of processes create light photons. The number of light photons Scattering, Thomson is proportional to the energy deposited in the interaction. The (Radiation Protection) See Thomson scattering crystal must be a high attenuating material to register most of the incident photons. Choosing the right crystal with highest Scene-based registration performance is an optimisation between a number of parame- (Nuclear Medicine) Scene-based registration, or projection-based ters, that is density, atomic number, light yield, decay time, sen- registration, aims to register calculated projections from one 3D sitivity to mechanical stress, peak emission, etc. Light photons data set to calculated projection from another 3D data set. These are transported through a light guide to the array of photomulti- methods can also apply to volume-to-image registration where plier tubes (PM tubes). The use of an array of PM tubes, instead projections of a 3D SPECT volume are registered to a planar scin- of single PM tube, makes it possible to localise the event using tillation camera image. By projecting the images onto the one- position logic circuitry. dimensional space the computational complexity is significantly The scintillation camera has a wide clinical use, for example reduced. whole body scans to localise cancer or metastases in the body and bone scans of the skeleton. Another use is for dynamic studies Schlieren where a number of images are acquired during an examination. (Ultrasound) Optical inhomogeneities in transparent material not Gated acquisition is used to examine different phases of the car- visible to the human eye are named schlieren. These inhomoge- diac cycle. neities can be visualised by the diffraction of visible light and can Related Article: SPECT easily be seen in air above a lit candle or a hot surface. Further Readings: Cherry, S. R., J. A. Sorenson and M. E. Schlieren can be used to visualise the characteristics of a field Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, from an operating ultrasonic transducer. The ultrasonic field Philadelphia, PA, pp. 211–213, 223–224; Anger, H. O. 1967. Scintillation detector 835 S cintillator Radioisotope cameras. In G. J. Hine (ed.), Instrumentation in Solid state scintillation detectors are widely used for detection Nuclear Medicine, Academic Press, New York, Chapter 19. and quantification of gamma emitters in nuclear medicine, e.g. in gamma cameras, SPECT, and PET machines. Liquid scintillation Scintillation detector counters are used to measure low-energy β− and α-particles. (Radiation Protection) The scintillation detector consists of a Scintillation detectors are very important in radiation protec- scintillator and a photomultiplier (PM) tube supplied with an tion, as their sensitivity is much higher than that of ionisation external high voltage (Figure S.14). The scintillator, e.g. NaI(Tl), chambers and Geiger counters. They lack the ability to identify mounted in special container with reflective powder (MgO), is in the type of radionuclide being detected, but they are ideal to optical contact with the PM tube. When an x-ray or gamma pho- quickly locate a lost gamma source or demarcate a contaminated ton interacts with the crystal, its absorption results in the emission area. of optical radiation (with a maximum wavelength of about 420 Abbreviations: MgO = Magnesium oxide, NaI(Tl) = Sodium nm) in all directions. The reflective powder protects the optical iodide with thalium impurities, PET = Positron emission tomogra- radiation photons from escaping the crystal. The light (the optical phy and SPECT = Single photon emission computed tomography. radiation photons) interacts with the photocathode by the photo- Related Articles: Gamma camera, Liquid scintillation, electric effect and its energy is transferred to the electrons.

The Photomultiplier, Pulse height analyser, Scintillator quantum efficiency, QE, of this process is defined as Further Readings: Brown, B. H. et al. 1999. Medical Physics and Biomedical Engineering, Institute of Physics Publishing, (Number of photoelectronsemitted) Bristol, UK, pp. 153–156; Dendy, P. P. and B. Heton. 1999. Physics QE = (Number of incident light photons) for Diagnostic Radiology, Institute of Physics Publishing, Bristol, UK, pp. 147–149; Graham, D. T. and P. Cloke. 2003. Principles of and its value is 20%–30%. The PM tube multiplies the number Radiological Physics, Elsevier Science Ltd., Edinburgh, London, of electrons thanks to the secondary electron emission from the UK, pp. 379–383; Knoll, G. F. 2000. Radiation Detection and system of electrodes (dynodes). The dynodes are held at the posi- Measurement, 3rd edn., John Wiley & Sons, Inc, New York, pp. tive potential (the first of several hundred volts) which accelerates 219–231; Saha, G. B. 2001. Physics and Radiobiology of Nuclear electrons. The multiple stage multiplication for 10 stages (dyn- Medicine. 2nd edn., Springer-Verlag, New York, pp. 89–91. odes) is about 107. The pulse height (amplitude) depends on the energy of the Scintillator gamma radiation. The measured spectrum can be analysed using (Nuclear Medicine) Radiation interacts with surrounding atoms a pulse height analyser (PHA), e.g. a single-channel analyser and molecules by causing ionisation and excitation. As the atoms (SCA). Figure S.15 shows an example of a simple gamma radia- and molecules undergo recombination and de-excitation, energy tion spectrum. is released, mostly as thermal energy. In a few materials, this energy is released as a number of low energetic photons (visible, S UV). Material with this ability is called a scintillator. These scin- tillation materials are used in scintillation detectors. HV Scintillation material can be divided into two groups: inor- R ganic scintillators and organic scintillators. Inorganic materials are solid crystals and the basic condition C for scintillation comes from characteristics in their lattice struc- K ture. Individual atoms or molecules do not scintillate. In organic materials on the other hand the scintillation is a molecular prop- Pulse erty rather than an effect of crystal structure. PM The general principle for all scintillators is that the amount of low energetic photons produced is proportional to the amount of FIGURE S.14 Schematic view of a scintillation detector. C, scintillator energy deposited during the interaction. (crystal); R, reflective powder (MgO); PM, photomultiplier tube; K, pho- Inorganic Scintillators: Many of the inorganic scintillators tocathode; HV, high voltage supply. are impurity activated which means that an impurity, consisting of a small amount of atoms from a different element is added to the crystal structure. Sodium iodine (NaI (Tl)) (1948 Hofstadter) and caesium iodine (CsI (Tl)) are two examples of crystals with a thallium impurity. Most commonly used in SPECT is the NaI (Tl) scintillator. Photopeak NaI (Tl) was first used for both SPECT and PET. Since NaI Noise (Tl) has a long decay time and a low effective atomic number, it Compton is unsuitable for PET imaging with its fast count rates and high- scattering FWHM energy annihilation photons. Over time a number of new scin- tillators have emerged, i.e. BGO (1975 Nestor and Hung), GSO (Takagi and Fukazawa) and LSO (1990 Melcher). The individual properties of these scintillators are displayed in Table S.1. The most common scintillator used in PET today is LSO due to its Height of pulse high effective atomic number which allows it to effectively stop high-energy radiation. FIGURE S.15 Example of gamma radiation spectrum obtained with Photon yield describes the number of photons produced per NaI (Tl). unit (keV) of deposited energy. Number of pulses Scintillator 836 S cintillator TABLE S.1 Properties of Different Scintillation Materials Scintillator Material Properties Density (g/ Effective Atomic Decay Photon Index of Peak Emission cm3) Number Time (ns) Yield Refraction Hygroscopic (nm) NaI (Tl) 3.67 50 230 38 1.85 Yes 415 BGO 7.13 74 300 8 2.15 No 480 LSO (Ce) 7.40 66 40 20–30 1.82 No 420 GSO (Ce) 6.71 59 60 12–15 1.85 No 430 The emitted low-energy photons must pass through the crys- Ia I tal and into the photomultiplier tube where the low-energy pho- e tons will strike a photocathode. The number of photons passing or being reflected at this connection depends on the refraction index of the two materials. Similar refraction indices allow more photons to pass without being reflected back (where the reflected photon might be reabsorbed by the crystal). Another important property, for both manufacturer and user, Ea E Ee E is that the scintillator is not sensitive to moisture and/or thermal Ee < Ea and mechanical stress. NaI (Tl), for example is very fragile and (a) (b) also hygroscopic. The electron yield at the photocathode is a function of the FIGURE S.16 (a) Spectrum of optical absorption. Ia, intensity of energy (wavelength) of the incident photons. For many photocath- absorbed radiation; E, energy of radiation; Ea, energy of maximum odes, the yield peaks at ∼400 nm. It is a desirable feature for a absorption intensity. (b) Spectrum of optical emission. Ie, intensity of scintillator to have its peak emission in the vicinity of 400 nm. emitted radiation; E, energy of radiation; Ee, energy of maximum emis- Organic Scintillators: The scintillators in this group are sion intensity. S in most cases liquid solutions, although there are some plastic organic scintillators. As previously mentioned the scintillation process of an organic scintillator is a property of individual mol- ecules or atoms and not a lattice effect. Conduction band The radioactive sample is placed in a solvent containing scin- Activator excited state tillator material. Electrons freed in an ionising interaction will transfer energy to scintillation molecules which will emit low- Eg Scintillation photon energy photons that are detected by one or more photomultiplier tubes. Activator ground state Related Articles: PET, Photocathode, Photomultiplier tube, Valence band Photon yield, SPECT Further Readings: Cherry, S. R., J. A. Sorenson and M. E. FIGURE S.17 Scheme of a scintillation photon emission. Eg, energy gap. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA, pp. 100–108; Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., Inorganic scintillators may be produced with different size New York. and shape and are used as gamma radiation detectors in a gamma camera for medical imaging. Scintillator Examples of inorganic scintillators are: NaI (Tl), CsI (Tl), (Radiation Protection) A scintillator is a substance in which mol- BGO, CdWO4 for photon detection and lithium and boron glasses ecules, absorbing the energy Ea of photons (gamma or x-ray) or as neutron detectors. A novel scintillator is Lu(Y)AP, a crystal charged particles and neutrons, emit ultraviolet or visible light with better image resolution and reduced scan time and random photons of energy Ee < Ea (Figure S.16) with high efficiency. The noise. Its main application is expected in the next generation of emission is performed as a prompt fluorescence process with a PET machines, especially for pediatric care, cancer pathologies, decay time of a few nanoseconds. heart and coronary artery diseases, neurological diseases. There are two kinds of scintillators: inorganic and organic. Organic Scintillators: There are pure organic crystals (e.g. Inorganic Scintillators: A crystal of NaI (sodium iodide) acti- anthracen) and liquid scintillators composed of a scintillation vated with Tl (thalium), in quantities of about 10−3 mole fraction, solute (primary solute) such as e.g. PPO, a solvent, e.g. toluene to enhance the light output (photon emission) of visible radiation or xylene, and often a secondary solute e.g. POPOP, called the (415 nm), is an example of inorganic scintillators, in which the wavelength shifter. Plastic scintillators may be shaped as fibres light efficiency is proportional to the radiation energy deposited (Figure S.18), ribbons and thin films. in crystal. Figure S.17 shows a scheme of the role of the activator Abbreviations: BGO Bi4Ge3O12 = Bismuth germinate, in an emission of a scintillation photon. PET = Positron emission tomography, POPOP = Compound Scoring 837 Screen speed Cladding Screen-film radiographic receptor Intensifying screen Core Film FIGURE S.18 Scheme of a plastic scintillation optical fibre (refractive index of core > refractive index of cladding). X-ray Light 1,4-bis-2-(5-phenyloxazolyl)-benzene and PPO = Compound 2,5-diphenyloxazole. FIGURE S.19 Screen film combination and effect. (Courtesy of Sprawls Foundation, www .sprawls .org) Related Articles: Gamma camera, Liquid scintillation (LS) counting, Scintillation detector Further Readings: Knoll, G. F. 2000. Radiation Detection screens. The function is illustrated in Figure S.19. The function of and Measurement, 3rd edn., John Wiley & Sons, Inc, New the intensifying screen is to absorb the x-ray photons and convert York, pp. 219–240; Hobbir, R. K. 1997. Intermediate Physics for them to light. Because of its composition and thickness the screen Medicine and Biology, 3rd edn., Springer-Verlag, New York, pp. is an effective x-ray absorber. The film is more sensitive to light 422–423. than to x-ray so the exposure to the film is ‘intensified’ by this process. Exposing the x-ray film both to the light from the screen Scoring and the x-rays from the beam increases the radiographic effect of (Nuclear Medicine) Semi-quantitative visual analysis was devel- the method, hence decreases the dose necessary for production of oped at Cedars-Sinai Medical Center for systematic and repro- radiographic contrast. ducible interpretation of the myocardial perfusion images. Their When two intensifying screens are used (with a double emul- approach uses three sets of slices from short axis views corre- sion x-ray film between these) the effect of crossover exposure sponding to the distal, mid-ventricular, and basal regions of the (or print-through exposure) is seen. This effect is due to the fact left ventricle as well as the apex visualised in a mid-ventricular that the light from one screen is not fully absorbed by the adja- long-axis image. A visual scoring system, based on either 17 or cent emulsion of the film. Part of this light passes through this 20 segments is usually used. emulsion (and the transparent film base) and exposes the other The distal, mid-ventricular, and basal regions of the myocar- emulsion as well. This leads to decreased resolution and unsharp dium are divided into six segments each, while the apex is repre- radiograph. This effect is about 30% when calcium tungstate S sented by two segments. Each of the segments is scored according screen is used and decreases to about 20% when yttrium tantalate to a 5-point scheme with the following definition: screen is used. Until the 1970s calcium tungstate was the typical fluores- • 0 is the normal cent material in radiographic intensifying screens. After this it • 1 is the slight reduction of uptake or equivocal had been replaced by a variety of materials/compounds with a • 2 is the moderate reduction of uptake, usually implies a rare earth element (such as lanthanum, gadolinium, and ytter- significant abnormality bium). Various screens produce different spectrums of light. • 3 is the severe reduction of uptake This requires the x-ray film to be sensitive to this particular light • 4 is the absence of tracer uptake spectrum. Usually there are specific combinations of screens and films. Related Article: Bullseye image Further Reading: Curry, T. S., J. E. Dowdey and R. C. Murry. Further Readings: Berman, D. S., A. Abidov, X. Kang, S. 1990. Christensen’s Physics of Diagnostic Radiology, Lea and W. Hayes, J. D. Friedman and M. G. Sciammarella et al. 2004. Febiger, Philadelphia, PA. Prognostic validation of a 17-segment score derived from a 20-segment score for myocardial perfusion SPECT interpretation. J. Nucl. Cardiol. 11(4): 414–423; Rozanski, A., G. A. Diamond, Screen selection J. S. Forrester, D. S. Berman, D. Morris and H. J. Swan. 1984. (Diagnostic Radiology) Intensifying screens are selected for spe- Alternative referent standards for cardiac normality – Implications cific clinical applications based on requirements for image detail for diagnostic testing. Ann. Intern. Med. 101(2): 164–171. and considerations for patient exposure. Two types of screens are illustrated on Figure S.20. Thin screens produce less blurring and Scout view provide better visibility of detail but do not absorb as much of the (Diagnostic Radiology) Scout view is a vendor name (GE) for the radiation as thicker screens. Thicker screens, described as being scan projection radiograph used in computed tomography. faster, can produce images with less exposure but with more Related Article: Scan projection radiograph blurring. SCR (silicon-controlled rectifier) Screen speed (Diagnostic Radiology) See Silicon-controlled rectifiers (SCRs) (Diagnostic Radiology)