Datasets:

krvhrv

/

crawl-books

like 0

time of the returning disease, e.g. cancer, over a time-period from all carcinogenic echoes and an assumption of the sound speed. Clearly, these factors is known as the absolute risk. conditions are contradictory as tissue is heterogeneous, but since When carrying out epidemiological studies to quantify the the sound speed variations between tissue types are relatively effect of a particular factor, e.g. exposure to ionising radiation, small, ultrasound imaging works well to a first order. in causing a biological effect such as cancer induction, it is nec- Aberration Correction: In ultrasound imaging, the beam- essary to know the absolute risk, and to be able to factor out the former phases and sums signals from individual transducer contribution of other carcinogens in order to isolate the effect of elements. Ideally these signals differ only in the arrival time, the radiation exposure as an excess risk to the exposed popula- which relates directly to the distance between the receiving tion, or to compare the relative risk of the carcinogen, with other element and the reflector. However, the propagation in tissue factors. breaks up the amplitude consistency, and makes also parts of the Related Articles: Excess risk, Relative risk wavefront arrive earlier or later than had the wave propagated in a homogenous medium. This is due to the combined effects of Absorbed dose scattering, reverberation, refraction and the cumulative differences (Radiation Protection, Radiotherapy) Absorbed dose is one of the in time delay associated with the passage through tissue layers of fundamental radiation dose quantities defined by the International different types and thicknesses. Studies of aberration correction Commission on Radiological Protection (ICRP). aim to correct for these tissue propagation effects, to recover the At an atomic level the variations in specific interactions full potential of the imaging system. between incident ionising radiation and individual atoms or Correction Procedures: A variety of correction methods have molecules in an absorber imply that the resultant distribution been devised to solve this problem, usually applied to sites like of energy absorption is a stochastic process. The definition of the abdominal wall and the chest wall. Usually, two steps are absorbed dose takes this into account by averaging out such involved: determining the degree of aberration and then to correct variations over a volume per unit mass of absorber, and can for it in an adaptive way. In astronomy, atmospheric aberration therefore be considered a macroscopic quantity. can be corrected by maximising the intensity integral in the image Absorbed dose is therefore formally defined as the mean plane through real-time adjustments of time delay. In ultrasound energy absorbed within a mass of absorber: imaging of the body, point targets (like stars in astronomy) are not readily available, so these methods have limited applications, de except for certain cases like kidney stones. However, random D = dm backscatter (as from tissue) can also provide measurement of the phase error. The usual approach is therefore to assume a where de− is the mean energy imparted in joules (J) over a mass phase screen in front of the transducer, i.e. to assume that the of tissue dm. main aberration effect is a phasing effect. By cross correlation of Related Articles: Mean absorbed dose in air, Mean energy signals on adjacent elements, an estimate of the phase error can imparted, Kerma, Air kerma be obtained. Once this is known, a delay of the opposite sign is applied to compensate for the aberration. Iteration can reduce the error further. To make an optimal correction, investigators have Absorbed dose conversion factor shown that also amplitude must be corrected for. At present, the (Radiation Protection) The absorbed dose conversion factor, fX,D, questions remain if aberration correction provides any significant is the relation between the absorbed dose in air and the exposure. improvement, and if the introduction of the massive computer That is, power to cope with the calculations is worthwhile. D = fX ,D*X Ablation (Radiotherapy) Ablation is a procedure that produces scars in a where small volume of tissue. D is the absorbed dose in air There are two techniques used to achieve this by introducing X is the exposure catheters into a vein or artery: Even if X is only well defined in air, sometimes, especially 1. Radiofrequency ablation – This uses radiofrequency for diagnostic radiology or nuclear medicine energies, where elec- waves to deposit energy, heating the tissues. tronic equilibrium is easily established, it is possible to obtain an 2. Cryablation – This uses a single catheter with a balloon acceptable approximation for absorbed dose in other media such containing a refrigerant. The balloon is inflated and then as water or muscle tissue using tabulated values of absorbed dose pre-cooled refrigerant is introduced to produce the scar. conversion factors for these media. Absorbed dose distribution 9 A bsorbed radiation In a similar way, a dose conversion factor, fK,D, can be defined for many years is one developed by the MIRD committee. The to calculate absorbed dose from air kerma measurements: phantom is a standard model of the human being with fixed organ A sizes and anatomic relationships. Even though the phantom is a D = fK ,D*K relatively good approximation of the average human, the devia- tion from the standard phantom can be large for specific patients. where Specific Absorbed Fraction: The specific absorbed fraction D is the absorbed dose Φ is given by the quotient between the absorbed fraction and the K is the kerma target organ mass (Equation A.3): At higher energies, for instance in radiotherapy, absorbed dose f conversion factors are usually associated to chamber calibration F = (A.3) m factors. t Related Articles: Exposure, Kerma, Absorbed dose Φ is the fraction of radiation emitted by the source organ that is absorbed per unit mass in the target organ mass and has the Absorbed dose distribution unit Gy. The absorbed dose to the target organ using the specific (Radiation Protection) Absorbed dose distribution is a descrip- absorbed fraction (insertion of Equation A.3 in Equation A.2) is tion, often a graphic description, of energy deposition across the given by matter where the ionising radiation impinges on. Due to the lack of homogeneity in the medium and to the radiation attenuation, the energy deposition may adopt different patterns, which may be D (r r A k ¬ h ) = åFi (rk ¬ rh )Di (A.4) more or less complex. i A usual way to represent the absorbed dose distribution is to use isodose curves. Dose Reciprocity Theorem: Consider an organ pair. The theo- Sometimes absorbed dose distribution may be determined rem states that the specific absorbed fraction is the same, regard- using a set of dosimeters (such as TLDs) located in the real less of which organ is the target and which is the source, i.e. the medium or in a simulated one. In many cases absorbed dose dis- energy absorption per gram is identical regardless of whether the tributions must be calculated by means of an appropriate math- radiation is travelling between rk and rh or rh and rk. This theo- ematical model or through a Monte Carlo simulation rem is useful when the absorbed fraction is not available for all Related Articles: Absorbed dose, Energy deposition source–target pairs. For example, when φi (rk ← rh) is known, then φi (rh ← rk) can be calculated according to Absorbed fraction (Nuclear Medicine) When determining the radiation dose f(rh ¬ rk ) f(rk ¬ rh ) = received by a target organ from a source organ, the final step is mh mk to determine the absorbed fraction ‘φ’. The absorbed fraction is (A.5) a measure of the fraction of the energy emitted from a source ( m f r r h h ¬ k ) = f(rk ¬ rh ) mk organ that is absorbed in the target organ. The absorbed fraction depends on the target organ composition and the amount of radia- tion reaching it (i.e. the distance between the two organs). For The absorbed fraction and the equilibrium absorbed dose constant example, a high absorbed fraction indicates that the source and are often combined into a mean dose per cumulated activity S target organ are located adjacent or near to each other and that value (see separate article) to simplify the calculation procedure. the target organ is ‘composed’ of a high attenuating tissue. The The S value is determined by the emission type, radiation energy absorbed fraction is therefore calculated for each emission type and anatomic relationship and is determined for each source–tar- and source-target organ pair. The notation φi (rk ← rh) (Gy kg) get pair and radionuclide. indicates the absorbed fraction from a source organ rh to the tar- Related Articles: Cumulated activity, Equilibrium absorbed get organ rk from the ith emission. Thus the dose absorbed by the dose constant, MIRD formalism, Mean dose per cumulated target organ is given by activity Further Reading: Cherry, S. R., J. A. Sorensen and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, ( A D rk ¬ rh ) = åfi (rk ¬ rh )Di (A.2) Philadelphia, PA, pp. 412–415. mt i Absorbed radiation where (Radiation Protection) When radiation goes through matter, a D (Gy) is the total dose to the target organ received from the variety of phenomena may happen. Those phenomena can be source organ classified in three classes: a number of photons or particles pass à (MBq s) is the cumulated activity through without interacting; some others may be scattered with mt (kg) is the target organ mass or without loss of energy; and finally a part of the incident radia- Δi (Gy kg/Bq s) is the equilibrium absorbed constant for the tion delivers all of its energy to the matter. This is the absorbed ith emission radiation. The non-penetrating radiation (photons with energy less than At low energies, absorbed radiation is mainly associated with 10 keV and electrons) is assumed to deposit all its energy locally, the photoelectric effect. Pair production also contributes to radia- i.e. in the source organ. tion absorption at higher energies. Values of the absorbed fraction are calculated using a phan- Related Articles: Scattered radiation, Photoelectric absorp- tom model of the human body. The phantom which has been used tion, Photoelectric effect, Pair production Absorber, broad-beam geometry 10 Absorber, narrow-beam geometry Absorber, broad-beam geometry where A (Radiation Protection) The geometry used for an x-ray exposure ρ is the density (kg/m3) will affect the measured transmission of the x-ray beam, and the m is the mass (kg) subsequently calculated attenuation on an absorber. Broad-beam v is the volume (m3) and narrow-beam geometry set-ups are shown in Figure A.1a and b. In the broad-beam set-up shown in Figure A.1a, radiation Absorber, linear attenuation coefficient that would have reached the detector (A) is scattered and does (Radiation Protection) Attenuation of photonic radiation is an not reach the detector. However, some radiation that would not exponential process that can be represented by the equation originally have reached the detector is scattered so that it does (B). This may lead to an increase in detector reading. For narrow- I = I e-mx beam geometries (Figure A.1b), it is assumed that the primary 0 x-ray beam has been collimated and therefore any scattered radia- tion will miss the detector. where The result of this difference in set-ups is that for a broad-beam I0 is the initial beam intensity geometry, the same absorber does not appear to attenuate as much I is the final beam intensity as if it were part of a narrow-beam geometry set-up. As a result, x is the distance travelled the value of the half-value thickness calculated for the absorber μ is the attenuation coefficient of the material (absorber) would be different for each set-up. This effect has practical impli- cations, e.g. the field size must be reduced when x-raying patients. The coefficient μ normally refers to the linear attenuation If a broader beam than necessary is used, extra scatter reaches the and describes the attenuation property of an absorber in terms detector and contrast is reduced. of attenuation per unit thickness. The unit for μ is 1/length (e.g. cm–1 It is important to note that for a certain absorber, the radiation ). absorbed will be the same for both narrow- and broad-beam set- See Attenuation. ups. It is the

scattering effects that cause the apparent decrease Related Articles: Attenuation, Mass attenuation coefficient, in attenuation for broad-beam geometries compared to narrow- Atomic attenuation coefficient beam geometries. Further Reading: Dendy, P. P. and B. Heaton. Physics for Absorber, mass attenuation coefficient Diagnostic Radiology 2011, 3rd edn. (Radiation Protection) The amount of attenuation in an absorber may be related to the mass or density of the absorbing material in Absorber density terms of its mass attenuation coefficient, given by (Radiation Protection) The density of a particular absorber is dependent on its mass and volume (Equation A.6). The more m dense the absorber, the more likely photon interactions are to r occur. An increase in density will lead to an increase in absorp- tion and scatter, and therefore an increase in attenuation. Such where increased absorption in denser materials implies that they receive μ is the linear attenuation coefficient a higher radiation absorbed dose: ρ is the density of the material m r = (A.6) For more information, see article on Attenuation. v Related Articles: Attenuation, Mass attenuation coefficient A Absorber, mean free path for photons (Radiation Protection) Mean free path for photons in absorbers is Absorbed B shorter than that for photons in gas due to the increased density photon of molecules within the absorber. This increased density is analo- B gous with an increased value of n in Equation A.7: Source A 1 l » A 7 p 2 ( . ) n r (a) Absorber where l is the mean free path n is the number of molecules Absorbed photon r is the collision radius of the molecule Related Article: Mean free path Absorber, narrow-beam geometry Source Scattered (b) Absorber photon (Radiation Protection) The geometry of the x-ray set-up may affect the output measured by the detector. This is described in the article Absorber, broad-beam geometry. FIGURE A.1 (a) Broad-beam geometry and (b) narrow-beam geometry. Related Articles: Absorber, broad-beam geometry Absorption 11 Absorption contrast tomosynthesis Absorption I = I0*(1- exp(-ma*x)) (Radiation Protection) Absorption is the process by which the A energy of a beam of radiation incident on a material is imparted Absorption coefficients depend on the medium composition, to the atoms or molecules of the material. Some of the energy material density and radiation energy. Material composition in the incident beam will possibly be transmitted out of the dependence is determined by the atomic number of the medium material. or by the so-called effective atomic number when a mixture of Any material that absorbs energy in this way is called an more than element is present. Density dependence may be avoided absorber. Materials can also scatter the incident radiation to dif- through the use of mass absorption coefficients. ferent directions through various processes (Compton scatter, Radiation absorption may arise from several interaction pro- elastic scatter, etc.). The scattered radiation may also be absorbed cesses. So, several absorption coefficients can be defined, for or may be transmitted out of the material. instance, associated with photoelectric absorption, μ The most important absorption process at diagnostic x-ray a,pe, and with pair production, μa,pp, etc. The total absorption coefficient is the energies is the photoelectric effect, and is the main mechanism combination of all of them: by which the energy of the radiation is transferred to the material, potentially leading to damage if the material is human tissue made up of biological molecules, cells and most importantly, DNA. ma = ma, pe + ma, pp + Related Articles: Photoelectric effect, Compton scatter In general, the attenuation of radiation is the result of absorption Absorption and scattering, so the absorption coefficients combine with scat- (Ultrasound) The intensity of a propagating ultrasound wave is tering coefficients as well. decreased by distance. This phenomenon is called attenuation Related Articles: Mass absorption coefficients, Scattering and is due to scattering, reflection, divergence and absorption of coefficients, Attenuation coefficients the ultrasound beam energy. Absorption is caused when there is imperfect relationship between the pressure changes in the sound Absorption contrast tomosynthesis wave and corresponding resulting density changes in the medium. (Diagnostic Radiology) Tomosynthesis is an x-ray imaging The lost mechanical energy is converted into heat. The sound modality relying on the movement of the x-ray source (along an energy is lost and cannot be recovered. Absorption is dependent arc with aperture/angle, which depending on the manufacturer, on tissue composition and structure and increases with frequency. may vary between 15 and 50º) and, in some cases, movement of Related Articles: Attenuation, Damping, Intensity, Scattering the detector in the horizontal plane. During the scan, a number of images are taken. The data are then reconstructed into a set Absorption coefficients of images each showing a slice at a specific depth in a patient in (Radiation Protection) For a given medium, the absorption coeffi- focus, with the background structures blurred out. cient, μa, is a measure of its photon radiation absorption efficiency. This allows cross-sectional information without the complex When a beam of intensity I0 traverses a medium of thickness x the equipment and the comparatively high doses required in CT. amount of radiation absorbed in it is given by The principle is illustrated in Figure A.2. FIGURE A.2 Schematic diagram of a tomosynthesis acquisition. Absorption cross section 12 AC motor Cross-sectional images may be reconstructed using filtered Transducer elements A back projection or iterative methods. Recently, deep learning approaches have been suggested (Moriakov et al., 2019). Pulse excitation Although sometimes used for chest (Ferrari et al., 2018) or Displacement dental imaging (Inscoe et al., 2018), the main field of application of tomosynthesis is breast imaging – digital breast tomosynthesis (DBT). Related Articles: Phase contrast tomosynthesis, Back projec- Undamped transducer element Long pulse length tion reconstruction, digital breast tomosynthesis Poor axial resolution Further Readings: Dobbins, J. T. III and H. P. McAdams. 2009. Chest tomosynthesis: Technical principles and clinical Pulse excitation update. Eur. J. Radiol. 72(2):244–251; Ferrari, A. et al. 2018. Displacement Digital chest tomosynthesis: The 2017 updated review of an emerging application. Ann. Transl. Med. 6(5):91–97; Inscoe, C. R. Backing et al. 2018. Characterization and preliminary imaging evaluation of a clinical prototype stationary intraoral tomosynthesis system. Med. Phys. 45(11):5172–5185; Moriakov, N. et al. 2019. Deep Damped transducer element Short pulse length learning framework for digital tomosynthesis reconstruction. In Good axial resolution Proc SPIE 10948, Medical Imaging 2019: Physics of Medical Imaging, p. 1094804. FIGURE A.3 The effect of backing a transducer element. (Courtesy of EMIT project, www .emerald2 .eu) Absorption cross section (Ultrasound) The absorption cross section is defined as the time- Magnetic averaged total absorbed power divided by the time-averaged inci- field dent intensity. The unit is in square meters. Physically this cross section corresponds to the area of the incident wave that contains Rotation the amount of power that is absorbed by an object. This means Sinusoidal that the absorption cross section divided by the geometrical cross alternating section of the object is a measure of how effectively the object voltage absorbs sound. See also Scattering cross section, Extinction cross section and Differential scattering cross section. FIGURE A.4 AC generator principle. Absorption efficiency (Radiation Protection) Absorption efficiency is a term used generates a sinusoidally varying electrical potential or which can to describe how efficiently a detector absorbs incident ionising be used as a source of electrical power. The frequency of the AC radiation. The term might be used either in the field of radiation power is directly related to the rotational speed of the rotor. protection (see Absorbed dose) and associated measurements, or The sketch shows a single coil and slip ring pair, though in the field of diagnostic radiology imaging (see DQE). mounting three coils and slip ring sets at 120° intervals on the Related Articles: Absorbed dose, Detective quantum rotor would enable the generation of three-phase power. efficiency. AC motor (General) An AC motor is a motor that is powered by alternating Absorptive backing current. The motor is made up of an outer stationary part (the (Ultrasound) Absorptive backing is used in an ultrasound trans- stator) which produces a magnetic field and an inner rotating part ducer to damp out vibrations in the transducer element in order to (the rotor) which produces an opposing field that generates a reac- produce short pulses (Figure A.3). tionary rotational force. To produce short pulses the backing material should be highly There are many types of AC motor designs, though they can absorptive and have acoustic impedance close to that of the piezo- be considered in two basic categories: the induction and the syn- electric material. However, this requirement runs counter to that chronous motor. of maximising the ultrasound output and therefore the amount of Synchronous motor – the rotor turns synchronously with the damping is a compromise between sensitivity and pulse length. In changing phases of the AC current, and the rotor is either a coil certain applications, such as in pulsed or continuous wave, sensi- powered through slip rings or can be a permanent magnet. One tivity is more important than spatial resolution and a less absorp- example is the shaded pole motor used in electric clocks. tive material can be used. Induction motor – here the rotor is not powered but has cur- Related Articles: Damping, Attenuation, Absorption, rent induced in it by the changing magnetic field provided by Matching layer, Bandwidth the stator. This type rotates slightly slower than synchronously. Common forms include the ubiquitous ‘squirrel cage’ rotor and AC generator the more expensive ‘wound’ rotor. These are found in most (General) An alternating current generator is based on turning a domestic equipment. Very large, powerful, electric motors are conducting coil in a magnetic field (Figure A.4). usually three-phase induction motors. The movement of the rotating coil causes the magnetic flux Related Article: AC generator through the coil to vary sinusoidally as the coil rotates. This Hyperlink: http://en .wikipedia .org /wiki /AC _motor Accelerated partial breast irradiation 13 Accessories Accelerated partial breast irradiation accelerated in a cyclotron and directed towards a target; hence, (Radiotherapy) Accelerated partial breast irradiation (APBI), the radionuclides produced in a cyclotron are referred to as accel- A also known as hypofractionated partial breast irradiation, is a erator-produced radionuclides. An alternative approach is to use localised form of adjuvant radiotherapy delivered after lumpec- photon-induced reactions by irradiating a target with high-energy tomy (breast-conserving surgery). It uses fewer fractions with a photons (~100 MeV). A downside with using photons is that the higher dose per fraction which shortens the treatment period, and most common nuclear interaction for low Z elements is the (γ, n). only delivers a therapeutic dose to the tumour bed rather than the For example, 18F is produced by irradiating 19F-gas. Since it is whole ipsilateral breast, which allows greater sparing of nearby almost impossible to separate two isotopes of the same element sensitive organs, including the heart and lungs. APBI may be one can never obtain carrier-free 18F with this approach. Carrier- delivered using external beam radiotherapy, high dose rate (HDR) free radionuclides are acquired when using a cyclotron to irradi- brachytherapy or intraoperative radiation therapy (IORT). ate 18O with a proton beam, giving a (p, n) reaction. Therefore Related Articles: External beam therapy, Intraoperative radia- cyclotron production is preferred to photon-induced production. tion therapy (IORT), Brachytherapy, Combining cancer therapies, Tables A.1 and A.2 contain clinically relevant accelerator- Hypofractionation produced photon and positron emitters. Related Articles: Carrier-free sample, Collimator, Cyclotron, Accelerating waveguide Positron emission tomography (Radiotherapy) See Wave guide Further Reading: Qaim, S. 2001. Nuclear data relevant to the production and application of diagnostic radionuclides. Acceleration compensation Radiochim. Acta 89(1):223–232. (Magnetic Resonance) During an ordinary MR imaging sequence, the object is expected to be at rest. If this is not the case, and Accelerators in film development the object moves during and/or between data acquisition, arte- (Diagnostic Radiology) The primary function of the accelerator, facts such as mispositioning and blurring may occur in the recon- typically sodium carbonate, is to soften and swell the emulsion so structed images. that the reducers can reach the exposed grains. Also, compounds Both moving and static spins will accumulate a phase offset such as ascorbic acid and thioether have been referred to as accel- when exposed to a gradient. However, the phase offset induced erators. See also Activators. on the moving spin is different from the phase of the static spin. Related Article:

Film processing Additionally, if the spin is accelerating or jerks, the spin phase will have an additional phase contribution. The different phase Acceptance test angle results in mispositioning and blurring of the object in the (General) The final quality control of equipment prior to its Fourier reconstruction. delivery is called acceptance test. The test is performed to make Flow compensation can be done by introducing additional gra- sure that all demands specified before the purchase have been dient lobes prior to the echo readout. The aim of these lobes is to met. In many cases the acceptance test is performed separately by null the phase offset of the spins induced by first-order motion both the system provider and the customer before transferring the (velocity). Likewise, by adding additional gradient prior to read- ownership. In such cases the customer test is called beta testing, out, second-order (acceleration) induced phase offsets and even user acceptance testing or end user testing. effects of jerk can also be nulled, although at the price of the utili- sation of rather complicated gradient lobe patterns and prolonged echo time. Accessible emission limit (AEL) Related Articles: Flow compensation, Phase contrast, Velocity (Non-Ionising Radiation) The accessible emission limit (AEL) mapping takes into account the laser output and access to the laser beam and is the maximum exposure a user has access to when using a Accelerator laser under normal use. This is usually stated in Watts for con- (General) Accelerators use electric fields to accelerate charged tinuous wave lasers and Joules for pulsed lasers. The AEL for the particles to high velocity and energy. There are two main types of treatment beam and aiming beam must be labelled on the laser as accelerators used in medical physics. shown in the standard. As a general rule, a higher AEL indicates Firstly linear accelerators in radiotherapy accelerate electrons a higher laser class. in a straight line to produce either a high-energy photon beam (by The manufacturer is required to provide AEL information and collision with a target) or an electron beam for treatment. Their accompanying information that allows for the calculation of ELV design and use are explained in more detail in the article Linear and NOHD and personal protective equipment requirements for accelerator. a particular laser. Secondly, cyclotrons are used in the production of radioiso- Related Articles: Nominal ocular hazard distance (NOHD) topes in nuclear medicine, notably PET imaging. A cyclotron accelerates a beam of charged particles (protons, deuterons, or Accessories alpha particles) along a circular path to gain energy and then (Radiotherapy) These are additional pieces of equipment which bombard targets of certain elements. This process produces pos- are attached to the exterior of a linear accelerator in order that itron-emitting radioisotopes such as 11C, 13N, 15O and 18F. This is the beam is modified to suit the treatment required. Some exam- explained further in the article Cyclotron. ples include electron applicators employed to provide a useable Related Articles: Linear accelerator, Cyclotron electron beam, physical wedges to change the dose distribution, and blocks or electron alloys to conform the treatment field to the Accelerator-produced radionuclides tumour shape and spare normal tissue. (Nuclear Medicine) Radionuclides are mainly produced in two Related Articles: Applicator, Electron applicator, Block tray, ways, either in a reactor or in a cyclotron. Charged particles are Low melting point alloy, Tertiary collimator, Wedge, Wedge filter Accidental coincidences of PET systems 14 Accumulator (storage battery) A TABLE A.1 Accelerator-Produced Photon Emitters Production Data Main γ-Ray Mode Off Energy (keV) Energy Thick Target Yield Radionuclide T½ Decay (%) (%) Nuclear Reaction Range (MeV) MBq(mCi)/μA h 67Ga 3.26 d EC (100) 93 (37) 68Zn (p,2n) 26 → 18 185 (5) 185 (20) (199.6) 67Zn (p,n) 111In 2.8 d EC (100) 173 (91) 112Cd (p, 2n) 25 → 18 166 (4.5) 247 (94) 111Cd (p, n) 123I 13.2 h EC (100) 159 (83) 123Te (p,n) 14.5 → 10 137 (3.7) 124Te (p,2n) 26 → 23 392 (10.6) 127I (p,5n) 123Xea 65 → 45 777 (21)b 124Xe (p,2n) 123Xea 29 → 23 414 (11.2)b 201Tl 3.06 d EC (100) 69–82 203Tl (p,3n) 201Pbc 28 → 20 18 (0.5)d (x-rays) (93) 166 (10.2) Note: These radionuclides are primarily used for scintillation camera and SPECT imaging. a 123Xe decays to 123I by EC (87%) and − (13%) emission. b 123I yield expected from 123Xe decay over an of time of approximately 7 h. c 201Pb decays to 201Tl by EC (100%). d 201Tl yield expected from the decay of 201Pb after 32 h. TABLE A.2 Accelerator-Produced Positron Emitters Production Data Mode Off Main γ-Ray Nuclear Energy Range Thick Target Yield MBq Radionuclide T½ Decay (%) Energy (keV) (%) Reaction (MeV) (mCi)/μA h 11C 20.4 min β+ (99.8) 511 (199.6) 14N (p,α) 13 → 3 3820 (103) EC (0.2) 13N 10.0 min β+ (100) 511 (200) 16O (p,α) 16 → 7 1665 (45) 15O 2.0 min β+ (99.9) 511 (199.8) 14N (d, n) 8 → 0 2368 (64) EC (0.1) 15N (p, n) 10 → 0 2220 (60) 18F 109.6 min β+ (97) 511 (194) 18O (p, n) 16 → 3 2960 (80) EC (3) 20Ne (d,α) 14 → 0 1110 (30) Note: These radionuclides are primarily used for PET imaging. Accidental coincidences of PET systems voltage. A cell consists of two dissimilar substances, a positive (Nuclear Medicine) Accidental coincidences refer to false coinci- electrode and a negative electrode, that conduct electricity, and a dences, i.e. the decay has occurred outside the line of response. third substance, an electrolyte, that acts chemically on the elec- Examples of accidental coincidences in PET are scattered, spuri- trodes. The two electrodes are connected by an external circuit; ous and random coincidences. To read more about each individual the electrolyte functions as an ionic conductor for the transfer event type, read the related articles. of the electrons between the electrodes. Batteries are classed as Related Articles: Object scatter events, Scatter coincidence, either dry cell or wet cell. In a dry cell the electrolyte is absorbed Random coincidence, True coincidence, Spurious coincidence, in a porous medium, or is otherwise restrained from flowing. In Event type in PET a wet cell the electrolyte is in liquid form and free to flow and move. Batteries also can be generally divided into two main types Accumulator (storage battery) – rechargeable and nonrechargeable, or disposable. Disposable (General) An accumulator or storage battery is a device that con- batteries, also called primary cells, can be used until the chemi- verts chemical energy into electrical energy, consisting of a group cal changes that induce the electrical current supply are complete. of electrochemical cells that are connected to act as a source of Rechargeable batteries, also called secondary cells, can be reused direct current. The cells are encased in a container and fitted with after being drained. This is done by applying an external elec- terminals to provide a source of direct electric current at a given trical current, which causes the chemical changes that occur in Accuracy 15 Acoustic power use to be reversed. A battery called a storage battery is gener- developer solution out of the emulsion and help stop the devel- ally of the wet-cell type, i.e. it uses a liquid electrolyte and can oper activity. A be recharged many times. The storage battery consists of several Related Article: Film processing cells connected in series. Each cell contains a number of alter- nately positive and negative plates separated by the liquid elec- Acoustic axis trolyte. The positive plates of the cell are connected to form the (Ultrasound) The ultrasonic beam axis is defined as the line fitted positive electrode; similarly, the negative plates form the negative to points of maximum acoustic pressure measured at increasing electrode. In the process of charging, the cell is made to operate in distances in the direction of propagation of a transmitted reverse of its discharging operation; i.e. current is forced through ultrasound field. the cell in the opposite direction, causing the reverse of the Further Reading: Report IEC 61390. Ultrasonics-real-time chemical reaction that ordinarily takes place during discharge, so pulse-echo systems-test procedures to determine performance that electrical energy is converted into stored chemical energy. specifications, standard number IEC/TR2 61390-1996, Batteries are made of a wide variety of electrodes and electrolytes International Electrotechnical Commission, Geneva, Switzerland, where the most common are lead and sulphuric acid, alkaline bat- 1996. tery (alkaline), nickel cadmium (NiCd), nickel hydrogen (NIH2), nickel metal hydride (NiMH), lithium ion (Li-ion), and lithium Acoustic impedance ion polymer (Li-ion polymer). (Ultrasound) The quantity characteristic acoustic impedance (Z) is defined as the ratio of the driving force (acoustic pressure, p) to Accuracy the response (local particle velocity, v) and is a measure of how (Diagnostic Radiology) The term accuracy is used in quality con- difficult it is for a particle to move within a medium. This relation trol assessments. It describes the ability of a system to keep its is analogues to Ohm’s law where the electrical impedance is the parameters exactly as set up. ratio of the voltage (driving force) to the current (response). The For example, x-ray tube kVp accuracy refers to the ability of acoustic impedance can also be expressed as the product of the the x-ray system (the generator and the tube) to produce exposures medium’s density (ρ) and its speed of sound (c). For a plane wave with accurate kVp (the measured kVp to be identical with the set with linear propagation, Z can be obtained by kVp). The minimal number of exposures used for calculating the accuracy of each parameter of an x-ray system is 4 (optimal is p 6). The kVp accuracy (%) in this case is calculated as the % ratio Z = = rc between the mean error and the real value for all 4 exposures (this v has to be made for each focal spot): The dimensions of acoustic impedance are M/L2/T. The units are (Mean error) Rayls, where 1 Rayl = 1 Pa s/m Accuracy = 100 * (Realvalue) Values of the acoustic impedance for some common materials and types of human tissue are shown in the table: Usually if the resultant figure is <10%, the kVp accuracy is accept- able. The resultant figure can be positive or negative (depending Material Z (kg/m2/s−1 on the error, i.e. if the system produces a larger or smaller kV ) p than expected). Water 1.48 × 106 Similar methods and calculations can be made to calculate, Air 430 e.g. the accuracy of the timer (again the lesser the % the better Blood 1.67 × 106 the accuracy). Muscle 1.71 × 106 The accuracy described earlier is a widely used practical term. Skull bone 6.47 × 106 However it can be misleading as it actually represents the error of accuracy (1% accuracy describes an excellent x-ray system, but actually it means that the system produces parameters which are The acoustic impedance is a very important tissue parameter 99% similar to the set parameters – i.e. very accurate). The term in diagnostic ultrasound as reflection of pressure waves occur accuracy is different from the term precision (see eponymous at boundaries with materials with different values of Z, see article). Reflection coefficient. The methods to perform various QC measurements are Related Articles: Speed of sound, Reflection coefficient described in detail in the EMERALD materials. Related Articles: Consistency, Precision Acoustic power Hyperlinks: www .emerald2 .eu (Ultrasound) Acoustic power is a measure of the rate at which an ultrasound transducer produces acoustic energy. Acoustic power Accuracy is measured in watts (W). Typical values of acoustic power for (General) See Receiver Operating Characteristic ROC diagnostic scanners are in the order of 10–100 mW. However, in some Doppler modes higher values can be reached. Acoustic Acetic acid in film processing power is a most important parameter when calculating the risk of (Diagnostic Radiology) Acetic acid has been used as a stop bath heating tissue. in film processing. The purpose of the stop bath is to quickly The output power is defined as time-averaged ultrasonic power stop the development process before the film is placed in the radiated by an ultrasonic transducer into an approximately free fixer. A stop bath is not used in the processing of radiographic field under specified conditions in a specified medium, preferably films. Rollers in the film transport system squeeze some

of the water. Acoustic pressure 16 Acoustics Acoustic power W can also be expressed as the integration of A intensity I over a specific area S: W = ò I dS Acoustic power is best measured using a force balance. Related Articles: Force balance, Radiation force, Intensity, Thermal index Further Reading: Report IEC 61161. Ultrasonics-Power measurement-Radiation force balances and performance require- ments, standard number IEC/TR2 61161–2006, International Electrotechnical Commission, Geneva, Switzerland, 2006. Acoustic pressure (Ultrasound) Acoustic pressure describes the pressure perturba- FIGURE A.5 Small-scale streaming near the oscillating object and the tion during the passage of a sound wave, as opposed to the static larger vortices represent the medium sized type of acoustic streaming. pressures (such as atmospheric and hydrostatic). The acoustic pressure is the pressure component detected by a hydrophone or ultrasound transducer. In common derivations of the wave equa- tion, it is assumed that the acoustic pressure variations are small Absorber compared to the static pressure. Acoustic radiation force impulse imaging (ARFI) (Ultrasound) Acoustic radiation force impulse imaging (ARFI) Sound uses ultrasound to create a 2D map of tissue stiffness (elasticity). source It generates a ‘push’ inside the tissue using the acoustic radiation force from a focused ultrasound beam. The amount the tissue along the beam axis is being ‘pushed’ down depends on its tissue stiffness; softer tissue is more easily ‘pushed’ than stiffer tissue. For more information, refer to point shear wave elastography (pSWE). FIGURE A.6 The larger type of acoustic streaming. Further Reading: Bruno, C., S. Minniti, A. Bucci and R. P. Mucelli. 2016. ARFI: From basic principles to clinical applications in diffuse chronic disease – a review. Insights Imaging 7(5):735–746. Acoustic working frequency (Ultrasound) The acoustic working frequency of an acoustic signal is based on the observation of the output of a hydrophone Acoustic streaming placed in an acoustic field at the position corresponding to the (Ultrasound) Acoustic streaming is a sound wave- or vibration- spatial-peak temporal-peak acoustic pressure. The signal is anal- induced time-independent flow. It results from a transfer of ysed either using the zero-crossing frequency technique or using momentum to the liquid as it absorbs acoustic energy. There are a spectrum analysis method. three main types of acoustic streaming: When analysing a pressure spectrum the acoustic working fre- quency is the arithmetic mean of the most widely separated fre- 1. The small-scale streaming that occurs near viscous quencies f1 and f2 at which the amplitude of the pressure spectrum boundary layers to an oscillating object. In this type of of the acoustical signal is 3 dB lower than the peak amplitude streaming the vortices are much smaller than the wave- (Figure A.7). length of the sound wave from the oscillating object. These types of flow can be seen in Figure A.5. 2. The medium-sized acoustic streaming occurs out- side the boundary layer of the oscillating object. The Symbol Unit streaming is also rotational in nature but the size of the fawf MHz vortices is in the same scale as the wavelength. These types of vortices are also shown in Figure A.5. 3. The final type of acoustic streaming is the largest where the vortices are much bigger than the wavelength. This Further Readings: International Electrotechnical type of streaming occurs when a beam of sound propa- Commission, Report IEC 61390; International Electrotechnical gates in a volume of liquid larger than the beam itself. Commission, Report IEC 60601-2-37 Figure A.6 depicts such a streaming. As can be seen the volume of liquid determines the size of the vortices. Acoustics (Ultrasound) Acoustics is the scientific study of the production Related Articles: Acoustic power, Safety and properties of sound waves. The word ‘acoustic’ often refers Acquisition modes for digital image 17 Action spectra (optical), AORD sequence, hence registering events only during specific parts of –3 dB0 the organ motion sequence. A –10 Related Article: Gated acquisition –20 –30 Acrylic (General) –40 –50 –60 Molar mass 15–500 kg/mol –70 Density at STP 1150–1190 kg/m 0 f1f2 5 10 15 20 25 30 Melting point 403–553 K Frequency (MHz) Boiling point 473 K Refractive index 1.4893–1.4899 FIGURE A.7 The acoustic working frequency is the arithmetic mean of CT number 110–130 HU the most widely separated frequencies f1 and f2 at which the amplitude of the pressure spectrum of the acoustical signal is 3 dB lower than the peak amplitude. (Courtesy of EMIT project, www .emerald2 .eu) Acrylic is known chemically as polymethyl methacrylate (PMMA) and by trade names such as Perspex and Plexiglass. It is to the entire frequency range while the word ‘sound’ is divided in a transparent thermoplastic, a synthetic polymer of methyl meth- infrasound (0–20 Hz), sound (20–20 kHz, human hearing inter- acrylate. Figure A.8 shows its chemical structure. It was devel- val) and ultrasound (>20 kHz). oped in 1928 and released commercially in 1933 by Rohm and The Greek word ‘acoustic’ means ‘able to be heard’ and ‘sonic’ Haas Company. is the Latin synonym. PMMA is usually used as an alternative to glass, because of its low cost and machine-ability. Its density is less than half that of glass, similar to other plastics. It is brittle when loaded and softer Acquisition modes for digital image than glass (i.e. easily scratched). A 3 mm thickness transmits up (Diagnostic Radiology) Imaging modalities use various modes to 98% of visible light entering through its surface, and reflects for acquiring digital images. The ‘acquisition mode’ for digi- around 4% of incident light from each of its surfaces. It has good tal images was a term used at the beginning of digital medical environmental stability compared to other plastics, but has a poor imaging when systems had only a limited number of acquisi- resistance to solvents. tion modes. For example, the fluoroscopic image from an image Medical Applications: PMMA has a good degree of compat- intensifier was acquired (and digitised) in two forms – inter- ibility with human tissue; therefore it can be used in a variety of laced and progressive mode. The interlaced mode (the one used medical applications. It is used as a replacement intraocular lens in broadcast TV) displays first the odd lines of the TV raster, in the eye when the original lens has been removed in the treat- then the even lines. This way the digital image is formed by ment of cataract. In orthopaedics, PMMA bone cement is used to two combined half-frames. The reason for using the inter- affix implants and to remodel lost bone. Dentures are often made laced mode is that it prevents perceived flicker in the viewed of PMMA, and can be colour-matched to the patient’s teeth and image because it fills the full display from top to bottom faster. gum tissue. Alternatively the progressive (non-interlaced mode) displays all PMMA is also extensively used in medical physics applica- lines of each frame in sequence. Hence the final digital image tions, predominantly for quality assurance tests. It acts as a cheap is formed by one frame that contains all of the lines. This mode alternative to specially developed phantoms to simulate patient however requires not only a change in the TV raster function conditions by providing scatter in radiation fields. (with doubled frequency), but also a more powerful analogue to Related Articles: CT Number, Plastic, Perspex characteristic digital converter. Another mode, not used today as a term, was the frame mode, which acquired separate images (frames) with certain speed (frames per second) – a mode similar to cut film Action spectra (optical), AORD changer operation. (Non-Ionising Radiation) Optical radiation can produce either photochemical or thermal damage, which depends on the energy of its photons, hence their wavelength and their penetration depth. Acquisition time (Nuclear Medicine) The acquisition time refers to the time span where the detector system counts and registers events. The length CH3 CH3 CH3 of the acquisition time is primarily determined by the detector CH count rates. For example, in order to attain an acceptable signal to 2 CH2 CH2 . . . C C C . . . noise ratio at low count rates, a long acquisition time is required. When imaging organs with sequential movements, like a beat- C C C ing heart and a breathing lung, the spatial resolution is severely O O O O O O degraded by motion distortions when using a continuous acquisi- tion. An alternative is to use gated acquisition. The acquisition CH3 CH3 CH3 time is divided into small portions of the total sequence time. Each acquisition time will be triggered at a certain point in the FIGURE A.8 Chemical structure of the PMMA polymer (C5O2H8)n. Amplitude (dB) Activation cross section 18 Activation cross section Action spectra are dimensionless functions of the wavelength, be induced in the metal components in the head of the linac as a A mostly varying between 0 and 1, which are used by photobi- result of photon and neutron activation. Figure A.10 shows that ologists to weigh hazards appropriately. The weights have been at 18 MV the main radionuclides produced are from such n,γ established primarily by empirically by either epidemiology stud- reactions. ies or laboratory experiments. The AORD refers to action spectra The likelihood of this activation process occurring is called agreed by ICNIRP. the activation cross section, and depends on the type and energy There are three key action spectra: of radiation being produced by the linac, other output parameters such as the monitor units (MU) and total beam-on time, and the • S(λ): an ultraviolet action spectrum, which is based on atomic number of the components within the linac head. the light-skin interactions Other articles such as furnishings and fittings within the treat- • B(λ): a blue light action spectrum, which takes into ment room may also have radioactivity induced in them by acti- account the photochemical interactions between visible vation. Similarly the air in the room may also become slightly light (primarily around the blue region) radioactive due to γ n reactions with nitrogen and oxygen, produc- • R(λ): a thermal action spectrum which takes into ing positron emitters with a short half-life. account the heating hazards on both the eye and skin of Related Articles: Cross section, Neutron, Neutron activation visible and infrared radiation cross section, Photon, Positron, Radionuclides Further Readings: NCRP Report No 151. Structural shield- The spectra are tabulated into the AORD Directive and reported ing design and evaluation for megavoltage X- and gamma-ray as shown in Figure A.9. radiotherapy facilities. December 2005; Rawlinson, J. A., M. In 2013, ICNIRP reviewed the thermal hazard and relative K. Islam and D. M. Galbraith. 2002. Dose to radiation thera- action spectrum, but the AORD Directive still refers to the one pists from activation at high-energy accelerators used for con- previously published. ventional and intensity modulated radiation therapy. Med. Phys. Related Articles: Action spectra, AORD, ICNIRP, Eye, Skin 29(4):598–608. Further Readings: ICNIRP. A closer look at the thresholds of thermal damage: Workshop report by an ICNIRP task group. Health Phys. 111(3):300–306; ICNIRP. 2013. Guidelines on limits of exposure to incoherent visible and infrared radiation. Health Probable Principal Phys. 105(1):74–91; ICNIRP. 2013. Guidelines on limits of expo- Half- Nuclear Decay Gamma-Ray sure to laser radiation of wavelengths between 180 nm and 1,000 Radionuclide Life Reaction Mode Energies (MeV) µm. Health Phys. 105(3):271–295; ICNIRP. 2004. Guidelines on Al-28 2.3 m 27Al(n,)28Al β−, γ 1.78 limits of exposure to ultraviolet radiation of wavelengths between Mn-56 2.6 h 55Mn(n,)56Mn β−, γ 0.85, 1.81, 2.11 180 nm and 400 nm (Incoherent Optical Radiation). Health Phys. Na-24 15.0 h 23Na(n,)24 87(2):171–186; ICNIRP. 2000. Revision of the guidelines on lim- Na β−, γ 1.37, 2.75 Sb-122 2.8 d 121 its of exposure to laser radiation of wavelengths between 400 nm Sb(n,)122Sb β−, β+, γ 0.51, 0.56 and 1.4 µm. Health Phys. 79(4):431–440. FIGURE A.10 Principal activation products in the head of a linear Activation cross section accelerator operating at 18 MV. (Reproduced from NCRP Report No (Radiation Protection) When linear accelerators (linacs) are 151, Structural shielding design and evaluation for megavoltage X- and operated above energies of 10 MeV significant radioactivity may gamma-ray radiotherapy facilities, December 2005.) FIGURE A.9 ICNIRP action spectra as published in the AORD Directive. Activation cross sections in radionuclide production 19 Activation formula thin target Activation cross sections in radionuclide production Further Readings: Helus, F. 1983. Radionuclides Production, (Nuclear Medicine) The amount

of activity produced when a Vol. 1, CRC Press, Boca Raton, FL, pp. 93–94; Krane, K. S. 1988. A target is irradiated with particles is described in the activation Introductory Nuclear Physics, 2nd edn., John Wiley & Sons, New formula and depends on the intensity of the particle beam, the York; Nordling, C. and J. Österman. 1999. Physics Handbook number of target nuclei and the reaction probability. This reaction for Science and Engineering, 6th edn., Student Literature, Lund, probability is expressed in terms of an effective area or a cross Sweden. section σ. The Système International unit for σ is m2 but it is com- monly expressed in barn or millibarn. (1 mb = 10−3 b = 10−31 m2.) Activation formula thin target Related Article: Activation formula (Nuclear Medicine) If a target is irradiated with particles it can Further Reading: Helus, F. 1983. Radionuclides Production, lead to the production of radioactive species. The rate, R at which Vol.1, CRC Press, Boca Raton, FL, pp. 93–94. this occurs depends on the intensity of the particle beam, the num- ber of target nuclei and on the reaction probability, i.e. activation Activation formula cross sections (for typical irradiation times and typical cross sec- (Nuclear Medicine) The amount of activity produced when a tions only a very small amount of the target material is consumed target is irradiated with particles depends on the intensity of the and can therefore be regarded as constant). particle beam I, the number of target nuclei N0 and the reaction The number of nuclear reactions per unit time in a thin target probability, i.e. activation cross sections σ. (Please see articles can be expressed as on Activation formula thin target and Activation formula thick target.) I × r × N × N = A s × x Related Articles: Activation formula thin target, Activation M formula thick target Further Reading: Helus, F. 1983. Radionuclides Production, where Vol.1, CRC Press, Boca Raton, FL, pp. 93–94. dN Activation formula thick target I = 0 dt (Nuclear Medicine) If a target is irradiated with particles it can lead to the production of radioactive species. The rate, R, at which N0 is the total number of particles impinging on the irradiated this occurs depends on the intensity of the particle beam, the area number of target nuclei and the reaction probability, i.e. activation ρ is the density cross sections. For typical irradiation durations and typical NA is Avogadro’s number cross sections only a very small amount of the target material is σ is the cross section of each individual atom in a target with consumed and this can therefore be regarded as constant. thickness x If a thick target is irradiated, the incident particles will M is the atomic mass of the target material decrease in kinetic energy as they progress through the target (in contrast with a thin target where the kinetic energy is assumed to The differential equation for the change of rate for the radio- be constant). This will result in a variation of corresponding acti- nuclide productions is vation cross sections as these depend on the energy of the incident particle. To express the activation in a thick target the excitation dR = N dt - lR dt function must be integrated. If we consider the thin target equa- tion (please see article Activation formula thin target) where N dt is the produced radionuclides over the time dt ( ) I × r × N × × = × A s x A t R l = (1 - e-lt ) λR dt is the radioactive decay during dt M The solution to this equation reads where DE R (t ) N = 1 l l ( - e ) x = dl = ¶E / ¶l This gives us a radioactivity of The equation can be rewritten as Estart A(t ) R N ( l I × r × N × × = = e = A s x × l 1 - ) (1 - el ) M A( ) I × r × N = A ( E 1 - e-lt ) s( ) t dE M ò (¶E /¶l) Ethreshold This formula is only valid for constant cross sections, i.e. thin targets. For thick targets one needs to integrate over the entire To calculate the maximum possible activity one must integrate energy range deposited in the target. over the whole energy range deposited in the target from Estart Related Articles: Activation formula thick target, Avogadro’s to the threshold energy Ethreshold. If not all energy is deposited in number, Activation cross sections the target material then one must integrate from Ethreshold to Eexit Further Readings: Helus, F. 1983. Radionuclides Production, instead. Vol.1, CRC Press, Boca Raton, FL, pp. 93–94; Krane, K. S. 1988. Related Articles: Activation formula thin target, Avogadro’s Introductory Nuclear Physics, 2nd edn., John Wiley & Sons, New number, Activation cross sections, Excitation function York; Nordling, C. and J. Österman. 1999. Physics Handbook Activation rates in radionuclide production 20 Active matrix array for Science and Engineering, 6th edn., Student literature, Lund, Related Articles: Deep inspiration breath hold, Gating A Sweden. – respiratory Further Reading: Wong, J. W., M. B. Sharpe, D. A. Jaffray, Activation rates in radionuclide production V. R. Kini, J. M. Robertson, J. S. Stromberg and A. A. Martinez. (Nuclear Medicine) See Activation formula 1999. The use of active breathing control (ABC) to reduce mar- gin for breathing motion. Int. J. Radiat. Oncol. Biol. Phys. Activators 44:911–919. (Diagnostic Radiology) The primary function of the activator (typically sodium carbonate) in film processing is to soften and Active device swell the emulsion so that the reducers can reach the exposed (Magnetic Resonance) An active device is any device that can grains. See also Accelerators. only serve its intended use with an external supply of power Related Article: Film processing by any means including electrical line, battery or gas power. Examples of active devices are ventilators, pacemakers, elec- Active breathing control troencephalographs, electrocardiographs and patient monitoring (Radiotherapy) devices. The fringe magnetic field at a relative low strength may Principle: Active breathing control (ABC) involves control- influence the functionality of an active device above a certain ling the patient’s respiratory cycle to reduce the effects of inter- field strength which depends on the specific device. Implantable nal anatomy motion on the delivered dose distribution in external active devices include cardiac pacemakers, neurostimulators, car- beam radiotherapy. A typical ABC approach is to use a mouth- diac defibrillators, drug infusion pumps, cochlear implants and piece with a flow sensor and value. The flow sensor is interfaced insulin pumps. The severity of any effect of the magnetic field to a control computer that monitors the patient’s breathing, pro- on the active device may vary from one device to another and ducing a display showing the breathing pattern as a sine wave-like there are also variations in the functionality thresholds for the trace. same active devices. Some devices may be labelled MR safe or Several approaches exist to implement breathing control. MR compatible but, this will only apply to the specific conditions Passive breathing control involves asking the patient to hold their stated in the manufacturer’s specifications. breath using an audio cue or by adjusting their breathing to fit a Related Articles: MR safe, MR compatible desired pattern using the trace displayed on the computer screen. A typical active breathing control approach is to use a mouthpiece Active implant with a flow sensor and valve. The flow sensor is interfaced to a (Magnetic Resonance) Active implants include any medical control computer that monitors the patient’s breathing, producing device that can only serve its intended use with the supply of a display showing the breathing pattern as a sine wave-like trace. power by any means including but not limited to line, battery or The valve is used to hold the patient’s breath at a desired position gas power. The use of an active implant is contraindicated for in the inhale or exhale part of breathing. MR imaging because they are either magnetically, electrically or Figure A.11 shows an ABC system. mechanically activated and this activation could be affected by Treatment Sites: Treatment sites affected by motion due to exposure to the fields present in a magnetic resonance scanner. breathing are the main candidates for the use of ABC. These Therefore patients with such devices should not be examined with include lung, liver and breast. ABC has also been used in the MR. These active implants include cardiac pacemakers, neuro- breast to exclude the heart from the radiation field to the left side stimulators, implantable cardiac defibrillators, implantable drug of patients. infusion pumps, cochlear implants and insulin pumps. The use of Abbreviation: ABC = Active breathing control, also Active electronically activated devices may also cause excessive heating breathing coordinator. that can result in burn injuries to patients undergoing MR proce- dures, as a result of conductive materials that have an elongated shape, such as electrodes, leads, guide wires and certain types of catheters (e.g. catheters with thermistors or other conducting components). Related Article: Implant Active matrix array (Diagnostic Radiology) The active matrix array is an integrated circuit formed out of a large number of photodetector elements connected to thin film transistors (TFTs). It can be produced as a large area matrix (currently in excess of 40 × 40 cm2), which allows it to be used as a fundamental constituent in modern digital x-ray imaging detectors. For medical imaging the active matrix array is used for both direct and indirect radiography. Arrays used for both types of imaging incorporate a two-dimensional array of imaging pixels, which consists of a switching element used for data read-out (typically a thin film transistor, TFT) and a sensing and storage element. FIGURE A.11 Photograph of an active breathing control system The active matrix utilises thin film technology which allows attached to a board used for breast treatment. The image shows the mouth- the deposition of hydrogenated amorphous silicon (a-Si:H), mak- piece, and the electrical connections to the valve and the flow sensor. ing it ideal for construction of both TFTs and photodiodes. Large Active matrix liquid crystal flat-panel display 21 Active matrix liquid crystal flat-panel display area arrays are formed by plasma deposition of thin layers of the array. In diagnostic radiology digital radiographs are usually appropriate materials (e.g. amorphous silicon) onto a glass sub- viewed on an active matrix flat-panel TFT LC display. They have A strate. Once deposited, they can be etched to the desired pattern become widely popular in the medical imaging industry because by a process called photolithography. they are smaller and lighter weight than their traditional cathode Figure A.12 shows a typical array used in a medical imaging ray tube (CRT) counterparts, offering equivalent, if not superior flat-panel detector. Within the array each pixel consists of a switch- resolution, contrast, viewing angle and response time. ing element and of an element able to detect incoming photons by The active matrix array is used to precisely control each indi- storing them as charge. The image read-out process is controlled by vidual pixel. Within a monochrome display each pixel consists altering the voltage applied across the switching element. Firstly, of the switching element (usually a TFT) which controls a pixel to allow each pixel to detect a signal during exposure, the volt- electrode (made of indium tin oxide – ITO, transparent electrode) age across each switching element is set to an ionisation or ‘off’ which in turn controls the transmission state of the LC and a stor- state. The signal is then read-out by changing the switching voltage age capacitor to maintain a constant voltage throughout the LC. row-by-row to the conducting or ‘on’ state which allows the charge Figure A.13 shows a pixel cross section of a typical LC display. stored in each pixel to be drained by the charge collector electrode and passed to the multiplexer. The voltage change is controlled by the gate line driver. As the read-out process is controlled by the Multiplexer ADC external circuitry, each row of pixels requires a separate control line driver to alter the switching voltage, and each column its own Charge amplifier amplifier. This process is called the active matrix read-out. The active matrix array allows the radiographic image signal Charge collector

to be read-out sequentially, line by line. Fluoroscopic images are electrode 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, TFT = Thin filmed transistor. Related Articles: TFT (thin film technology), Amorphous sili- Switch, con detector diode or TFT Further Reading: Beutel, J., H. L. Kundel and R. L. Van Metter. 2000. Handbook of Medical Imaging: Physics and Psychophysics, Vol. 1, SPIE, Bellingham, WA, pp. 79–159. Gate line Active matrix liquid crystal flat-panel display (Diagnostic Radiology) An active matrix thin film transistor liq- uid crystal display is a thin, flat screened display whose pixels Data line are created from an array of liquid crystal (LC) cells which are each individually controlled by a separate TFT in an active matrix FIGURE A.12 An active matrix array and peripheral electronics. Black matrix Colour filter Polariser Common substrate Common electrode (ITO) LC capacitance TFT Gate Source Drain (G) (S) (D) Pixel electrode (ITO) Storage capacitor Active matrix substrate Polariser Diffuser Backlight FIGURE A.13 The cross section of a typical TFT display pixel using a twisted nematic liquid crystal configuration between the two ITOs. Gate line driver Active shielding 22 A ctivity outside the scanner, the strength of which falls with distance. A Data line driver The fringe field presents a safety hazard due to both the missile effect on ferromagnetic objects and the potential for interfer- G ence with the operation of electro-medical devices. In general, D Storage pixel S MRI installations are designed so that fringe fields no higher electrode (ITO) than 0.5 mT (5 gauss) are present outside the scan room, or at Common least outside an area designated as a controlled zone. Prior to electrode (ITO) the introduction of active shielding MRI installations had to routinely incorporate ‘passive’ steel shielding in the scan room walls to contain the magnetic field. Passive shielding can add tens of tonnes to a room design and is a significant consideration in structural design. With active shielding a set of magnetic field TFT coils in the MRI counters the field external to the machine. This reduces the footprint of the fringe field and the 0.5 mT (5 gauss) contour may be completely contained within a room of accept- ably small dimensions without the need for passive shielding. Static field shielding is achieved with a set of superconducting coils external to the main static field superconductor windings. FIGURE A.14 An active matrix array used for LCDs. The shield produces a field opposing the field produced external to the MRI by the main field coil. In gradient field switching, the resulting rapid changes in mag- The active matrix forms the image by controlling the pixel ele- netic field strength induce local eddy currents in conductive parts ments, row-by-row, shown in Figure A.14. Each row of pixels is of the MRI machine. These eddy currents generate magnetic selected by applying the appropriate select voltage to the gate line fields which interfere with the applied gradient field, resulting in connecting all the TFT gates in one row. When a row of pixels is gradient field distortion. In relation to the bore of the magnet, the selected, the desired signal is applied to each pixel in that row via gradient field coils sit internal to the main field coils and cryo- the data lines. In that way, each pixel is selected individually, and stat. With active shielding of the gradient an intermediary set of if the switching element works ideally, each pixel is addressed shielding coils sits between the gradient coils and main field coils with no cross-talk between adjacent pixels. and cryostat. These active shielding coils are designed to counter For monochrome displays each pixel is formed from a single the gradient field external to the gradient coils preventing eddy LC cell; in a coloured display each pixel is formed from three current induction in the scanner, while having minimal effect of sub-pixels of red, blue and green, Figure A.15. To create a colour the field internal to the gradient coils. pixel a coloured filter is usually imbedded on the second glass Further Reading: Chapman, B. L. W. 1999. Shielded gradients. substrate (Figure A.12); however, some development has been And the general solution to the near field problem of electromagnet done to allow the filter to be integrated into the TFT substrate. design. Magn. Reson. Mater. Phys. Biol. Med. 9:146–151. The final coloured image is then created by altering the intensity and brightness of each coloured cell, using a similar principle to Active transport of tracers the coloured phosphor screens used in CRT displays. (Nuclear Medicine) Active transport of tracers refers to transport Related Articles: TFT (thin film technology), Matrix array, mechanisms that require energy. Actively transported substances Active matrix array, LCD (liquid crystal display), Viewing angle. are able to move against concentration gradients. Examples of active transport are the sodium–iodine pump and the renal tubu- Active shielding lar reabsorption of glucose. (Magnetic Resonance) Active shielding refers to technical meth- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. ods used in MRI scanner design to counter undesirable effects of Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, both the static and gradient fields. Philadelphia, PA, p. 387. Ideally the static field should be contained completely inside the bore of the scanner. In real systems a fringe field will exist Activity (Nuclear Medicine) The activity of a radionuclide is the aver- age decay rate. It is measured in disintegrations per second. It is essentially a measure of how radioactive a substance is. Switching element The average decay rate, ΔN/Δt in a sample containing N atoms (TFT) is determined by the decay constant λ. The decay constant is the probability for disintegrations of a single atom per second (s−1). The mathematical expression is Sub-pixel cell DN = -lN (A.8) Dt As the minus sign indicates, the number of atoms N decreases Dot pitch with time. The activity A refers to the number of disintegra- tions per second of N radionuclides. The SI unit is the Becquerel FIGURE A.15 Coloured display pixel configuration. [Bq] where a sample with activity of 1 Bq decays at a rate of 1 Gate line driver Actual focal spot 23 Adaptive processing disintegration per second. The number of atoms N at a specific is known as aperture ratio – an analogue of the fill-factor of the time point t depends on the number of atoms at t = 0 and the pixels in flat panel detectors in digital radiography. Usually, this A decay factor. The decay factor e−‘λ’t is an exponential function that ratio is of the order of 0.9. depends on the decay constant ‘λ’ and time t. The mathematical The difference between the actual pixel size and nominal expression is pixel size of CRT monitors is larger compared with LCD moni- tors. Here, the actual pixel size is measured as the diameter of the 50% of the luminance profile (FWHM of the light spread around N (t ) = N (0)e-lt (A.9) the centre of the pixel). In CRT monitors there is some overlap of neighbouring pixels due to the spread of light. Since activity is proportional to the number of atoms, Equation Related Articles: Nominal pixel size monitors, Detector fill A.9 can be translated into factor Further Reading: AAPM Report assessment of display per- A(t ) = A(0)e-lt (A.10) formance for medical imaging systems. Each radioactive nuclide is associated with a certain probabil- Acute morbidity ity for decay. A high probability suggests that the atom is likely (Radiotherapy) Complications due to a radiotherapy treatment to decay and vice versa. When considering a greater number of with an onset within 90 days. atoms, it is more suitable to talk about the half-life T½ of the sam- Related Articles: Adverse effects (Radiotherapy), Long-term ple; namely the time it takes for radioactivity to decrease to 50% morbidity, Tolerance, Probability of complications, Sigmoid of its initial activity level. The half-life and the decay constant are dose-response curve, Dose response model related according to Adaptive collimation ln2 T (Diagnostic Radiology) One of the issues related to spiral CT, or 1/2 = l helical CT, is the exposure of regions of the patient before and after ln2 (A.11) the region of interest along the axial direction (overscanning). l = T Adaptive collimation consists of dynamically adjusting the 1/2 aperture of the collimators, so that it is smaller at the beginning and at the end of the scan, thus limiting the exposure to the In tables of radionuclides the half-life is often listed instead of the respective regions of the patient. decay constant. The concept is illustrated in Figure A.16. The activity of radionuclides used in nuclear medicine tends Abbreviations: CT = Computed tomography. to be multiples of becquerels, e.g. kilobecquerels (1 kBq = 103 Related Articles: CT, Helical scanning, Overscanning disintegrations/s), megabecquerels (1 MBq = 106 disintegrations/s) Further Reading: Deak, P. et al. 2009. Effect of adaptive and gigabecquerels (1 GBq = 109 disintegrations/s). section collimation on patient radiation dose in multisection The curie is another unit of activity. One curie is equivalent to spiral CT. Radiology 252(1):140–147. 3.7 × 1010 disintegrations/s. One curie was originally defined as the activity of 1 g of 226Ra. Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Adaptive processing Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, (Ultrasound) Adaptive processing in ultrasound covers a wide range Philadelphia, PA, pp. 32–33. of techniques designed to enhance the ultrasound signals from tis- sue features while reducing artefactual signals in the image, partic- Actual focal spot ularly those from speckle. Adaptive processing algorithms permit (Diagnostic Radiology) See Focal spot actual changes in response to local changes in time and space and may be based on the statistical analysis of signals. A number of filter- Actual pixel size, monitor ing techniques have been studied for ultrasound images including (Diagnostic Radiology) The actual pixel size (AS) of digital (e.g. as median filters, Weiner filters and wavelets. Increased computa- LCD) monitors is smaller than the nominal pixel size (NS). This tional power has enabled the implementation of several different is due to the space taken by the electronics of the pixel (i.e. not adaptive processing methods in commercial systems. An example all area of the pixel produces light). The ratio between the AS/NS of Philips XRES system in 2007 is shown in Figures A.17 and A.18. FIGURE A.16 Patient exposure in spiral CT with static collimation (left) and with adaptive collimation (right). Adaptive radiotherapy 24 Adaptive responses and hormesis for subsequent treatment fractions of a course of radiotherapy can A be modified to compensate for inaccuracies in dose delivery that cannot be corrected for by simply adjusting the patient’s position- ing. The causes of these inaccuracies may include tumour shrink- age, patient weight loss and increased tumour hypoxia resulting during the course of fractionated treatment. Image-guided radio- therapy (IGRT) is a requirement for this technique. Abbreviations: ART = Adaptive radiation therapy, cl-PTV = Confidence limited planning target volume, CT = Computed tomography, IGRT = Image guided radiotherapy and PTV = Planning target volume. Related Articles: Target volume, Planning target volume Further Readings: Martinez, A. A., D. Yan, D. Lockman, D. Brabbins, K. Kota, M. Sharpe, D. A. Jaffray, F. Vicini and J. Wong. 2001. Improvement in dose escalation using the process of adap- tive radiotherapy combined with three-dimensional conformal or intensity-modulated beams for prostate cancer. Int. J. Radiat. FIGURE A.17 Longitudinal image of a kidney without adaptive processing. Oncol. Biol. Phys. 50:1226–1234; Podgorsak, E. B. 2003. Review of Radiation Oncology Physics: A Handbook for Teachers and Students, International Atomic Energy Agency, Vienna, Austria. Adaptive responses and hormesis (Radiation Protection) The current internationally accepted framework for radiation protection is based on a model of potential harm from exposure to ionising radiation called the linear no-threshold model. This model suggests that at any level XRES of received radiation dose, harm may be caused – i.e. a cancer may be induced. The risk of harm (stochastic effects) is proportionate to the dose received (i.e. linear with dose). This is described in the diagram (Figure A.19). However, more recent published evidence would suggest that this model is too simplistic and that for a number of

reasons low- level exposure to ionising radiation may be more or less harmful than predicted by a linear extrapolation, and indeed may actu- ally be beneficial. Models predicting less than expected damage to radiation exposure may be attributed to adaptive responses by the cell. If the radiation exposure provides benefit in terms of FIGURE A.18 Longitudinal image of a kidney with XRES adaptive resistance to future carcinogenetic events, then it is called hor- processing. mesis. This hormetic effect appears in epidemiological evidence from mortality rates amongst British radiologists, survivors of Chernobyl, and studies in Taiwan. The diagram (Figure A.20) Adaptive radiotherapy describes the possible consequence of hormesis to our under- (Radiotherapy) Adaptive radiotherapy (ART), as its name sug- standing of the dose–response curve for stochastic effects. gests, involves delivering a radiotherapy treatment that is adapted Possible reasons for such a response have been postulated. to changes during the treatment course. Two basic concepts exist: Firstly, it is recognised that life on earth has evolved whilst adaptive planning and adaptive treatment delivery. continuously exposed to ionising radiation for natural sources, Adaptive Planning: The first approach developed for ART both terrestrial and from space. Secondly, it is recognised involved acquiring a set of planning scans of the patient on several that the genome of humans and other animals seem to include imaging visits to get a measure of the variation in the anatomy expected from day to day, such that a target volume may be con- structed based on an estimate of the expected inter-fraction varia- tion for the patient’s anatomy. This was first demonstrated for the prostate using a set of five CT scans. The combined information from the scans may be used to generate a confidence limited plan- ning target volume, cl-PTV. This enables a probabilistic approach to be used in treatment planning of target coverage. Adaptive radiotherapy delivery Linear, no threshold (LNT) model (Radiotherapy) A newer approach to ART is to measure the patient’s anatomy and estimate the dose distribution during each treatment fraction, accumulate this information and plan subse- Dose quent treatment fractions to correct for discrepancies between planned and delivered treatment. In this process the dose delivery FIGURE A.19 The linear no-threshold model. Probability of effect ADC 25 Adenosine triphosphate (ATP) A Hormetic (beneficial) effect Dose 1 FIGURE A.20 The hormetic model of dose response. Alu 2 mm ‘anti-cancer’ genes – genes that if ‘switched on’ appear to pro- tect from the effects of being exposed to carcinogens. Ionising radiation exposure at low levels seems to be a trigger to switch on these anti-cancer genes. FIGURE A.21 Added filtration − 2 mm Al disk in front of x-ray tube The International Commission for Radiological Protection housing (to be placed at opening 1). has stated in introducing its latest (2007) recommendations that although it is accepted that hormesis is probably real, such a dose–response curve could not be used as the basis for a frame- NH2 work for occupational, medical, or public radiation protection because of the difficulties in specifying where the boundary lies O O N N between radiation doses that are beneficial and those higher doses HO P O P O that are deleterious. N H H O N Related Articles: Linear no-threshold model, Radiobiological O O models, Stochastic effects β α Further Readings: Berrington, A. et al. 2001. 100 years of observations on British radiologists: Mortality from cancer Phosphate groups and other causes 1897–1997. Br. J. Radiol. 74:507–519; Chen, OH OH W. L. et al. Spring 2004. Is chronic radiation an effective pro- phylaxis against cancer? J. Am. Phys. Surg. 9(1):6–10; The FIGURE A.22 Molecular structure of ADP. 2007 Recommendations of the International Commission on Radiological Protection, ICRP Publication, p. 103, Ann. ICRP 37 (2–4), 2007. group experiences a different electronic environment and hence gives rise to a resonance peak at a different chemical shift in a 31 ADC P spectrum. (General) See Analogue-to-digital converter (ADC) These resonances are generally not seen however, because they overlie the resonances due to the γ and α phosphate groups in ATP and because most of the ADP (around 80%) is bound to Added filtration proteins and hence is not NMR-visible. (Diagnostic Radiology) The total filtration of an x-ray beam aims The importance of ADP in 31P NMR arises from its role in the to reduce the unnecessary low energy x-ray photons. It is pro- body’s energy metabolism. ADP is the product of dephosphoryla- duced by a combination of two filter components – inherent and tion of ATP, the ‘currency’ of intracellular energy. Although ADP added. The inherent filtration consists of existing components of concentration usually cannot be directly measured using NMR, it the x-ray system through which the x-ray beam passes (tube win- can be estimated using the formula dow, oil in tube housing, beam locator mirror, etc.). The added filtration of the x-ray beam is formed by sheets of metal (most often Al) that are added to provide the required total filtration. éëATPù éCrù éëADPù û û = ë û Added filtration is applied either in front of the x-ray tube hous- K + f éëPCrùû éëH ùû ing (Figure A.21) or inside the light beam diaphragm (collimator). Additive colour model where Kf is the forward rate constant of the creatine kinase (General) See RGB (red green blue) reaction. Related Articles: ATP, Magnetic coupling Adenosine diphosphate (ADP) (Magnetic Resonance) ADP (adenosine diphosphate) is a chemi- Adenosine triphosphate (ATP) cal compound that features in in vivo phosphorus (31P) NMR (Magnetic Resonance) ATP (adenosine triphosphate) is a chemi- spectra. The ATP molecule contains two phosphate groups, des- cal compound that features in in vivo phosphorus (31P) NMR ignated α and β in order of increasing distance from the remainder spectra. The ATP molecule contains three phosphate groups, of the molecule (Figure A.22). The phosphorus nucleus in each designated α, β and γ, in order of increasing distance from the Probability of effect Adherographic printing 26 Adhesive remainder of the molecule (Figure A.23). The phosphorus nucleus uses a laser beam which exposes the film by scanning it. The A in each group experiences a different electronic environment and pixel values of the digital image modulate the intensity of the hence gives rise to a resonance peak at a different chemical shift laser beam. in a 31P spectrum (Figure A.24). These chemical shifts are sensi- One of these printing methods uses adherographic film (hence tive to intracellular pH, Mg2+ concentration and temperature, and adherographic printing). The adherographic film process is gen- can be used to estimate these parameters. erally similar to the Polaroid photographic process. The adhero- Furthermore, the γ and α resonances appear as doublets and the graphic film has two layers – laser sensitive adhesive layer and β resonance as a triplet due to homonuclear J-coupling between another layer with carbon particles (also known as ‘pells’). Both phosphorus nuclei within the molecule. For this reason, the ATP layers are sandwiched between two polyester sheets. resonances appear to have artefactually short T2 values, so that When the laser beam (modulated with the intensities of the echo-based techniques generally cannot be used in 31P MRS. digital radiograph) scans the film, it causes the adhesive layer to Another complication is the fact that the γ peak contains con- attract carbon and stick it to the polyester sheet. Higher intensity tributions from the β resonance of ADP and the α peak contribu- of the laser beam (i.e. brighter pixel) attracts more carbon par- tions from the α resonance of ADP and also from NAD/NADH. ticles. As a result, there are two sheets, one with more and one Thus the β ATP resonance is normally used for quantitative with less carbon – i.e. with positive and negative images. The first purposes. sheet is the film that carries the radiographic image. The other The importance of ATP in 31P NMR arises from its role in the sheet is negative to the first one and is disposed. body’s energy metabolism. ATP is the ‘currency’ of intracellular The adhesion process is binary, and the grey tone (nuance) is energy, transporting chemical energy between different locations produced by dithering. Normally, a cell of 16 × 16 pells makes a within the cell for use in metabolism. pixel with 256 grey levels. This requires a very thin laser beam Related Article: Magnetic coupling and small pells (5 μm). This way, a pixel of 16 pells, each 5 μm, will have 16 × 5 = 80 μm. An image with 80 μm pixels will pro- Adherographic printing duce an image resolution of 6.25 lp/mm. (Diagnostic Radiology) The laser film printer (also known as This process is suitable for digital radiography. Before print- dry laser imager or laser camera) uses different types of films ing, the image should have been windowed in order to visualise to record the image from a digital x-ray detector. The printer the important anatomical features with the necessary contrast. In such cases, an indicative window width between 100 and 200 would deliver sufficient image contrast. Due to this reason, the H2N small contrast dynamics of the film (only 256 grey levels) is ade- quate for diagnosis. Adherographic films are often used in dental N radiography. N Related Articles: Laser film printer, Dithering, Window O O O N O– P O P O P O N Adhesive O– O– O– O (General) Adhesive, more commonly known as glue, is a com- pound that bonds two objects together. Adhesives come from γ β α either natural or synthetic sources. Natural or bio-adhesives are OH OH produced from inorganic mineral sources or biological sources Phosphate groups including plant matter, starch, resins and animal tissues. Synthetic adhesives can be elastomers, thermoplastics and thermosets. FIGURE A.23 Molecular structure of ATP. Adhesives can be further categorised depending on their method of adhesion. Drying adhesives usually are mixtures of polymers in a solvent, such as white glue and rubber cements. The adhesive hardens as the solvent evaporates. Drying adhesives are generally weak and tend to adhere to different materials to varying extents. ATP phosphate groups Contact adhesives are applied to both surfaces and are allowed γ α β to dry before the surfaces are put in contact, often requiring sev- eral hours to dry. Natural rubber and polychloroprene are com- mon contact adhesives and are used in laminates and footwear. Hot adhesives, such as the common ‘glue gun’, are thermoplas- tics which are applied when hot and harden on cooling. Reactive adhesives function by either chemical bonding with the material’s surface or hardening due to the polymerisation of two chemicals. Examples of such adhesives are two-part epoxy, silane, and metallic cross-links. They are used to prevent loosen- ing of bolts in moving assemblies such as engines. UV light curing adhesives or materials (LCMs) experience 25.000 20.000 15.000 10.000 5.0000 .00000 –5.0000 –10.000 –15.000 –20.000 –25.000 rapid curing, strong bonding and have the ability to withstand (ppm) high temperatures. They are therefore suitable for the manufac- ture of products in the electronics, telecommunications, medi- FIGURE A.24 31P NMR spectrum of the human brain showing ATP cal and aerospace industries. They are also used to seal and coat resonances. products. Adiabatic RF pulse 27 ARSAC Pressure-sensitive adhesives (PSAs) form a bond by the appli- z΄ cation of pressure due to a balance between flow and resistance A to flow. The bond forms when the adhesive is soft enough to flow (i.e. it is wet) and it has strength when it is hard enough to resist Δω/γ flow when under an applied stress. This is due to van der Waals forces, which determine the bond’s strength. PSAs are either permanent or removable, and are used for labels, tape, damping Beff α films, ‘blu-tack’, plastic wrap (‘cling film’) and plasters. M The strength of adhesion depends on many factors, includ- ing the means by which adhesion occurs. Adhesion can occur mechanically by the adhesive running into the pores of the sur- face, or by chemical mechanisms. A chemical bond may occur between the adhesive and surface. Electrostatic or van der Waals y΄ forces may hold the surfaces together. Adhesion may also be due to the moisture-aided diffusion of the adhesive into the surface followed by hardening. Some strong adhesives are important in modern construction and industry. Medical Applications: Adhesives are used in various medi- B1(t)

cal applications including prosthetic adhesives for catheters, den- x΄ tures and cosmetic purposes. PSAs are often used in skin contact uses, such as wound dressings, ECG electrodes, and analgesic and FIGURE A.25 The effective magnetic field rotates by modulating the transdermal drug patches. LCMs may also be used in medical B1-field. If the adiabatic condition is satisfied, the net magnetisation vec- equipment due to its strong bonding and ability to withstand high tor follows and precesses around the effective magnetic field vector. α is temperatures the resulting flip angle. Related Article: Gel One limitation of adiabatic RF pulses is the long pulse dura- Adiabatic RF pulse tion needed to fulfil the adiabatic condition. (Magnetic Resonance) The adiabatic phenomenon in magnetic Related Articles: Fluid attenuation inversion recovery resonance was described as early as 1946 by Bloch in experiments (FLAIR), Short tau inversion recovery (STIR) of nuclear induction (Bloch, 1946; Bloch et al., 1946). Further Readings: Bloch, F. 1946. Nuclear induction. Phys. In the common amplitude-modulated RF pulses used widely Rev. 70(7–8):460–474; Bloch, F., W. W. Hansen and M. Packard. in magnetic resonance, the carrier frequency of the applied RF 1946. The nuclear induction experiment. Phys. Rev. 70(7–8):474– field is held constant, while the RF amplitude is modulated. In 485; Hajnal, J. V. et al. 2001. Reduction of CSF artifacts on FLAIR this way, the resulting flip angle of the net magnetisation becomes images by using adiabatic inversion pulses. Am. J. Neuroradiol. dependent on the pulse duration and the B1 amplitude. In contrast, 22(2):317–322; Tannus, A. and M. Garwood. 1997. Adiabatic adiabatic RF pulses modulate both the frequency and amplitude pulses. NMR Biomed. 10(8):423–434. of the B1 field (Tannus and Garwood, 1997). When a RF field is applied, a spin population in the rotating frame experiences an effective field Beff, which is composed of Adjuvant therapy the B1-field vector, and a vector in the B0 direction with amplitude (Radiotherapy) Adjuvant therapy is a treatment that is given to proportional to the RF frequency offset. The effective field can patients after the initial primary treatment. Types of treatment change direction (rotate) in space if the applied RF field is fre- that are used as adjuvant therapy in cancer treatment include che- quency and amplitude modulated (Figure A.25). motherapy, hormone therapy, radiation therapy, immunotherapy If the effective magnetic field vector rotates sufficiently slowly and targeted therapy. during the radiofrequency pulse, the net magnetisation vector will Related Article: Combining cancer therapies precess around the effective magnetic field vector and thus follow it by T1rho relaxation. This condition is called the adiabatic condi- Administration of Radioactive Substances tion, and can be expressed as Advisory Committee (ARSAC) (Radiation Protection) Article 5(a) of the 76/579/Euratom da w Directive requires a process of authorisation prior to the admin- eff (t) dt istration of a radioactive substance to a person for the purposes of diagnosis, therapy or research. In the United Kingdom, this in which ωeff is the frequency of the effective B-field, and α is the is achieved by requiring doctors or dentists who wish to admin- angle of the magnetisation vector. One of the most common ways ister radioactive substances to first apply for a licence issued by to modulate the B1 field in order to fulfil the adiabatic condition the Department of Health. The applications process is handled is to use a hyperbolic secant pulse, in which the B1-field is modu- by ARSAC and they advise health ministers on the granting or lated by a hyperbolic secant function. renewal of licences. Licences for diagnosis or therapeutic uses Adiabatic radiofrequency pulses are able to achieve uniform last for 5 years; licences for research projects ordinarily last for flip angles, even if the RF field is nonuniform. This advantage can 2 years. be utilised in inversion recovery sequences such as STIR (short ARSAC was established by the Medicines (Administration of tau inversion recovery) and FLAIR (fluid attenuated inversion Radioactive Substances) Regulations 1978 and currently consists recovery), in which the pulses can be used to achieve uniform of 21 members, the majority of which are medical doctors. Medical inversion across the imaged object (Hajnal et al., 2001). physicists, radiopharmacists, radiographers and other staff groups ADP (adenosine diphosphate) 28 Adverse effects also sit on the committee. The committee can also recommend the Typical clinical examples of acute effects are radiation-induced A revoking or suspension of a previously issued licence. dermatitis, mucositis, and bone marrow depletion. Clinical late Hyperlinks: ARSAC website: www .arsac .org .uk effects include telangiectasia, fibrotic reactions (skin, lung), and in rare cases when exceeding tolerance doses, osteopathy and ADP (adenosine diphosphate) radiation myelitis. (Magnetic Resonance) See Adenosine diphosphate (ADP) Acute effects occur often in rapidly proliferating tissue (e.g. mucosa) where radiation deteriorates the balance of cell produc- Adverse effects tion in the germinative tissue layers and the cell loss following the (Nuclear Medicine) Adverse effects, also known as side effects, final differentiation into functional cells. The regulation mecha- refer to unwanted negative effects following diagnostic treat- nism of this balance is not yet fully understood. The stem cell ment or therapy. Radiopharmaceuticals may have an affinity for pool itself seems to trigger the proliferation. organs or biochemical processes other than the intended target Three different categories of interacting effects seem to be organ/process. Organs, other than target organs, that accumulate responsible for the late response of tissue: (1) in the target cell the administered radiopharmaceutical are referred to as critical model the manifestation of tissue damage depends on the char- organs. The amount of administered activity is limited by the acteristics of individual target cells (i.e. proliferation kinetics, dose to the critical organs. Following an administration three repair capacity) and tissue structure; (2) another mechanism doses are of particular interest: is the indirect or reactive effect, e.g. damage of vasculature may be followed by destruction of the parenchyma cells of an organ; (3) finally, radiation may interact at a molecular level 1. Total body dose, which is correlated to the risk of leu- via specific signalling pathways and activation of gene expres- kaemia and cancer sion to induce growth factors and certain proteases. The com- 2. Gonadal dose as a measure of the hereditary effects bination of all three mechanisms leads to the manifestation of 3. Dose to critical organs which can be several times larger tissue radiation response. In general, late effects may occur in than the total body dose. Organs, and their respective tissues with low proliferation rates like the connective tissue tissues, vary in their sensitivity to radiation and a spe- and brain. cial consideration must be given to the critical organs. In the early days of radiotherapy when conventional x-ray and Cobalt units were used, severe side effects, in particular skin Special consideration should always be taken for pregnant reactions were observed. Holthusen in his pioneering work (1936) women. The effects of fetus irradiation depend strongly on how on radiotherapy optimisation established the sigmoidal dose– far the pregnancy has progressed and the effects may manifest response curves for tumour and normal tissue which defines the as mental retardation, smaller head size and an increased risk of general strategy of radiotherapy; balancing tumour cure against carcinogenesis. the probability of adverse effects. Hence, minimising the adverse Further Reading: Hall, E. J. 2000. Radiobiology for the effects whilst achieving high tumour cure probabilities calls for Radiologist, 5th edn., Lippincott Williams & Williams, precisely conforming the dose distribution to the shape of the tar- Philadelphia, PA. pp. 218–229. get volume. Nowadays, with advanced technologies such as, 3D-CRT and Adverse effects IMRT based on image-guided individual treatment planning and (Radiation Protection) Ionising radiation causes biological delivery the occurrence of adverse effects is rare. Consideration damage at a cellular level. When such damage is expressed as of the potential induction of adverse effects is an essential part of symptoms either in the person exposed that has not been planned the radiotherapy treatment process, particularly at the planning (i.e. not associated with radiotherapy), or is seen in the exposed stage where the doses to all organs at risk are evaluated alongside person’s progeny, it may be referred to as adverse effects, or that given to the tumour. This has been aided by the availabil- adverse radiation effects. ity of dose-volume histograms (DVH) in modern computerised Related Articles: Bioeffects treatment planning systems. Many centres have correlated the incidence of late adverse events with DVH parameters to obtain Adverse effects ‘dose objectives’ for organs at risk to use as a guide for the plan- (Radiotherapy) Since any radiation treatment inevitably also ning process. All clinical trials with a radiotherapy component affects normal tissue, radiotherapy may cause radiation-induced now include some guidance on doses to organs at risk in their complications which are known as adverse or side effects. These protocols. effects depend on the general status of the patient [e.g. age, co- The significance and probability of acute adverse effects can morbidities, performance status (Karnovsky index)], the clini- hardly be predicted in the individual case, but mostly they disap- cal situation (tumour site, type, stage, classification, treatment pear shortly after the course of treatment. Initial information by scheme, combined treatment modalities, previous treatments) and the physician, appropriate behaviour and sometimes supportive on the physical and technical treatment parameters such as radia- medication may help the patient to cope with the acute adverse tion type, beam quality, target volume and in particular the spatial effects. and temporal dose distribution. Abbreviations: 3D-CRT = Three-dimensional conformal Adverse effects may occur during the treatment (acute effects, radiotherapy and IMRT = Intensity modulated radiation treatment. typically up to about 90 days after start of treatment) or months Related Articles: Probability of complications, Sigmoid dose– and years after completing the course of treatment (late effects). response curve Adverse effects range from mild effects like tiredness, mood Further Reading: Holthusen, H. 1936. Erfahrungen über swings, mild forms of nausea, light skin reactions, up to more die Verträglichkeitsgrenze für die Röntgenstrahlen und deren severe effects such as skin ulceration, and the potential induction Nutzanwendung zur Verhütung von Schäden. Strahlenther 57: of cardiovascular diseases and secondary cancer (carcinogenesis). 254–269. Adverse radiation effects 29 Agatha phantom Adverse radiation effects Related Articles: Scintillator, Scintillation detector (Radiation Protection) See Adverse effects Further Reading: Knoll, G. F. 2000. Radiation Detection and A Measurement, 3rd edn., John Wiley & Sons, Inc., New York, pp. AEC (Atomic Energy Commission) 232–240. (General) See Atomic Energy Commission (AEC) Afterloading AEC (automatic exposure control) (Radiotherapy, Brachytherapy) (Diagnostic Radiology) See Automatic exposure control (AEC) Source Handling and Loading: Brachytherapy source/s must be handled and loaded into the applicators for treatment and, over time, many methods have been used. These methods have been Affinity developed primarily to reduce the dose to the personnel, but also (Nuclear Medicine) A measure of the strength of binding of a to improve the quality of the treatment itself. ligand to another molecule. The reciprocal of affinity is called Kd Afterloading of brachytherapy sources is performed in the fol- (the equilibrium dissociation constant). Thus the higher the affin- lowing steps: ity that the ligand has for the receptor, the lower its Kd will be. 1. Applicators, needles, catheters, etc. are inserted AFOMP 2. Correct applicator positions are verified using dummy (General) The Asia-Oceania Federation of Organizations for sources Medical Physics (AFOMP) was founded in 2000 as a regional 3. The sources are inserted into the applicators organisation of IOMP. As of 2019, the federation consists of 18 national member organisations (plus three affiliated members), representing about 6500 physicists and engineers working in the Generally, afterloading techniques make it possible to devote time field of medical physics. to correct placement of applicators and verification of applicator Since its inauguration, the main objective of AFOMP has been positions. to harmonise and promote the best practice of medical physics in Related Articles: Brachytherapy, Source loading in the Asia and Oceania region. The aims and objectives of AFOMP brachytherapy, Manual loading, Manual afterloading, Remote are: afterloading, Remote afterloading unit • To promote the co-operation and communication Agatha phantom between medical physics organisations in the region (Diagnostic Radiology) The Agatha phantom was created to • To promote medical physics and related activities in the address the need for a digital breast tomosynthesis phantom for region constancy testing, as previously available phantoms were unsuit- • To promote the advancement in status and

standard of able having been designed for planar imaging. practice of the medical physics profession Agatha consists of a PMMA slab measuring 160 mm × 220 • To organise and/or sponsor international conferences mm and is 45 mm thick, thus is attenuation equivalent in the and regional and other meetings/courses mammography x-ray energy range to the 53 mm standard breast • To collaborate or affiliate with other scientific thickness. It contains two positioning aids at the chest wall, con- organisations sisting of cuboids with 70% glandular and 30% adipose tissue equivalency to ensure correct positioning. The phantom includes details to assess tumour and calcifica- AFOMP includes the medical physics societies from the follow- tion like features, with the presence of a 3 mm and 5 mm diam- ing countries: Australia, Bangladesh, China PR, China Taipei, eter sphere of 70% glandular and 30% adipose tissue equivalency China Hong Kong, India, Indonesia, Iran, Japan, South Korea, and an aluminium disc of 0.5 mm diameter, respectively. Digital Malaysia, Mongolia, Myanmar, Nepal, Philippine, Singapore, breast tomosynthesis is reliant on the smooth motion of the x-ray Thailand and Vietnam. Additionally, AFOMP has three affiliated tube in order to ensure that no breast tissue is missed out. With societies from Bangladesh, Pakistan and Malaysia. the lack of structure in breast tissue, it may not be obvious when Hyperlinks: www .afomp .org any irregularity in the movement arc results in missed tissue. The three spherical inserts assist in assessment of tomographic Afterglow motion, as well as the analysis of 3D MTF. Given the limited (Radiation Protection) The afterglow phenomenon occurs in number of projections, DBT is particularly susceptible to image luminescent materials, e.g. scintillators. The energy of ionising artefacts. These spheres also enable assessment of the artefact radiation (x-ray or gamma photons, charged particles) absorbed in spread function. the scintillator causes excitation of orbital electron energy states The phantom also contains a ‘perpendicular’ wire through- in the material. The average lifetime of these excited energy out the thickness of the phantom, allowing for measuring states tends to be ~10−8 s, before decay of the electrons back to the signal difference to noise ratio through the acquired slices. ground state occurs involving the emission of optical radiation Considering DBT is a tomographic method, a tilted tungsten (scintillation) in the visible range. This process is called fluores- wire is present to test z-direction sensitivity profiles (ZSP). cence. Sometimes the excited state is metastable and its average Additionally, a set of four nylon wires can be used for line lifetime is longer, e.g. from microseconds to hours, depending object spread function (LOSF). Finally, the phantom also on the scintillator material. The de-excitation is delayed in such contains a homogenous central area for daily verification materials and this process is called afterglow or phosphorescence. (Figure A.26). Afterglow is a source of background light (noise) in scintillation Related Articles: Mammography, Digital mammography, detectors. Tomosynthesis AI (Artificial intelligence) 30 Air gap Medical Applications: In medical physics, air equivalent A materials can be used to construct the walls and electrodes of ionisation chambers, such that the number of ionisation events is similar to that generated in a free-air ionisation chamber. Air equivalence requires both the mean mass energy absorption coefficient of the photon spectrum present and the mean mass collision stopping powers of the secondary electron spectrum present to be the same as that of air. Certain plastics are found to be suitable with a graphite surface layer to enable electrical conduction. Related Articles: Air, Free-air ionisation chamber, Plastic Air gap (Diagnostic Radiology) The air gap is space or distance added FIGURE A.26 Agatha phantom. (Image courtesy of Leeds Test Objects: www .l eedst estob jects .com/ index .php/ phant om /ag atha/ ) between a patient’s body and the x-ray image receptor as shown on Figure A.27. The air gap reduces the intensity of the scattered radiation Further Reading: Jacobs, J., N. Marshall, L. Cockmartin, reaching the receptor. This is not by filtration or absorption but F. Zanca, R. van Engen, K. Young, H. Bosmans and E. Samei. a geometric effect in which the scatter diverges at a greater rate 2010. Toward an international consensus strategy for peri- than the primary x-ray beam. Therefore, the intensity of the scat- odic quality control of digital breast tomosynthesis systems. In tered radiation is reduced when it reaches the receptor. Medical Imaging 2010: Physics of Medical Imaging (Vol. 7622, p. Air gaps are present when geometric magnification techniques 76220G). International Society for Optics and Photonics. are used, as in magnification mammography. Although magnifi- cation is used to enhance image detail, there is the added benefit AI (Artificial intelligence) of reduced scatter reaching the receptor. (General) See Artificial intelligence (AI) Air gap Air (Radiotherapy) In external beam radiotherapy, the distance (General) between the exit of the treatment unit and the patient’s skin is known as the air gap. The air gap is important for the treatment as, particularly in Molar mass 0.029 kg/mol high energy x-ray beams, it allows sparing of the dose to the skin Density at STP ~1.2 kg/m3 and hence reduced side effects, such as reddening, of the skin. The size of the air gap determines other characteristics of the dose dis- Melting point ~63 K tribution such as the output factor, particularly for electron beams. Boiling point ~77 K Related Articles: Dosimetry, External beam therapy CT number −1000 HU Air is a mixture of gases making up the Earth’s atmosphere. Dry air consists of around 78.08% nitrogen, 20.95% oxygen, 0.93% argon, 0.038% carbon dioxide, and trace amounts of other gases. Air also contains a variable amount of water vapour, approxi- mately 1%. Unfiltered air additionally includes particulates and industrial pollutants. The atmosphere acts to protect life by Primary retaining heat via the greenhouse effect, absorbing ultraviolet solar radiation and minimising temperature extremes. The atmo- sphere gradually decreases in density with distance from the Earth’s surface, and it is normally categorised into several layers. The troposphere is the lowest layer of the atmosphere extending 7–17 km from the surface and containing around 80% of the total mass of the atmosphere. The average temperature and pressure of air at sea level is 288 K and 101.3 kPa, respectively. Related Articles: Air equivalent composition, Air gap, Air kerma, Air kerma strength, Air-cored transformer, CT number, Scatter Entrance surface air kerma (ESAK), Equivalent tissue air ratio (ETAR), Free air ionisation chamber, In air calibration factor, Air gap Mean absorbed dose to air, Reference air kerma rate – RAKR, SAR (scatter air ratio), TAR (tissue air ratio) Air equivalent composition Receptor (General) A material with an air equivalent composition is designed to sufficiently mimic the chemical and physical proper- FIGURE A.27 Air gap between the table top and receptor (film). ties of air depending on the purpose it is designed for. (Courtesy of Sprawls Foundation, www .sprawls .org) Air kerma 31 AIUM Air kerma 1. Specification of contained activity (Radiation Protection) Kerma (kinetic energy released per unit a. Mass of radium; mg Ra A mass) is used to describe energy loss in a medium. Thus air kerma b. Contained activity; Ci, Bq represents the kinetic energy transferred to charged particles per 2. Specification of output unit mass of irradiated air when indirectly ionising (uncharged) a. Equivalent mass of radium; mg Ra eq radiations such as photons traverse the volume of air. b. Apparent activity Related Article: Kerma c. Reference exposure rate d. Reference air kerma rate Air kerma e. Air kerma strength (Radiotherapy) The kerma (K) is defined as the mean energy transferred from the indirectly ionising radiation to charged Task Group 43 of the AAPM defines ‘air-kerma strength’: particles in the medium per unit of mass at a point of interest ‘Air-kerma strength has units of μGy m2/h and is numerically without concern as to what happens after this transfer. As the identical to the quantity Reference Air Kerma Rate recommended energy transferred to charged particles depends on the irradiated by ICRU 38 and ICRU 60’ (ICRU 38, 1985; ICRU 60, 1998). For medium, a statement of kerma is incomplete without a reference convenience these unit combinations are denoted by the symbol to the material concerned. Air kerma results from photon interac- U where U = 1 μGy m2/h1 = 1 cGy m2/h. The National Institute tions with air. of Science and Technology (NIST) maintains the US primary air- It is worth showing a relationship of air kerma with exposure, kerma standards for x-rays in the energy range of 10–300 keV and another dosimetric quantity related with air interactions. for photon-emitting radionuclides such as 137Cs, 192Ir, 103Pd and Air kerma is related to the energy fluence by 125I. Air-kerma strength, SK, is the air-kerma rate, Kδ(d), in vacuo due to photons of energy greater than δ, at a distance d, multi- m plied by the square of this distance, d2. The distance d is measured Kai = Y tr r r along a perpendicular bisector of the source. where SK = K d 2 d ( )* d ψ is the fluence energy μtr/ρ is the mass energy transfer coefficient In modern brachytherapy dosimetry, reference air kerma rate or air kerma strength is the quantity used to calculate absorbed dose. and the exposure is given by See Source strength for a full description of specification of source strength. æ m Abbreviations: AAPM = American Association of Physicists en ö e X = Yç r ÷ è ø W in Medicine and ICRU = International Commission on Radiation air air Units and Measurements. where Related Articles: Source strength, Mass of radium, Contained ψ is the fluence energy activity, Equivalent mass of radium, Apparent activity, Reference (μ air kerma rate (RAKR) en/ρ)air is the mass energy absorption coefficient of air Wair is the average energy required to produce an ion pair in air Further Readings: Nath, R. et al. 1995. Dosimetry of inter- e is the electron charge stitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No 43. Med. Phys. therefore 22:209–234; Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hansson, M. S. Huq, G. S. Ibbott, M. S. Mitch, R. Nath and J. F. Williamsson. Update of AAPM Task Group No 43 Report: A XW K air æ mtr /r ö XW e air = = air / e ç ÷ revised AAPM protocol for brachytherapy dose calculation. Med. è men /r ø 1 - g air Phys. 33:633–674. where g is the fraction of electron energy lost in bremsstrahlung Air-cored transformer production. (Diagnostic Radiology) See Transformer The fraction (g) is significant only at high energies. Its value for 60Co photons is 0.003. AIUM Related Articles: Kerma, Collision kerma (Ultrasound) The American Institute for Ultrasound in Medicine was formed in the early 1950s to promote the emerging science Air kerma strength of medical ultrasound and its diagnostic and therapeutic applica- (Radiotherapy, Brachytherapy) Calibration of source strength tions. Since then it has developed into a leading multidisciplinary is a very important part of a comprehensive brachytherapy organisation with a membership of physicians, technologists, quality system. The instruments, ion-chambers and electrom- sonographers, scientists, physicists and engineers. The AIUM eters, used for source strength determinations, should have works with users, manufacturers and government to produce calibrations that are traceable to national and international guidelines for the safe and effective use and practice of medi- standards. cal ultrasound, many of which are available through its website. Specification of Source Strength for Photon Emitting The Institute produces the Journal of Ultrasound in Medicine and Sources: Source strength for a photon emitting source can be runs a large annual meeting and other smaller educational meet- given as a quantity describing the radioactivity contained in the ings in North America. source or as a quantity describing the output of the source: Hyperlink: AIUM: www .aium .org Alanine 32 Aliasing Alanine ALFIM includes the medical physics societies from the fol- A (General) See Alanine dosimeter lowing countries: Argentina, Brazil, Chile Colombia, Cuba, Ecuador, Mexico, Panama, Paraguay, Peru and Venezuela. Alanine dosimeter (General) Alanine is an amino acid [C3H7O2N] that can exist in Algebraic reconstruction technique (ART) two structural forms: α-alanine [CH3-CH(NH2)-COOH] and (Diagnostic Radiology) The algebraic reconstruction technique β-alanine [CH2(NH2)CH2-COOH]; α-alanine exists in three struc- (ART) is an iterative reconstruction technique used in projec- tural forms: L-α-alanine,

D-α-alanine and DL- α-alanine. All the tion reconstruction techniques such as e.g. computed tomography α-alanines form very stable free radicals when subjected to ionis- (CT). Gordon, Bender and Herman (1970) first showed its use in ing radiations, and therefore, the three forms can be used as dosim- image reconstruction by exploiting what is better known in linear eters, even if for economical cost, DL- α-alanine is the most used algebra as the Kaczmarz method. dosimeter. A non-destructive read-out of the radiation-induced Consider the system: measuring signal is performed by electron spin resonance (ESR). With the proper adjustment of the ESR-spectrometer and using p = Af dose values in the range 1 Gy–100 kGy, doses can be determined with the overall uncertainty of 2–3% at k = 2. EPR response is also where linear as a function of dose up to 10 kGy; the α-alanine dosimetric signal is not very dependent on dose rate and radiation energy. Moreover, the composition of alanine is very close to that p = ( T p1,…, pM ) of biological tissues, and therefore, it can be considered a tis- sue equivalent material. Advantages of alanine-EPR dosimetry is the acquired projection data (sinogram) arranged as a singular are that results fade slowly over a period of years, it is easy to column vector, handle, non-toxic and is available in different shapes. It also has a low-temperature coefficient of irradiation and results are highly stable and reproducible. The disadvantages are that it is sensitive f = ( T f1,…, fN ) to water, humidity and to light if illuminated for long periods. In addition, EPR spectrometers are expensive. is the image to reconstruct (arranged as a singular vector) and Related Articles: Electron spin resonance (ESR) A is the system matrix (see Iterative image reconstruction). The ART approach in its simplest form means applying iteratively the Alanine gel formula: (General) Alanine gel consists of small alanine crystals ran- domly suspended in a gel-forming agar matrix. When alanine gel k +1 k p - f + i a × k = i f f a is exposed to ionising radiations, alanine crystals trap the free × i a i ai radicals and act as a dosimeter. Electron paramagnetic resonance analysis of the relative concentrations of free radicals trapped where ∙ denotes the dot product between vectors, ai is the i-th inside the alanine determine the absorbed dose. Alanine gel can row of the system matrix A, and for each iteration k, we have i = be modified by additives to provide elemental equivalency with k mod M. Usually, regularisation parameters might be added into that of tissue. the formula and an initial guess (also, for instance, the output of Related Articles: Electron spin resonance (ESR) an FBP reconstruction) rather than zeros might be used at first iteration. Moreover, rather than exploring A sequentially, the i-th ALARA row of the system matrix to process might be chosen randomly. (Ultrasound) The principle of ALARA (as low as reasonably Related Articles: Iterative image reconstruction, Filtered back achievable), as applied in ultrasound, includes using low output projection (FBP) powers where patient safety is an issue and using lower intensity Further Reading: Gordon, R., R. Bender and G. T. Herman. modes (e.g. B-mode) before adding higher output modes such as 1970. Algebraic reconstruction techniques (ART) for three- colour flow and spectral Doppler imaging. dimensional electron microscopy and x-ray photography. J. Theor. Biol. 29(3):471–481. ALARA (Radiation Protection) See As low as reasonably achievable Algorithm (General) An algorithm is a series of well-defined mathematical ALARP operations, i.e. a calculation method. Each operation or algorithm (Radiation Protection) See As low as reasonably practicable state yields a result that may affect the choice of the subsequent stages, as in a flow chart. A simple flow chart for a scintillation ALFIM camera reconstruction is seen in Figure A.28. (General) The Association of Latin American Physicists in Medicine (ALFIM) was founded in 1985 as a regional organisa- Aliasing tion of IOMP. As of 2019, the federation consists of 11 national (Magnetic Resonance) Aliasing, ‘phase wrap’ or ‘foldover’ is an member organisations, representing about 1200 physicists and image artefact that occurs in MRI where there is anatomy outside engineers working in the field of medical physics. of the user defined field of view (FOV) in the phase encoding Since its inauguration, the main objective of ALFIM has been direction (Figure A.29). to harmonise and promote the best practice of medical physics in Application of a phase encoding gradient creates a spatial South America, Central America and the Caribbean region. variation in temporal frequency given by Aliasing 33 Aliasing After rearrangement, Raw data A 1 GPhaseMax = (A.16) Pinhole reconstruction Collimator? Parallelhole 2Dy ×T g reconstruction Over the course of the scan, the phase encoding gradient ranges OSEM MLEM FBP between +/−Gphasemax. For Np pixels in the phase encoding direc- OSEM MLEM FBP tion, the phase encoding gradient changes by 2G DG PhaseMax 2 Phase = = (A.17) Postprocessing NP Dy ×T gN p at each phase encode. The change in phase at any given location y between phase encodings is then (combining [A.12] and [A.16]) Reconstructed image D ( py j y) 2 FIGURE A.28 A flowchart describing the reconstruction process of raw = (A.18) Dy × N data acquired by a scintillation camera. p As FoV N p = (A.19) Dy The phase change between phase encodings at any location y can y N be expressed as p Df( py y) 2 = (A.20) FoV For anatomy outside of the field of view in the phase encoding (a) (b) direction, FIGURE A.29 (a) Image of a test object and (b) an aliased image of FoV the same object. In the aliased case, the field of view (FOV) in the phase y > (A.21) encoding direction is set as shown, with some of the object outside the 2 FOV. so the phase change between phase encodings (A.19) is > π. Phase changes with magnitude >π are ambiguous, so phase encoding of points at all locations Dw =2pgyGp (rad/s) (A.12) FoV y = n × n = 0,1,2,¼ where (A.22) 2 y is distance from isocentre along the direction of the applied gradient G is ambiguous and these points are mapped back into the FOV p γ is the gyromagnetic ratio causing the foldover artefact. Similarly in the frequency encoding direction, any anatomy On completion of phase encode step i of duration T, beyond the edges of the user set FOV generates signal at frequen- the variation of phase with location y in the phase encoding cies that violate Nyquist sampling requirements. However, alias- direction is ing in the frequency encoding direction can be prevented simply by low pass temporal filtering to exclude signal originating from beyond the FOV. fi (y) = 2pgyTGp(rad) (A.13) Aliasing If Gmax is the magnitude of the maximum gradient applied, the (Ultrasound) Aliasing is a well-known phenomenon to all who maximal variation of phase with position is have studied sampling theory. The Nyquist theorem states that the sampling frequency must be at least twice the highest frequency fmax(y) = 2pgyTGmax (A.14) component of the input signal, or else aliasing will occur. By this, it is meant that the reconstructed signal will appear at a lower On completion of the maximal phase encoding step, the phase frequency than it originally had. Usually this is illustrated with difference between two adjacent pixels separated by the pixel two sinus-signals, where the slower one is reconstructed from width Δy is 180°: sampling of the faster waveform at the equidistant points marked in the figure. Usually in connection with Doppler measurements, this Dfmax(Dy) = p = 2pgDyTGmax (A.15) illustration is not entirely applicable. In this limiting case, when FOV phase Alpha beta ratio 34 Alpha beta ratio the input frequency is exactly half the sampling frequency, the data from cell suspensions, animals and patient treatment A same frequency will be reconstructed. If the input frequency f0 together with the interpretation of these data by means of empiri- increases an amount ε, the reconstructed frequency will be f0−ε. cal models have provided insight into the mechanisms of radiation If that would be the case in Figure A.30, illustrating carotid flow, effects on living organisms. For instance, it turned out from the the peak velocities would appear as ‘folded’ and interfere with early radiobiology data that double strand break (DSB) of DNA the waveform. Instead they show up as negative velocities. Why is considered the most important type of cellular damage which it appears different in a Doppler sonogram is due to the fact that induces chromosome aberrations, phenotype changes and cell usually a quadrature detection is made. This results in a pair of sterilisation. Translation of this fundamental knowledge gained signals (I and Q channels), and the subsequent Fourier transfor- from first radiobiology investigations had a significant impact on mation can be thought of as performed on a complex signal with the progress of radiotherapy. The initial radiobiology experiments these as the real and imaginary parts, respectively. The result is established the relation of cell survival and radiation dose. The a nonsymmetric spectrum, with a representation of both negative typical continuously bending shape of cell survival curves has and positive velocities, as opposed to an FFT performed on a real been successfully interpreted based on the target theory which signal, which results in a symmetric spectrum. The image of the follows from Poisson-statistics. Particularly, in the low-dose spectrum of a sampled signal as one that repeats itself around region up to about 3 Gy the linear-quadratic model turned out to multiples of the sampling frequency is probably more useful in describe dose–response data best: this case. Aliasing can also be found in the same way in colour Doppler images, see Figure A.31. SF(D) = exp( - aD - bD2 ) Alpha beta ratio where SF(D) is the fraction of cells surviving a dose D. The (Radiotherapy) Radiation causes damage to cells via interaction parameters α (1/Gy) and β (1/Gy2) determine the ‘bendiness’ of with molecules, most importantly with the DNA. Experimental a survival curve (see Figure A.32). The ratio α/β has the unit Gy, and in a semi-log plot of SF(D) it is the dose where both the lin- ear and the quadratic components of the survival curve are equal. Many mechanistic interpretations of the α and β parameters have been suggested, for instance the linear component was claimed to represent single-track events whereas the quadratic part might describe two-track events. Curtis (1986) in his unified repair model of cell killing proposed two different causes that are ulti- mately responsible for cell kill: the reparable and the nonrepara- ble lesions. The nonreparable lesions such as DSB, or in terms of the target theory the single-hit lethal effects, are associated with the linear term exp(−αD) of cell killing, whereas the repair itself is a balance of successfully repaired events and binary misrepair, the latter giving rise to the exp(−βD2) term. However, all these mechanistic approaches to understand the cell survival became to some extent obsolete, particularly in the light of molecular radiobiology. In the clinical environment the α/β-ratio is widely used to assess the importance of acute and late effects in normal tissue. High α/β-values (usually 10), i.e. nearly straight survival curves, FIGURE A.30 Aliasing in a spectral Doppler display using pulsed represent tissues with limited potential for repair from radiation Doppler. damage. On the other hand, normal tissues with low α/β-values (usually 3), i.e. with a significant shoulder of the survival curve at low doses, are characterised by their high recovery potential. 1 αD 0.5 βD2 0.1 0.05 α/β Dose (Gy) FIGURE A.31 Aliasing in a colour Doppler image. FIGURE A.32 Schematic survival curve with definition of the α/β-ratio. Surviving fraction Alpha emission 35 A lternating current (AC) Hence, tissues with low α/β-ratios are sensitive to the dose frac- ionising helium atoms to produce alpha particles and accelerating tionation scheme (fractionation effect) whereas those with high them in a cyclotron. A α/β-ratios are hardly affected by dose fractionation. Furthermore, Related Articles: Alpha particle emitter, Radioactivity, the α/β-ratio is clinically applied to calculate isoeffective doses Radioactive decay. when changing the fractionation schedule. For example, replacing a fraction dose d1 by d2, the total dose for the new fractionation Alpha radiation scheme is

(General) Alpha radiation is composed of alpha particles, emit- ted from radionuclides undergoing radioactive decay through the (a/b + d decay process normally referred to as ‘alpha decay’. Alpha radia- D = 1 ) 2 D1 (a/b + d 2 ) tion can also be generated by ionising helium atoms to produce alpha particles and accelerating them in a cyclotron. As many normal tissue tolerance values refer to a standard frac- Alpha particles are charged particles, with a charge of two tion dose of 2 Gy the formula can be used to assess the radiation and a mass of four, which consist of two protons and two neu- response when applying an alternate treatment regime. trons bound together. Alpha particles are identical to the nuclei of Abbreviations: DSB = Double strand break and SF = helium atoms ( 4 2He), and either the symbol α (first letter of Greek Surviving fraction alphabet) or He2+ can be used to denote an alpha particle. Related Articles: Linear-quadratic model, Radiobiological Related Articles: Alpha decay, Alpha particles, Radioactive models, Adverse radiation effects, Biological effective dose, decay, Radionuclide. Surviving fraction Further Readings: Curtis, K. H. 1986. Lethal and poten- Alternating current (AC) tially lethal lesions induced by radiation: A unified repair model. (General) Alternating current is a form of electric current in Radiat. Res. 106:252–270; Podgorsak, E. B. 2003. Review of which the direction of flow changes or alternates regularly. This Radiation Oncology Physics: A Handbook for Teachers and distinguishes it from ‘direct current (DC)’, where the current Students, International Atomic Energy Agency, Vienna, Austria; flows in one direction and is typically constant in value. Steel, G. G. 2002. Basic Clinical Radiobiology. Published by Alternating current is easy to generate, and AC power may be Arnold, London and co-published in the United States by Oxford transferred efficiently using transformers. Most domestic electri- University Press. cal power is AC power, typically supplied at a potential of 110– 230 V RMS. Alpha emission Alternating current sources are usually sinusoidal in form and (Nuclear Medicine) In α-decay an α-particle, consisting of produced at a frequency of 50–60 Hz (Figure A.33), though the two neutrons and two protons, is emitted from a nucleus. The term AC and the theory of AC conduction apply to currents and α-particle has high kinetic energy, typically 4–8 MeV, but voltages at any frequency. due to frequent interactions with the surrounding material It is necessary to define AC voltage and current values in terms the range is limited to as little as a few μm in solid materials. of their DC equivalents, so that the power provided into a resistive Therefore most α-emitters are never used for clinical imaging. load will be the same. This is defined as the RMS or ‘root-mean- On the other hand, α-particles are very efficient in inducing cell square’ AC value, and for sinusoidal signals is 1/√2 of their peak death and are therefore an option when designing therapeutic values. methods. The electrical power (in watts) in a DC circuit can be calcu- lated simply by multiplying the potential difference across a load Alpha particle emitter (in volts) by the current through the load (in amps). Similarly, the (Radiation Protection) An alpha particle emitter is any element apparent power in an AC circuit can be deduced using the Vrms that emits alpha particles as a result of radioactive decay. Alpha and Irms values. emitters either occur naturally – there are three series of natu- However, true AC electrical power calculations are more com- rally occurring substances (uranium, thorium and actinium) – or plex as the AC current does not necessarily flow ‘in phase’ with can be created by bombarding specific elements with high-energy the applied potential. particles or by fusion induced by a neutron in a nuclear reactor (or nuclear device). Related Articles: Alpha particles, Nuclear fusion AC current and potential waveforms 400 Alpha particles (General) Alpha particles are charge particles, with a charge of 200 two and a mass of four, which consist of two protons and two neu- trons bound together. Alpha particles are identical to the nuclei of 0 helium atoms ( 4 2He), and either the symbol α (first letter of Greek 1 4 7 10 13 16 19 22 25 28 31 34 37 40 alphabet) or He2+ can be used to denote an alpha particle. –200 Alpha particles were discovered by Ernest Rutherford (1871– 1937) in 1899 and are highly ionising due to their mass and double –400 charge. Time in milliseconds Mass 6.644656 × 10−27, equivalent to 3.727738 GeV. Alpha particles are emitted in some forms of radioactive decay Potential in Volts Current in Amps RMS value (alpha decay) where the radionuclide has an excess of neutrons and protons. Beams of alpha particles can also be produced by FIGURE A.33 Typical alternating current and voltage signals. AC potential Alternating voltage 36 AMBER The relative phases of the AC potential and current must also Aluminium A be taken into account where a load possesses reactive properties (General) (capacitance or inductance). In such cases the power delivered to the load is given by Symbol Al ACpower = ACvoltagerms * ACcurrentrms * cos(phaseangle) Element category Metal Mass number A 27 where the phase angle represents the difference in phase between Atomic number Z 13 the voltage and current. Atomic weight 26.9815 g/mol Related Articles: Alternating voltage, Apparent power, Direct Electronic configuration 1s2 2s2 2p6 3s2 3p1 current, Direct voltage Melting point 933.47 K Boiling point 2792 K Alternating voltage Density near room temperature 2.70 g/cm3 (General) A form of electrical potential which changes polarity regularly, and is usually sinusoidal in amplitude. This distin- guishes it from ‘a direct voltage’ where the potential remains at a History: Aluminium salts were used in Ancient Greek and constant fixed value. Roman medicine for wound dressing. Aluminium was first Alternating voltages cause alternating currents (AC). Most extracted in pure metal form by Friedrich Wöhler in 1827, by domestic electrical power is AC power, typically supplied at a reacting potassium and anhydrous aluminium chloride. potential of 110–230 V RMS with a frequency of 50–60 Hz. Although aluminium is the official international spelling, alu- Alternating voltages may be specified in terms of their peak- minium is widely used in the United States and Canada, and is to-peak voltage swing (Vpk-pk), the peak swing (Vpk), or by the accepted as an alternative spelling by many chemical societies. rms or root-mean-square voltage (Vrms) which has the equivalent Isotopes of Aluminium: Pure aluminium is highly reactive power capability as the DC voltage of the same value. and is therefore uncommon in nature. However, it is abundant in For a sinusoidal alternating voltage: Vpk-pk = 2 · Vpk = 2 · √2 · Vrms compound with other molecules, particularly as oxides and sili- Abbreviations: AC = Alternating current, DC = Direct current cates. Stable 27Al is by far the most common naturally occurring and RMS = Root-mean-square. isotope, with more than 99.9% relative abundance. The remain- Related Articles: Alternating current, Apparent power, Direct der consists of radioactive 26Al, produced in the atmosphere by current, Direct voltage cosmic ray photons reacting with argon. There are seven further isotopes, with mass numbers between 23 and 30, but these do not Alumina occur naturally and must be synthesised. (Nuclear Medicine) Alumina (Aluminium oxide) is used in Medical Applications: Solid-state lasers – Aluminium forms technetium generators. In the generator, molybdenum-99 is in part of the yttrium-aluminium-garnet (YAG) crystal that is used the form of MoO 2- 4 ions that are bound to alumina on columns. in many solid-state medical lasers. This crystal can be doped with Technetium-99m can easily be separated from its parent nucleus neodymium (Nd-YAG), ermium (Er-YAG) or other rare earth ele- molybdenum-99, because of the different chemical properties of ments, to produce light of different wavelengths. Nd-YAG and the two elements. Molybdenum-99 decays to technetium-99m, Er-YAG lasers (with wavelengths of 1064 and 2940 nm, respec- which is then found as pertechnetate ions TcO - 4 . Molybdenum-99 tively) have many uses in medicine, including ophthalmology, has an affinity to the alumina of the columns, while pertechnetate dentistry and cosmetic corrective treatments. does not bind to the alumina. Radiation shielding: Aluminium is traditionally used to atten- Aluminium ions (Al3+) might be washed out during elution, uate ionising radiation (particularly beta radiation), both for radia- which is often referred to as ‘aluminium breakthrough’. While it tion protection purposes and in the shielding of equipment from is a rare event, excessive aluminium in the eluate is an indication external radiation that may affect its performance. The shielding for a problem with the generator. A common test for assessing the performance of any material can be expressed in terms of ‘mm Al’, concentration of Al3+ in the eluate is the use of a colourimetric test the thickness of aluminium that would be required to provide the kit. A drop of a standard solution of Al is placed on an indicator equivalent reduction in radiation flux. strip, together with a drop of the eluate. Comparing the intensity of the two drops, the concentration of Al3+ in the eluate can be Aluminium equivalent qualitatively estimated to smaller or larger than the concentration (Diagnostic Radiology) Both the inherent and total filtration of the standard solution. through which an x-ray beam passes can be expressed in mm Related Articles: Technetium generator, Aluminium of aluminium. This, the aluminium equivalent, is the thick- Further Reading: Zolle. 2007. Technetium-99m Pharma- ness of aluminium that would provide the same absorption/ ceuticals: Preparation and Quality Control in Nuclear Medicine, filtration. Springer, Berlin Heidelberg. Related Articles: Filtration inherent, Filtration total Alum in film processing AMBER (Diagnostic Radiology) Alum is aluminium salt. Such salt is used (Diagnostic Radiology) Advanced multiple-beam equalisation in x-ray film processing – especially as hardener (e.g. potassium radiography (AMBER) is a technique developed by Kodak to alum). The aluminium salt prevents excessive softening of the reduce the problem of the wide range of exposures coming from emulsion (which may damage it during the washing or drying a patient’s body, especially for thoracic imaging. In the chest the processes). low density lung areas produce high exposures to the image recep- Related Articles: Hardening agent, Developer tor and the more dense mediastinum produces low exposures. The Ambient lighting 37 Amorphous selenium Electronic configuration 1s2 2s2 2p6 3s2 3p6 3d10 4s2 4p6 5s2 4d10 5p6 4f14 5d10 6s2 6p6 5f7 7s2 A Melting point 1449 K Boiling point 2880 K Density near room temperature 12000 kg/m3 (12 g/cm3) History: Americium is an artificial metal that was first obtained by Glenn Seaborg and colleagues in 1944. The team, who were based at the University of Chicago, created the isotope 241Am by bombarding plutonium with neutrons. Named for the Americas, Americium is now used worldwide in smoke detectors and as a neutron source in industrial moisture gauges. Isotopes of Americium: All of americium’s isotopes (237Am through to 246Am) are radioactive. The isotope of interest in medi- cine is 241Am. Isotope of Americium 241Am Half-life 432.2 years FIGURE A.34 An illustration of the segments and the modulator set- Mode of decay α, plus γ in ≈ 80% of decays tings for several anatomical regions. Maximum decay energy, Emax α: 5.638 MeV, γ: 13.9 keV or 59.5 keV problem is that this range of exposure can exceed the optimum contrast producing range, or latitude, of radiographic film. The decay product, neptunium-237, is also radioactive with a two The AMBER system uses a scanning slit x-ray beam which million year half-life. is divided into 21 beam segments. The beam intensity in each Medical Applications: Fluorescence thyroid scanning – exter- segment is modulated based on the intensity of the beam pen- nal sources of 241Am can be used to bombard stable iodine in the etrating the patient’s body in that area and measured by indi- thyroid with gamma rays, causing it to fluoresce. The fluorescence vidual x-ray detectors. The output from each detector controls photons released quantitatively correlate with the iodine content the corresponding modulator. The AMBER principle and some of the imaged tissue, such that a plot of the thyroid’s iodine content modulator settings for different anatomical areas are shown in can be produced (low iodine content can be an indicator of thyroid Figure A.34. cancer). As this technique involves minimal radiation exposure,

it For areas of high intensity, such as in the lungs, the beam may prove particularly useful in the diagnosis of paediatric/preg- modulator attenuates more of the beam and reduces the intensity nant patients, but unfortunately it is not widely available as yet. reaching the receptor. The result is a reduced range of exposures Gynaecological brachytherapy – In the past 241Am intracavi- reaching the receptor from the different areas of the chest. tary applicators were used for the treatment of gynaecological cancer, but now 192Ir and 137Cs applicators are more usual (their shorter half-lives allow for easier disposal and reduced security Ambient lighting risk). (Diagnostic Radiology) The level of ambient lighting or illumina- Related Articles: Fluorescence, Brachytherapy, Brachytherapy tion in a room can have an effect on the visibility of contrast in sources displayed digital images. Some display devices or monitors have photocells which measure the light in the room and change the brightness of the monitor to produce optimum viewing condi- A-mode tions. AAPM reports some typical ambient lighting levels for a (Ultrasound) The amplitude mode (A-mode) is the most basic number of image display devices (e.g. typical 2–10 lux for x-ray form to display ultrasound pulse-echo measurements. The echoes diagnostic reading stations). detected by a transducer in a single line measurement are dis- Further Reading: AAPM On-Line Report No. 03 (2005), http: played on an oscilloscope with the echo-amplitudes on the y-axis / /www .aapm .org/ pubs/ repor ts /OR _03 .p df (accessed 31 July 2012). and the time delay (depth) on the x-axis (Figure A.35). A-mode measurements were common on early ultrasound Americium systems. Nowadays the technique is used for detection of sinus (General) infection and when measuring eyeball length and skin thick- ness. In some systems, the operator can select the speed of sound appropriate to the tissue under investigation to optimise distance measurement accuracy. Symbol Am Related Articles: B-mode, M-mode Element category Actinide Mass number A 241 and 243 (no stable isotope Amorphous selenium known) (Diagnostic Radiology) Selenium (Se) is a nonmetal chemical ele- Atomic number Z 95 ment with an atomic number (Z) of 34. It is found in rare minerals Atomic weight 241 and 243 such as crooksite and clausthalite but is generally obtained as a Amorphous selenium photoconductive layer 38 Amorphous selenium photoconductive layer Transducer Kasap, S. O., M. Z. Kabir and J. A. Rowlands. 2006. Recent A advances in X-ray photoconductors for direct conversion X-ray image detectors. Curr. Appl. Phys. 6:288–292; Rowlands, J. A. and J. Yorkston. 2000. Flat panel detectors for digital radiog- raphy. In: Handbook of Medical Imaging, vol. 1, Physics and Psychophysics, eds., J. Beutel, H. L. Kundel and R. L. Van Metter, Received signal SPIE Press, Bellingham, WA, pp. 223–313. Time Time/distance Amorphous selenium photoconductive layer of pulse (Diagnostic Radiology) Selenium (Se) is a non-metal chemical transmission element of atomic number 34. It is found in rare elements such as crooksite and clausthalite and is generally obtained from the FIGURE A.35 Principle of A-mode display. The display shows the anode metal in electrolytic copper refineries. amplitude of echoes and the distance between them. Distance assumes Selenium is used as a photoconductor to detect x-rays in known speed of sound in tissue. flat-panel detectors in medical imaging. It is the most widely employed photoconductor in flat-panel direct-conversion x-ray imaging as it is easily manufactured into a large area continu- by-product from the anode metal in electrolytic copper refiner- ous photoconductor with low dark current and high x-ray sensi- ies. It exhibits both photoconductive and photovoltaic properties; tivity, with the detector formed by electronically connecting the in the former the number of charge carriers increases when it photoconductor, an active matrix array, that stores and reads out absorbs certain wavelengths of electromagnetic radiation increas- the signal. In other industries selenium has been widely used as ing the electrical conductivity, in the latter incident light generates a photoconductor in photocopiers, in a traditional x-ray imaging a voltage across the material. technique called xeroradiography. In medical imaging selenium is used as a photoconductor in Figure A.36 shows the electronic band structure of a pho- flat-panel detectors as it is easily manufactured into a large area toconductor. The valence band immediately below the forbid- continuous photoconductor layer with low dark current and high den energy gap is almost completely full while the conduction x-ray sensitivity. It is the most widely employed photoconductor band is usually completely empty. When photons of a large in flat-panel direct-conversion x-ray imaging. In other industries enough energy are absorbed by the photoconductive material, selenium has been widely used as a photoconductor in photo- a bound electron in the valence band is given enough energy copiers, in a traditional x-ray imaging technique called xerora- to move to the conductive band across the forbidden region. diography, and as a photovoltaic material in solar cells. Although This creates an electron–hole pair, in which the electron can selenium is a very useful photoconductor it has many other appli- freely move in the conductive band and the electron ‘hole’ cre- cations including, converting alternating current to direct current ated in the valence band moves through the material similar to in a rectifier and as a decolourising green glass and producing that of a physical charged particle. The increased number of ruby coloured glass. charge carriers reduces electronic resistance and increase the The structure of amorphous selenium (a-Se) is described by conductivity. the random chain model as a twofold coordinated chain structure Table A.3 lists several properties of stabilised amorphous sele- in which the angle between two adjacent bonding planes φ (dihe- nium. X-ray sensitivity of a photoconductor can be defined as the dral angle) is constant in magnitude but randomly changes sign. charge collected per unit incident radiation per unit device area. This creates regions of ring-like and chain-like structures which Two factors affect the sensitivity: (1) the radiation energy W± are randomly distributed throughout the material. This compares absorbed by the medium to create a free electron and hole pair; to crystalline selenium which occurs in one of two forms, either in and (2) the x-ray absorption coefficient α of the material. From rings of 8 atoms Se8 (α-monoclinic or α-Se), or chains of n atoms the absorption coefficient the absorption depth δ is defined as 1/α, Sen (trigonal or γ-Se). which is the depth at which 63% of the incident x-rays are absorbed Within imaging detectors a-Se is usually used in a stabilised form, this can be produced by alloying it with 0.2%–0.5% arse- nic (As) and doping with chlorine (Cl) in 10–20 ppm range. The arsenic is introduced to prevent the structure from re-crystallising Conduction band while the Cl compensates for the hole traps introduced by the As. Electron The amorphous nature of selenium is advantageous in the manu- facture of flat-panel detectors as it can be easily deposited on a suitable substrate by conventional vacuum deposition techniques to form large area photoconductive film of thicknesses up to 1000 μm. This is in contrast to polycrystalline structures which are Eg Forbidden gap X-ray excitation difficult to grow large enough to cover large area detectors (e.g. 40 × 40 cm). Related Articles: Stabilised a-Se, a-Se photoconductive layer, TFT (thin film technology), Flat-panel detector, Xeroradiography. Further Readings: Belev, G. and S. O. Kasap. 2004. Hole Amorphous selenium as an X-ray photoconductor. J. Non-Cryst. Solids 345:484–488; Kasap, S. O., C. Haugen, M. Nesdoly and J. Valence band A. Rowlands. 2000. Properties of a-Se for use in flat panel X-ray image detectors. J. Non-Cryst. Solids 266–269(part 2):1163–1167; FIGURE A.36 An energy level diagram for a photoconductor. Ampere 39 A mplitude attenuation coefficient describe the charge of the electrons, emitted by the cathode (ther- TABLE A.3 mal electrons) and bombarding the target of the x-ray tube: A Properties of Stabilised Amorphous Selenium Value Units 1A *1s = (1C/1s)*1s = 1C(here 1A = 1C/1s) Density (d) 4.3 g/cm3 Absorption depth (δ) for mean x-ray energy 20 KeV 48 μm This way, the quantity of x-rays is directly proportional to the Absorption depth (δ) for mean x-ray energy 60 KeV 976 μm quantity of charge of the thermal electrons (Q = mA × s) – the Forbidden gap energy (Eg) 2.22 eV anode current (mA) and the time of the exposure (s). The mA-s is measured as the anode current integrated over the W± for an applied external electric field of 10 V/μm 45 eV time of the exposure (using an integrator). Usually a 25% change W± for an applied external electric field of 20 V/μm 20 eV of the mAs presents an x-ray film image with a clearly visible Resistivity (ρ) 1014–1015 Ω cm change in contrast (optical density). Such a change is often related to one exposure point (such points are still used in some of the exposure tables used in the radiography practice). Related Article: mAs selector by the photoconductor. Amorphous selenium is favoured as a pho- toconductor in flat-panel detectors as it has a high x-ray absorption coefficient due to its high atomic number, better transport prop- Amplification factor erties for electrons and holes compared to other selenium-based (Nuclear Medicine) The amplification factor is the signal multi- compounds and low-dark current due to a large energy gap, Eg. plication factor in a photomultiplier (PM) tube. Two properties As selenium is used in its amorphous state it can easily be determine the amplification factor, namely the number of dynodes deposited onto the active matrix substrate, used in flat-panel and multiplication factor at each dynode. The signal multiplica- detectors, by conventional vacuum deposition techniques at low tion factor in a PM tube with 10 dynodes and electron multiplica- temperatures of below 70°C. This is in contrast to crystalline tion factor of 6 at each dynode is ~6 × 107. It is important that the materials, which require advanced and thus expensive manufac- amplification factor is constant and not dependent on the amount turing techniques and high temperatures for both annealing and of energy deposited in the photon interaction. optimisation, which may thermally damage the substrate below. Related Article: Photomultiplier (PM) tubes Although amorphous selenium is the most widely used pho- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. toconductor for flat-panel technology, at present several other Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, alternatives have been investigated. Amorphous silicon (a-Si:H) Philadelphia, PA, pp. 101–102. has a lower atomic number than a-Se (Z = 14) which means that the absorption coefficient is much smaller than selenium and Amplifier consequently a thicker layer would be required for the similar (Nuclear Medicine) A nuclear pulse amplifier is used to convert absorption properties making fabrication much more difficult. a low-amplitude pulse from a radiation detector to one with suf- Polycrystalline photoconductors such as Te and PdI2 have a ficient amplitude and the proper pulse shape to drive the pulse- higher atomic number and have been manufactured in the crys- selecting elements of the counting system. The amplifier must tal sizes required. However, their grain boundaries limit charge have enough gain to drive the pulse selector while still enabling transportation. Growth of single crystal materials is currently not the detector to operate in its most favourable operating range. Its suitable for flat-panel detectors as they have only been manufac- gain must be stable. tured in diameters of 5–10 cm and are too small for large-area The capacitance introduced by a signal cable running from detection. the radiation detector to the amplifier will in many cases signifi- Related Articles: Stabilised a-Se, Amorphous selenium, cantly attenuate and distort the transmitted electrical impulse. Selenium detector, TFT (thin film technology) Preamplifiers placed near the detector will minimise this input Further Readings: Belev, G. and S. O. Kasap. 2004. capacity and may be designed to provide impedance matching at Amorphous selenium as an X-ray photoconductor. J. Non-Cryst. their output so that long cables can be used. Solids 345–346:484–488; Kasap, S. O., M. Z. Kabir and J. A. In any modern detectors the signal after the preamplifier is Rowlands. 2006. Recent advances in X-ray photoconductors digitised for further processing. for direct conversion X-ray image detectors. Curr. Appl. Phys. Further Reading: Orvis, A. L. 1967. Systems for data accumu- 6:288–292. lation and presentation. In: Instrumentation in Nuclear Medicine, ed., G. Hine, Academic Press, New York, pp. 119–161. Ampere

(General) Ampere is the unit for electric current. It is one of the Amplitude attenuation coefficient seven SI base units from which all other units can be derived. (Ultrasound) When sound travels through a medium, its intensity Consider a setup with two parallel infinitely long conductors sep- diminishes with distance. Signal amplitude is reduced not only by arated by 1 m. When a current is applied to the two conductors, the spreading of the wave but also by scattering and absorption. they will assert a force on one another. With these settings 1 A The combined effect of scattering and absorption is called attenu- is defined as the current needed to produce a force between the ation. Ultrasonic amplitude attenuation is the decay rate of the conductors equal to 2 × 10−7 Newton per metre of length (N/m). wave as it propagates through material. The amplitude change of a decaying plane wave can be expressed as Ampere-second (Diagnostic Radiology) The ampere-second term is used in diag- nostic radiology (mainly as milliampere-second, or mA-s) to A = A0 * exp( - a * z) Analogue image 40 Analyser-based imaging where analogue signal, which has a continuous range of possible voltage A A0 is the unattenuated amplitude of the propagating wave at values, to a digital pulse which can be sorted into one of a finite some location number of energy windows, or channels. This converter circuitry A is the reduced amplitude after the wave has travelled a dis- is called an analogue-to-digital converter (ADC). If the height tance z from that initial location of the analogue pulses is in the range of 0–10 V, a 1000 chan- The quantity α is the attenuation coefficient of the wave travel- nel analyser would divide the voltage evenly over each channel. ling in the z-direction For example, channel 1 corresponds to 0–0.01 V, channel 2 to 0.01–0.02 V and so on. For every channel, the MCA has a spe- Analogue image cific memory storage location, so that each pulse that is regis- (General) An image where the darkness varies continuously with tered within a certain energy window is recorded. There are two the irradiation intensity. An analogue image is also continuous types of ADC commonly used in nuclear medicine; these are the in space, i.e. no pixels. The opposite of the analogue image is the Wilkinson converter and the successive approximation converter. digital image consisting of several image elements, i.e. pixels, and Further Reading: Cherry, S. R., J. A. Sorenson and M. E. each pixel is associated with a discrete number proportional to its Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, corresponding signal intensity. Philadelphia, PA, pp. 117–121. In early radiography, films covered with light-sensitive emul- sion (silver compound) were used to acquire images. The limiting Analyser-based imaging factor for the spatial resolution in a conventional film is the physi- (Diagnostic Radiology) Analyser-based imaging (ABI) is a cal spread of each individual silver compound. The highest pos- phase-contrast imaging technique that relies on an analyser crys- sible amount of darkness produced depends on the density of the tal placed between the sample and the detector. The analyser acts silver compound and the greyscale varies continuously. as a narrow angular band-pass filter, which is sensitive to angular deviations in the order of some microradians (1 microradian cor- Analogue signal responds to a 1 mm detail seen at a distance of 1 km). Thanks to (General) An analogue signal is a variable signal with continuity this high-angular sensitivity, x-rays deviated in the microradians in time and amplitude rather than a pulsed or discrete nature. A range due to the refractive properties of the sample have differ- variation in the signal is a representation of another time vary- ent probabilities to be collected by the detector compared with ing quantity. In electronics physical properties measured are volt- the undeviated x-rays. This translates into measurable intensity age, phase, frequency and charge. A disadvantage with analogue modulations on the detector, providing extra contrast in addition signalling is that most systems are typically sensitive to noise. to x-ray attenuation. When an analogue signal is copied and re-copied or transmitted A typical ABI setup implementation (see Figure A.37) requires over long distances the noise increases which eventually renders more crystals upstream from the sample for beam preparation in the signal useless, overshadowed by the noise. A digital signal addition to the analyser. In synchrotron experiments, these are is much more resistant to noise compared to an analogue signal. usually provided by double crystal monochromators ensuring col- Another way to convey an analogue signal is to use modulations limation and monochromaticity. The intensity at the detector can of the signal. A signal, typically a sinusoidal carrier wave, has one be modulated by displacing the diffraction plane angle of the anal- of its properties modulated, e.g. amplitude or frequency. yser with respect to the monochromator: this characteristic curve, referred to as rocking curve, expresses the shape of the angular Analogue tracer filter and it is key in the image formation process. The rocking (Nuclear Medicine) Analogue tracers are compounds created to curve is a bell-shaped function resulting from the convolution of mimic the properties of natural compounds. An analogue tracer the intrinsic reflectivity curves of both monochromator and analy- can be tailored so that the tracer only takes part in selected com- ser crystals, plus the contribution of the beam divergence. ponents of a biological process. This can minimise the number of In its simplest form, the intensity reaching the image plane in variables in a process, thus increasing the specificity and accu- an ABI application can be written as: racy of the measurements. Another use of an analogue tracer is to tailor a compound that I(q0; x, y) = IR(x, y)R(q0 + DqR(x, y)) it can be labelled with an element (radioisotope) that is not present in the natural biological compound so that the tailored compound Where R is the rocking curve, θ0 is the angular displacement can be used to study the behaviour of the natural compound. between monochromator and analyser or working point, ΔθR is The biochemical properties of an analogue tracer are not the refraction angle due to the sample and IR is called apparent always identical to the naturally occurring compounds. To com- absorption intensity and accounts for the sample absorption and pensate for this one uses correction factors. scattering to angles larger than the rocking curve width. The pre- One of the most common analogue tracers is FDG (18F-2- vious equation emphasises how the working point θ0 affects the fluoro-2-deoxy-d-glucose) which measures glucose metabolism. image contrast due to refraction. If θ0 = 0, i.e. the working point is Related Article: Isotope effect the top of the rocking curve, the refracted photons will fall some- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. where in the flanks of the rocking curve, hence they will have Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, a lower probability of being reflected, producing a reduction in Philadelphia, PA, p. 380. the detected intensity. Conversely, if θ0 is such that the working point falls in one slope of the rocking curve, the photons refracted Analogue-to-digital converter (ADC) to angles towards the top (tail) of the rocking curve will have a (General) A problem in conventional pulse height analysis is higher (lower) probability of being reflected, thus producing a when some applications require a simultaneous acquisition and positive (negative) image contrast (see Figure A.38). separation of events into different voltage or energy windows. In a In ABI, images collected at different working points on multi channel analyser (MCA), the pulses are transferred from an the rocking curve can be combined to selectively highlight Analyser-based imaging 41 Analyser-based imaging A FIGURE A.37 Sketch of a typical ABI setup. The rocking curve is obtained with no sample in the beam. When the sample is put onto the beam, the rocking curve selectively reflects x-rays according to their deviation angle ΔθR(x,y) upon exiting the sample. FIGURE A.38 Sketch of three working positions corresponding to the top and the two slopes of the rocking curve. different physical effects. In particular, the diffraction- related to sub-pixel scale features within the sample, by evaluat- enhanced imaging (DEI) algorithm combines the two images ing effects of the sample on the width of the rocking curve. acquired on both slopes of the rocking curve and yields sepa- Related Articles: Phase-contrast imaging, Diffraction- rately the (apparent) absorption IR(x,y) and refraction angle enhanced imaging ΔθR(x,y) maps of the imaged object. Since the refraction angle Further Readings: Chapman, D. et al. 1997. Diffraction is directly proportional to the gradient of the sample-induced enhanced x-ray imaging. Phys. Med. Biol. 42(11):2015; Endrizzi, phase shift, ABI is sometimes referred to as a differential Marco. 2018. X-ray phase-contrast imaging. Nucl. Instrum. phase-contrast imaging technique, while ΔθR(x,y) is called a Methods Phys. Res. A 878:88–98; Olivo, A. and E. Castelli. X-ray differential phase map. phase contrast imaging: From synchrotrons to conventional sources. Besides absorption and refraction, the development of these Rivista del nuovo cimento 37(9):467–508; Rigon, L. 2014. X-ray algorithms soon leads to the possibility of also quantifying the imaging with coherent sources. In Brahme A. (ed.) Comprehensive effect of the ultra-small-angle-scattering (USAXS), which is Biomedical Physics 2, Elsevier, Amsterdam, pp. 193–220. Anatomical body planes 42 Anatomical relationships Anatomical body planes A (General) To describe anatomical planes imagine a person stand- ing in an upright position and dividing this person with imaginary 1 vertical and horizontal planes. Anatomical planes can be used to describe a body part or an entire body. Sagittal Plane: Picture a vertical plane that runs through the Anatomical noise body from front to back or back to front. This plane divides the 0.5 Quantum noise body into right and left regions. This plane will pass approxi- mately through the sagittal suture of the skull, and hence, any plane parallel to it is termed a sagittal plane. Coronal Plane: Think of a vertical plane that runs through the 0 centre of your body from side to side at right angles to the sagittal Log (radiation dose) arbitrary units plane. This plane divides the body into front (anterior) and back (posterior) regions. This plane runs through the central part of the coronal suture or through a line parallel to it; such a plane is FIGURE A.39 Relative impact of anatomical and quantum noise upon the detection of anatomy by a human observer. (Adapted from Månsson, known as a coronal plane. L. G. et al., Rad. Prot. Dos., 114(1–3), 298, 2005.) Axial or Transverse Plane: This horizontal plane divides the body into the upper (superior) and lower (inferior) regions by run- ning through the midsection of the body. Although the efficiency and performance of the system can be modelled quantitatively by the modulation transfer function Anatomical landmark (MTF), noise power spectrum (NPS) and detective quantum effi- (Radiotherapy) Anatomical landmarks refer to distinctive struc- ciency (DQE), the detection of pathology by a human observer tures in the patient’s anatomy that may be easily localised. relies upon the total noise which they perceive. Consequently, Anatomical landmarks are generally classifiable as rigid and there is much debate over the best way to assess radiological mobile. Rigid landmarks are those which do not move as the system performance and whether to include qualitative observer- patient breathes and include bony structures. Mobile landmarks based assessment to allow the inclusion of anatomical noise include bronchi which may be used to determine how anatomy sources. Such qualitative assessments include receiver operating moves when the patient breathes. characteristics (ROC) analysis and visual graded analysis (VGA) Treatment Setup: Landmarks for setup for treatment may and visual graded characteristics (VGC). include Related Articles: Quantum noise, Noise power spectrum (NPS), DQE (detective quantum efficiency) 1. External points on the skeleton, which may be used for Further Reading: Månsson, L. G., M. Båth and S. Mattsson. basic treatment set-up with tattoos and lasers 2005. Priorities in image optimisation of medical x-ray imaging: 2. Internal bony structure, which may be used with x-ray A contribution to the debate. Rad. Prot. Dos. 114(1–3):298–302. imaging for setup 3. Internal soft tissue, which may reveal the extent of Anatomical reference point motion of soft tissue relative to bony anatomy (Radiotherapy) The anatomical reference point is an

anatomical landmark used for treatment set-up, often using imaging of inter- Image Registration: Often a multi-modality approach is used nal anatomy. It is often chosen to be stable relative to the position in medical imaging, in order to yield extra information compared of the treatment target, e.g. in treatment of the head and neck, to a single modality. The images are often registered spatially. setup may be achieved using x-ray imaging of bony anatomy. The This often involves identifying anatomical landmarks in each anatomic reference point would be a vertebral body close to the image, calculating the transformation needed to register them and tumour rather than the jaw, which is a mobile structure. interpolating/extrapolating for other regions. Related Article: Anatomical landmark Related Articles: Imaging, Multi-modality imaging, Laser localisers Anatomical relationships (General) Directional anatomical terms describe the relation- Anatomical noise ship of structures relative to other structures or locations in the (Diagnostic Radiology) Anatomical noise is a psychophysical body. noise source in radiographic images. It is caused by the projec- tion of overlaying small anatomical structures on the radiograph. Anterior: In front of, front (e.g. the kneecap is located on These overlaying structures cannot be distinguished by the the anterior side of the leg). human observer as specific anatomical detail and can be mod- Posterior: After, behind, following, toward the rear [e.g. elled as another source of image noise as it limits the detection of the shoulder blades (scapula) are located on the poste- small pathology. rior side of the body]. Figure A.39 illustrates how anatomical and quantum noise Distal: Away from, farther from the origin (e.g. the hand is affects the detection of pathology by a human observer. It assumes located at the distal end of the forearm). that quantum noise is described by the Rose model and is thus Proximal: Near, closer to the origin (e.g. the proximal end proportional to the absorbed detector dose. As the radiation dose of the femur joins with the pelvic bone). to the detector increases, the quantum noise signal-to-noise ratio Superior or Cephalic: Above, over or towards the head increases. However, the reproduction of small anatomical struc- (e.g. the elbow is superior to the hand). tures also increases in clarity with radiation dose, which in turn Inferior or Caudal: Below, under or towards the feet (e.g. increases the anatomical noise. the foot is inferior to the knee). Relative impact on contrast needed for detection of pathology v Anechoic 43 A nger logic Medial: Towards the mid-line, middle, away from the side provides detailed images prior to treatment. MRI and MR and CT (e.g. the middle toe is located at the medial side of the foot). angiography as well as conventional arteriography may be used to A Lateral: Towards the side, away from the mid-line (e.g. the image cerebral artery aneurysms. little toe is located at the lateral side of the foot). Treatment: The traditional treatment for an abdominal aortic Contralateral: On different sides of the midline. The right aneurysm is to replace the affected artery surgically with an arti- shoulder and left hip are contralateral to each other. ficial graft; this effectively excludes the aneurysm from the cir- culation. A newer procedure, endovascular aneurysm repair, uses Anechoic stents introduced via arterial catheters to exclude the aneurysm. A (Ultrasound) Anechoic, or hypoechoic, describes areas of an cerebral artery aneurysm may be surgically clipped or embolised ultrasound image which contain no visible echoes. This is nor- with metallic coils introduced via arterial catheter under radio- mally a result of imaging fluid where there is little or no backscat- graphic control. tering. An example is shown in Figure A.40. Related Article: Aneurysm clips Aneurysm Aneurysm clips (General) An aneurysm is an abnormal local enlargement, or (General) Aneurysm clips are metallic, often titanium, surgical dilatation, of a blood vessel. The term usually refers to an artery devices used to prevent the rupture of intracranial aneurysms. They although locally enlarged veins may also be described as aneurys- take the form of a clip with blades and a coiled spring which ensures mal. Aneurysms are found in several arteries though they are most closure of the blades. The clips are designed to isolate balloon-like commonly seen in the abdominal aorta. A pseudoaneurysm or false aneurysms from the intracranial arterial circulation. The clip is aneurysm is a contained leak from an artery, usually iatrogenic and applied under visual control using an operating microscope through often as a result of femoral artery puncture during arteriography. a surgical opening in the skull called a craniotomy. The clips work Clinical Consequences: Typically aneurysms may grow best on aneurysms with a distinct neck that separates the aneurysm in size and possibly rupture; the clinical consequence of this from the artery. An alternative to clipping is coil embolisation where depends upon the site. In the brain the bleeding may be into the small platinum coils are introduced into the aneurysm under radio- brain itself or into the surrounding subarachnoid space, which can graphic control through an arterial catheter. The coils cause the result in a variety of symptoms including stroke and may be fatal. aneurysm to thrombose, thus isolating it from the circulation. In the abdominal aorta there may be extensive blood loss into the Related Article: Aneurysm abdomen which is often fatal. Abdominal aortic or cerebral artery Hyperlinks: Clip. http: / /www .mayfi eldc linic .com/ PE -Cl ippi n aneurysms often have no or non-specific symptoms. It is there- g .htm ; Coil. http: / /www .brai naneu rysm. com /a neury sm -tr eat me nt fore difficult but important to correctly diagnose their presence. .ht ml If this is done, an elective procedure has a much higher success Further Reading: Louw, D. F., W. T. Asfora and G. R. rate than emergency treatment after rupture. An aneurysm may Sutherland. 2001. A brief history of aneurysm clips. Neuro. surg. become lined with thrombus, throw off emboli or even occlude. Focus 11(2):E4. An acutely occluding popliteal artery aneurysm can result in criti- cal ischaemia and limb loss. Anger logic Imaging: Ultrasound is used to screen for abdominal aortic (Nuclear Medicine) The event localisation process in a scintilla- aneurysm; it is a simple, noninvasive test with a high sensitivity. It tion camera system is referred to as anger logic. In conventional may be performed as a screening test on those particularly at risk, analog scintillation camera systems the position is determined by men aged 65 years and older. Abdominal aortic aneurysms may be splitting the PM-tube signal into four different output lines. The imaged with conventional arteriography but if lined with throm- signal of each output line is denoted as X+, X−, Y+, Y−. Each output bus the true diameter may be underestimated. CT angiography line is associated with a resistor and the value of the resistance differs between the different output lines (see Figure A.41). The output line signals are used to determine the X and Y position over the entire detector surface. The X-position is given by the ratio between the difference in X+ and X− signal and the total X signal (X+ + X−). The same goes for the Y position: Y Y+ 1 2 3 R + 4 5 6 7 Y R + X X 8 9 10 11 12 X– X+ 13 14 15 16 RX– R – 17 18 19 Y Y– (a) (b) FIGURE A.40 Anechoic region in the renal pelvis as a result of obstruc- tion. There are no or very weak echoes in the fluid. Weak reverberation FIGURE A.41 (a) A schematic representation of the 19 PM tubes in is evident and enhancement deep to the anechoic region due to reduced a scintillation camera. (b) The output lines from a single PM tube. The attenuation through the fluid. PM-tube signal is divided to four output lines by four resistors. Anger scintillation camera 44 A ngular sampling intervals in computed tomography + Related Articles: Scintillation camera, Scintillation crystal A (X - X - ) X = Hyperlinks: Hal O. Anger: interactive .snm .org /index .cfm? (X + + X - ) PageID=4577&RPID=969 ( Further Reading: Anger, H. O. 1967. Radioisotope cam- Y + - Y - ) (A.23) eras. In: Instrumentation in Nuclear Medicine, ed., G. J. Hine, Y = (Y + + Y - ) Academic Press, New York, Chapter 19, pp. 516–517. Angiogram The positions are normalised so that the calculated position does (Magnetic Resonance) Angiogram is an image showing vessels, not depend on the energy deposited, i.e. pulse height. Y and X obtained by, e.g. x-ray or MRI. To achieve this result, contrast can range from −1 to +1 and in a perfect scintillation camera the media are injected into the vessels (e.g. contrast media with values would change linearly when moving from the lower left Iodine for x-ray angiography, or paramagnetic nanoparticles in corner (−1, −1) to the top right corner (+1, +1). Non-linearities will MRI). See also Magnetic resonance angiography (MRA) and give rise to either pincushion or barrel distortion. digital subtraction angiography (DSA). In digital cameras the signal from each PM tube is digitised Related Articles: Magnetic resonance angiography (MRA), DSA and the position is calculated using software. The digital event localisation is analogous to the resistor read out but it also allows Angle of beam incidence for more complicated algorithms. One commonly used approach (Radiotherapy) See Oblique incidence to improve the positioning accuracy is to discriminate PM tubes with low signals from the position calculation. The largest signal Ångström contribution in these PM tubes is due to noise rather than event- (Nuclear Medicine) An angstrom (ångström) is a unit of length induced signal, i.e. they contain very little positional information. equal to 1 × 10−10 m (or 0.1 nm) and is abbreviated as Å. It is Another advantage with this approach is that when only a few PM used in the field of spectroscopy, atomic physics and chemistry tubes surrounding the interaction position are used for position- where the size of atoms, visible light spectra and length of chemi- ing, the other PM tubes can be used for simultaneous acquisition, cal bonds are sometimes measured in ångström. thus increasing the count rate performance. Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Angular anisotropy effect Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, (Nuclear Medicine) See Anisotropy Philadelphia, PA, pp. 215–217. Related Articles: SPECT, Barrel distortion, Pincushion dis- Angular sampling intervals in computed tomography tortion, Photomultiplier (PM) tubes (Diagnostic Radiology) On 3rd generation CT scanners the x-ray tube and detectors rotate around the object being imaged, Anger scintillation camera to acquire attenuation data at different angular positions (Nuclear Medicine) The scintillation camera was invented by Hal (Figure A.42). The number of ‘views per rotation’ refers to the O. Anger (1920–2005) (shown in the following picture) in the number of times the detectors are sampled during one tube rota- mid-1950s and is also referred to as Anger scintillation camera. tion which determines the angular sampling interval. The angular In some texts also the term ‘gamma camera’ is used. sampling interval affects the circumferential scan plane spatial For a detailed description – see Scintillation camera. resolution (i.e. the resolution along concentric rings in the image) of the CT scanner. On some CT systems the views per second (sampling fre- quency) remains constant with rotation time, so that for fast rotation times the number of views per rotation decreases. See Example A. This leads to a reduced circumferential spatial resolution. Related Article: Computed tomography 1 2 3 FIGURE A.42 Diagram of CT x-ray fan beam at three different angular Hal Anger, inventor of Anger camera. positions. Anisotropy 45 Annihilation coincidence detection Anisotropy where (Nuclear Medicine) Anisotropy is the property of being direction- E is energy A ally dependent as opposed to isotropy which is homogeneous in kB is the Boltzmann’s constant all directions. Examples of anisotropy are the fission fragments in T is the temperature a nuclear reactor. The fission products of 239Pu are most likely to yield a fission product couple where the lighter nuclei have a mass The life time of the trapped state can be large (up to hundreds of 90–100 u and the heavier of 130–140 u. of years) if E is large. Before using a TLD for dose measurement it is

necessary to remove electrons and holes from its trapping centres by a heat EXAMPLE A: treatment (Figure A.43b). This procedure is called annealing. Abbreviation: TLD = Thermoluminescent detector. Related Articles: Thermoluminescent dosimeter, Dose, Rotation Time (s) (Sampling Frequency [Hz]) Radiation dosimetry Views per Second Samples per Rotation Further Readings: Knoll, G. F. 2000. Radiation Detection 1 2000 2000 and Measurement, 3rd edn., John Wiley & Sons Inc., New York, 0.5 2000 1000 pp.731–736; Stabin, M. G. 2008. Radiation Protection and Dosimetry. An Introduction to Health Physics, Springer, New On other systems, the sampling frequency changes with York, pp. 157–159. rotation time, so that the number of times the detectors are sampled remains constant and spatial resolution is main- Annihilation tained. See Example B. (Nuclear Medicine) The interaction between a beta emitted posi- tron and an electron. As a result of the interaction the particles are annihilated, hence the name of the process. The interaction takes place when the beta particle is brought to a near or complete stop and as a result two photons are emitted from the point of interac- EXAMPLE B: tion, each photon with an energy equivalent to the electron rest mass of 0.511 MeV. Because of the conservation of momentum the two photons are emitted in an almost exactly opposite direc- Rotation Time (s) (Sampling Frequency [Hz]) tion (Figure A.44). Views per Second Samples per Rotation Beta emitters are used in PET imaging. 1 2000 2000 Related Articles: PET, Beta decay 0.5 4000 2000 Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA, pp. 25–26. Annealing Annihilation coincidence detection (Radiation Protection) In thermoluminescent detectors (dosim- (Nuclear Medicine) The spatial localisation process in PET is eters) the absorbed energy of the ionising radiation is stored by called annihilation coincidence detection (ACD). During the elevating electrons from the valence to the conduction band; the annihilation process a positron undergoes mutual annihilation electrons are then captured at trapping centres (metastable states) with an electron which produces two photons with identical within the band gap, below the bottom of the conduction band. energy (511 keV). The photons are emitted simultaneously with The holes created in the valence band can move through the an almost 180° opposing direction. The two annihilation photons crystal and be trapped by the trapping centres situated above the are registered simultaneously and the system can localise the valence band (Figure A.43a). The probability that an electron escapes from the trap is pro- portional to a Boltzmann factor: 0.511 MeV æ -E ö expç ÷ è k BT ø Conduction band e+ e– Valence band (a) (b) 0.511 MeV Electron trap Hole trap FIGURE A.44 Schematic representation of an annihilation interaction FIGURE A.43 (a) Formation of electron–hole pair and its trapping; (b) between a positron and an electron. Following the annihilation two pho- removing of electron and holes from traps by annealing. tons are emitted with a 180° angle to each other. Annihilation photons in positron decay 46 Anode A Line of radioactive material taken into the body of an adult worker by response inhalation or ingestion. The annual limit of intake (ALI) is the Detector intake of a given radionuclide by inhalation, ingestion or through ring the skin in a year by the reference man that would result in a com- Point of mitted dose equal to the relevant dose limit. The ALI is expressed annihilation in units of activity. ALI is the smaller value of intake of a given radionuclide in a Registered year by the reference man that would result in a committed effec- photon tive dose equivalent of 50 mSv or a committed dose equivalent of 500 mSv to any individual organ or tissue. The ALI of any radionuclide depends on the following factors: FIGURE A.45 Annihilation coincidence detection. The annihilation pho- (1) the type of radiation emitted, (2) energy of the radiation and tons are simultaneously detected by two opposite detectors, and an event is assumed to have occurred along the line of response. The dotted line is that of any radioactive progeny, (3) the selective biodistribution the LOR, and it represents the actual path of the two annihilation photons. and accumulation in specific organs or tissues and (4) the effec- tive half-life. Note: ICRP intended to replace ALI with Dose Coefficients origin of the event along a line between the two detectors. This with Publication 68 (1995), but ALI is still in use. line is referred to as the line of response (Figure A.45). Abbreviation: ICRP = International Commission on Radiation When one photon is detected the coincidence processor exam- Protection. ines events in opposite detectors during a specified coincidence Further Readings: Annals of the ICRP. 1982. Limits for timing window, which is typically 6–12 ns. A coincidence is Intakes of Radionuclides by Workers, Publication 30, Elsevier, assumed to have occurred when a number of simultaneous events New York; Annals of the ICRP. 1995. Dose Coefficients for are detected within the coincidence timing window. Intakes of Radionuclides by Workers, Publication 68, Elsevier, When using ACD to localise events, there is no need to use New York. absorptive collimation, which is necessary in SPECT to get any spatial resolution. ACD can be seen like an electronic collima- Annular array tion instead of a more physical approach as in SPECT. Since no (Ultrasound) Annular arrays are transducers with a concentric absorptive collimation is used PET sensitivity is much higher than ring of transducer elements. By altering the time or phase between for SPECT or conventional planar imaging. The high sensitivity elements, the depth of best focus can be altered in the imaging makes PET more suitable for relatively fast dynamic studies and plane and in the elevation plane (Figure A.46). The transducer also reduces the artefacts due to patient movement. is rotated about an axis to sweep through a volume to produce a Related Article: PET sector image. Conventional annular arrays are no longer available Further Reading: Cherry, S. R., J. A. Sorenson and M. E. on most commercial scanners. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Experimental annular arrays have been used for intravascular Philadelphia, PA, pp. 325–327. imaging. By arranging elements in a ring/annular configuration in a catheter, imaging has been shown to be possible both radially Annihilation photons in positron decay and in a forward direction. (Nuclear Medicine) See Annihilation radiation Anode Annihilation radiation (General) The positive electrode (e.g. of a detector). (Nuclear Medicine) Annihilation radiation refers to the process when a particle and its antiparticle collide. In this process the two particles are annihilated and radiate photons or particles. In nuclear medicine, annihilation radiation refers to the process where an electron collides with its anti particle, the positron. The positron is usually emitted from a β+-emitting radionuclide like 18F or 68Gd and when brought to a halt it pairs up in an annihila- tion process with an adjacent electron that emits two photons. In order to conserve momentum and energy the two photons are emitted in opposite directions with total energy equal to the rest mass of the annihilation particles, i.e. 511 keV for each photon. The fact that the annihilation is followed by emission of two photons with opposite direction constitutes the foundation for PET imaging. After registering an event in a detector the opposite detectors are scanned for a potential detection of the second anni- hilation photon within a time frame. If such an event occurs, a line of response is recorded between the two detectors. Related Article: Positron emission tomography (PET) FIGURE A.46 (a) Diagrammatic representation of an annular array transducer. By altering the timing between adjacent concentric rings, the Annual limit of intake (ALI) zone of best focus can be brought nearer to or further from the transducer. (Nuclear Medicine) Annual limit on intake, ALI, is defined by (b) The transducer can be used to insonate in the forward as well as the the ICRP (Publication 30) as the derived limit for the amount of radial direction. Anode acceleration 47 Anode cooling curve Anode acceleration Some contemporary x-ray tubes (especially mammographic (Diagnostic Radiology) Anode acceleration is very important for tubes) use anode target perpendicular to the beam of thermal A minimising the time between activating the exposure (pressing electrons (i.e. no bevel of anode surface). In this case the tubes the button) and the actual radiograph. This time is usually 1–2 s are mounted in a tilt position (in comparison to the image plane), and depends on the motor of the rotation anode, the bearings and this way presenting an effective anode angle (incorporating the the mass (kg) of the anode disc. Almost all x-ray tubes with a rota- tilt angle). This requires the anode angle to be defined in relation tion anode use an induction electrical motor with several speeds to the central x-ray beam – the anode angle is the angle between of rotation. The available speeds depend on the frequency of the the central x-ray beam and the anode target. This definition satis- electricity supplying the motor. All x-ray generators have a spe- fies the classical geometry in Figure A.47 (angle α), as well as the cial accelerating circuit (anode starting device) initially supplying situation when the x-ray tube is tilted. the motor with higher frequency (to quickly reach the necessary Figure A.48 represents three cases: on the left is a normal revolutions per minute [rpm]). In high frequency generators this x-ray tube with bevelled anode (actual anode angle 15°); in the is achieved very easily, as all circuits there use varying high fre- middle is the same tube, but tilted at angle β = 6°, thus presenting quency. This acceleration problem does not exist in x-ray tubes an effective anode angle of 21° (sum of 15° + 6°); the tube on the with liquid metal bearings. right is with perpendicular anode (i.e. no anode bevel), but tilted Related Articles: Anode, Rotation anode, Anode rotation to 25° (β = 25°), thus acting as x-ray tube with anode bevel of 25°. speed, Anode starting device, Bearing, High voltage generator, Related Articles: Stationary anode, Rotating anode, Target, Medium frequency generator Line focus principle, Biangular anode disk, Focal spot actual, Focal spot effective, Focal spot Anode angle (Diagnostic Radiology) The anode of the x-ray tube is normally Anode cooling chart angled; the range of the angle (bevel) is most often between 6° and (Diagnostic Radiology) See Anode cooling curve 20° (Figure A.47). The angle of the x-ray tube determines both the size of the effective focal spot Fe and the size of the actual Anode cooling curve focal spot Ft. The link between these focal spots and the sine of (Diagnostic Radiology) This is the chart showing the dissipation the anode angle α is known as line-focus principle: of the heat stored in the anode with time (in short this character- istic is also known as cooling curve). Usually this chart is pre- Fe = sina × Ft sented together with the chart showing the build-up of heat into the anode (anode heat storage charts, showing the tube load time). A smaller anode angle produces a smaller effective focal spot The heat imparted to the anode is absorbed by its structures and therefore better spatial resolution in radiographs. However, if and by its surrounding; due to this reason the heat storage capacity the anode angle is very small, then the useful field (over the film/ is a very important parameter which is naturally linked with the detector), covered by the x-ray beam, will be too small (as the cooling. The anode heat storage charts represent the link between conical x-ray beam will be too narrow). Often x-ray tubes used the heat units [HU] and the time (minutes) for their absorption. in CT have a very small anode angle, as they only need to cover These charts (input curves – HU/s or J/s [W]; 1 HU = 1.4 J) show a narrow slice and not a large field. Some tubes are made with the heat stored in the anode as a result of a long exposure or a surfaces forming two Anode angles (see Biangular anode disc). sequence of multiple exposures. These charts are very useful for fluoroscopy. One has to remember that in rotating anode tubes the thermal

path (focal spot track) gradually increases its tempera- ture, but in case of a single-phase generator the momentary actual focus (projected focal spot) increases its temperature by pulses in Anode ‘hot spots’ of the thermal path (depending on the anode rotation speed – rpm). The cooling curve on the same chart shows the α time (in minutes) for cooling the x-ray tube. The combined chart is presented in Figure A.49. Cathode Sometimes there is an additional cooling curve for the anode W-Re housing (which depends on the type of cooling oil in the housing). target In contemporary x-ray equipment the tube load/cooling is con- α-anode trolled automatically depending on the cooling curve. angle There are special rating charts and heat storage capacity charts α for computed tomography x-ray tubes. These are related not only Actual focal spot Effective focal spot } β }β Central x-ray beam towards the patient FIGURE A.48 Concept of effective anode angle, incorporating the FIGURE A.47 Anode angle of an x-ray tube and its relation with the anode bevel and the tube tilt angle (note the tube at right is with 0° actual effective focal spot and the actual focal spot. anode angle). Anode heel effect 48 A node heel effect X-ray beam span A 150,000 Rated anode capacity (degrees) 30° 120,000 Cooling 667 (472) (354 W) 90,000 500 Hu/s 2 1 Central beam 60,000 15° Anode 200 (141) 30,000 25% 50% 75% 0° 1 2 3 4 5 6 7 8 1 Time (min) 00% X- r r e a l y a t b iv e e a FIGURE A.49 Anode heat storage capacity chart and cooling curve. m intensity See the example on how to use these curves. (Courtesy of Sprawls Foundation, www .sprawls .org) FIGURE A.50 X-ray tube intensity spatial distribution for new (curve 1) and old (curve 2) x-ray tube. The x-ray intensity is shown in relation with the maximum intensity (100%) at the middle of the central beam of to the focal spot size (as for all tubes), but also to the scanning a new tube. time. These charts use the same concept, but are for much greater values, as most contemporary CT tubes withstand more than 1 million HU. of x-ray beam intensity (up to 50%) at the anode side of the beam. This is due to lesser production of x-ray photons at this direction (mainly due to absorption of the x-rays in the anode itself at the EXAMPLE FROM FIGURE A.49 lower end of the target surface). This decreased intensity of radia- tion at the anode site of the beam (if one looks it from the place of The heat storage curves represent the anode build-up power the patient) is known as ‘heel effect’. in watts and corresponding HU/s. The heel effect is more prominent with old x-ray tubes, If fluoroscopy is performed with 100 kVp and 3.6 mA, where all intensity of the beam decreases up to 50%. This over- this will result in 1.4 × 360 W = 500 HU/s – for 4 min this all decrease of beam intensity is primarily due to loss of x-ray fluoroscopy will import to the anode ~70,000 HU (by the radiation inside the cracks on the target surface. These cracks are heat storage curve). result of the thermal stress of the target after thousands of expo- If we additionally perform two exposures with param- sures (i.e. cycles of heating and cooling). The second example in eters 100 kVp, 500 mA, 0.2 s, the total heat delivered by Figure A.50 (curve 2) presents an old x-ray tube (used ~10 years at these will be 2 × 100 × 500 × 0.2 W = 20,000 HU. In this normal daily workload) where the maximal intensity has dropped case the total heat to the anode will be 70,000 + 20,000 = at a half of new x-ray tube and has shifted to 20° measured from 90,000 HU. This is safe (in this case the maximum tube the anode surface. This has produced a much more noticeable heat capacity by Figure A.49 is about 140,000 HU). heel effect. Note that the ‘age’ of an x-ray tube (measured with the The cooling curve shows that these 90,000 HU will number of exposures produced) depends greatly on the tube type be dissipated in 5 min. This is found by subtracting 2 min (target surface alloy, construction, etc.), as well as of the power (where the cooling curve crosses 90,000 HU) from 7 min of exposures. (where the cooling curve reaches 0 HU). For most radiographs the heel effect is almost unnoticed, as it is covered (absorbed) by the lead diaphragm attached to the x-ray tube. However the effect is seen with large area films (e.g. 35 × 43 Related Articles: Tube rate charts, Stationary anode, Rotating cm2) and specially when the x-ray tube is aged. Figures A.51 and anode, Target, Tube load time A.52 (detail) illustrate the heel effect (seen as drop of film optical Hyperlinks: Sprawls Foundation: http://www .sprawls .org / density) of an old x-ray tube and field 24 × 30 cm2. resources FFD Dependence: The heel effect is better seen when the Further Reading: Forster, E. 1993. Equipment for Diagnostic focus-film distance (FFD) is shorter, as in this case the lead dia- Radiology, MTP Press. phragm which specifies the radiographic field is naturally more Anode heel effect (Diagnostic Radiology) The x-ray anode generates radiation in all directions (only a fraction of it is at the direction of the patient). At diagnostic energy, this fraction is mainly at direction 90° from the direction of the incident electron beam (anode current) in the x-ray tube. The intensity of the radiation beam towards the patient has significant spatial variation. Figure A.50 (curve 1) presents an example where the maximal intensity of a new x-ray tube (marked with 100%) is at direction 15° measured from the anode surface FIGURE A.51 Heel effect seen with old x-ray tube (drop of optical den- (this depends on the type of the x-ray tube). There is a notable loss sity at the right side). (Courtesy of EMERALD project, www .emerald2 .eu) Stored heat (HU) Anode (of an x-ray tube) 49 A node (of an x-ray tube) Anode (of an x-ray tube) (Diagnostic Radiology) The anode of the x-ray tube is located A opposite the cathode at ~25 mm distance. In most tubes the anode is angled, the range of the angle (bevel) is normally 10°–20° (Figure A.54). The anode is at positive potential relative to the cathode (in some x-ray tubes the anode is grounded, but the cathode has high negative potential). This way the anode attracts the thermal elec- trons produced by the cathode. The small region of the anode, which is bombarded by the thermal electrons and produces x-rays is called the target. Almost 99% of the energy imparted to the tar- get by the electrons is converted to heat and secondary electrons are generated (strictly speaking the energy converted directly to heat is ~75%). Due to this reason the material of the target is normally tungsten – a material with a very high melting point. Tungsten also has a high atomic number, which is important for the effective conversion the energy of electrons to x-rays. This ‘bremsstrahlung generation efficiency’ (h) of the anode is dis- cussed in the article Target of the x-ray tube. The anode is not made entirely from tungsten as it would be too expensive. Tungsten is also not as effective at conducting heat away from the target as other metals. The anode stem is made from copper, molybdenum or other materials with good thermal conductivity and heat storage capacity (allowing cooling outside FIGURE A.52 Zoomed detail of the Heel effect from Figure A.51. (Courtesy of EMERALD project, www .emerald2 .eu) the x-ray tube). The cooling of the anode target is directly related to the power of the x-ray tube, which in turn is directly related to the construction of the anode assembly. Although a number of anode constructions exist (e.g. dynamax x-ray tube, field emis- sion x-ray tube, Straton tube, etc.) there are two general types of x-ray tube constructions – with stationary anode and with rotating anode (Figure A.55). Figure A.54 shows a stationary type anode (removed from a broken x-ray tube). Note the radiator at the end of the copper Anode stem (Cu) with radiator FIGURE A.53 Heel effect influenced by FFD. (Courtesy of EMERALD project, www .emerald2 .eu) widely opened (wider beam angle); hence, the edges of the beam are displayed. Compare the un-exposed ends of the films in Figure A.53 (upper with FFD = 100 cm and lower with FFD = 150 cm). FIGURE A.54 X-ray tube anode (stationary anode removed from the Related Articles: Anode, Heel effect x-ray tube envelope). Note the radiator at the end of the copper anode stem. Hyperlinks: EMERALD (DR module), www.emerald2.eu The x-rays are produced only from the small tungsten plate at the forehead Further Reading: Dendy, P. and B. Heaton. 2002. Physics for of the anode – on the right segment. Note it has been melted from the high Diagnostic Imaging, IOP Publishing, Bristol, UK. temperature. (Courtesy of EMERALD project, www .emerald2 .eu) Anode rotational speed 50 Antibodies and the actual x-ray generation (very important for targeted quick A 1 exposures of moving organs). Contemporary x-ray tubes with liquid metal bearings have very low friction, and in fact the anode reaches its rpm after switching on the equipment (e.g. in the morning) and constantly rotates (keeping these rpm) during the whole working day/period. Related Articles: Anode, Rotation anode, Anode acceleration, 2 Cooling curve, Bearing, Liquid metal bearing 5 3 Anode starting device (Diagnostic Radiology) See Starting device Antagonist 4 (Nuclear Medicine) An antagonist is a molecule that binds to a receptor without activating it. Instead the receptor is blocked and therefore hindered from performing its normal activity. An irre- FIGURE A.55 X-ray tube with rotating anode (low power); (1) tungsten/ versible antagonist binds permanently with a covalent bond to a rhenium (W-Re) coating (anode target material); (2) molybdenum body receptor. The receptor is then unable to function. An antagonist of the anode (often backed with graphite in high power x-ray tubes); (3) that binds to a specific subgroup of receptors is called a selective anode stem (often made of molybdenum); (4) anode rotor (with bearings); receptor antagonist. (5) anode support. (Courtesy of EMERALD project, www .emerald2 .eu) Anterior (General) Directional anatomical terms describe the relation- anode stem. Despite the cooling through the radiator the tungsten ship of structures relative to other structures or locations in the target plate of this particular stationary anode had been melted body. due to the high temperature (the melted area is seen on the right Anterior: In front of, front (e.g. the kneecap is located on the image segment). anterior side of the leg). Figure A.55 shows a rotation type anode (low power), removed Related Article: Anatomical relationships from a broken x-ray tube. The elements of the anode are: (1) tung- sten/rhenium (W-Re) coating (anode target material); (2) molyb- Anteroposterior (AP) projection denum body of the anode (often backed with graphite in high (General) There is a convention where the radiographic technique power x-ray tubes); (3) anode stem (often made of molybdenum); projection is identified by the direction of the x-ray beam. In the (4) anode rotor (with bearings); (5) anode support. antero-posterior projection the x-ray tube produces an x-ray beam Related Articles: Stationary anode, Rotating anode, Target which passes through the back to the front of the patient to pro- Hyperlinks: EMERALD, www .emerald2 .eu duce an image. Related Article: Technique projection Anode rotational speed (Diagnostic Radiology) The speed of x-ray tube anode rotation Anthropomorphic phantom is proportional to the power of the tube (see Rotation anode). A (Nuclear Medicine) A phantom designed to mimic the morphol- higher rotation speed leads to a more even heat distribution over ogy of the human body is referred to as an anthropomorphic the thermal track and therefore improved anode cooling. phantom. Anthropomorphic phantoms are used for Almost all x-ray tubes with a rotation anode use an asynchro- nous induction electrical motor with several speeds of rotation. At • Evaluation of system parameters in patient-like situations low power x-ray exposures (e.g. fluoroscopy) the motor is supplied • Educational

use with ~20 Hz electricity and rotates with ~1200 revolutions per • Evaluation of artefacts minute (rpm). When radiographic exposures are made the anode • Dose evaluation/calculation of delivered dose has to absorb more heat. This requires a higher speed of rotation, ~3000 rpm, which is achieved by supplying the induction electri- Examples of anthropomorphic phantoms are torso (lung, breast cal motor with ~50 Hz electricity. and cardiac), skull and pelvis phantoms. High power exposures (or sequence of exposures in angi- ography) significantly increase the temperature of the anode, Antibodies and quicker rotation is necessary to allow its even distribution (Nuclear Medicine) Antibodies are proteins found in blood or over the thermal track. In this case the motor is supplied with other bodily fluids. They are a part of the human immune sys- 150 Hz and thus rotates with ~9000 rpm. In reality the rpm tem and they identify and target foreign biomaterial, i.e. bacteria is a bit less than the earlier-given figure (e.g. ~8500 rpm, not and viruses. The antibodies are produced by one type of white the theoretical 9000 rpm). Some special x-ray tubes use up to blood cell called a B-cell. Most antibodies have a similar general 17,000 rpm. structure, except for a small tip of the protein which is extremely The speed of rotation is also related to the type of anode bear- variable. Each of these tips can bind to a specific structure, or ings. This is of special importance for the initial rotation of the an antigen. Antibodies produced by a specific type of B cell are anode. All x-ray generators have special circuitry which allows called monoclonal antibodies. These antibodies can be biologi- the exposure only after certain rpm have been reached. Due to cally engineered so that they target a specific biophysical process this reason the acceleration of the anode rotation is of importance and they are therefore very useful for nuclear imaging and radio- for minimising the time between pressing the exposure button nuclide therapy. Anticoincidence circuit in single channel analyser 51 Apex Further Reading: Imam, S. K. 2005. Molecular nuclear therapy. The approach uses antibodies which bind with high affin- imaging: The radiopharmaceuticals (Review). Cancer Biother. ity to the targeting agent in order to form a molecular complex. A Radiopharm. 20(2):163–172. The complex is taken up and degraded by the reticuloendothelial system (RES) which is a part of the immune system. A large por- Anticoincidence circuit in single channel analyser tion of the excessive target agent in systematic circulation can (Nuclear Medicine) An anticoincidence circuit is an electronic be removed by using this method. A possible application for this circuit which only produces an output pulse if the input pulse technique is to first administer radiolabelled antibodies and then occurs at one predetermined level. The circuit will not produce an anti-idiotype antibodies to achieve a RES-mediated clearance. output pulse if another input is sent simultaneously. Abbreviation: RES = Reticuloendothelial system. An example of the use of anticoincidence circuitry is in radia- Related Articles: Radionuclide uptake in tumour cells, tion detection. A scintillation detector produces a voltage pulse Extracorporeal elimination that is proportional to the energy of the incident radiation. These Further Reading: Carlsson, J., E. F. Aronsson, S.-O. Hietala, pulses are sorted using a single-channel pulse height analyser T. Stigbrand and J. Tennvall. 2003. Tumour therapy with radionu- where two discriminators D1 and D2 are connected to an anti- clides: Assessment of progress and problems. Radiother. Oncol. coincidence circuit. Only pulses above a minimum level will be 66(1):107–117. transmitted by the discriminators through to the anticoincidence circuit. If these voltages are V1 and V2, respectively (where V1 < Antiscatter grid V2), then the following is true: (Diagnostic Radiology) See Grid, Bucky • If the input pulse is less than V1 it will not be transmit- AORD (Artificial Optical Radiation Directive) ted by either discriminator. (Non-Ionising Radiation) See Artificial Optical Radiation • If the input pulse is greater than V1 but less than V2 it Directive (AORD) will be transmitted by D1 but not by D2. Therefore the coincidence circuit will produce an output pulse. • If the input pulse is greater than V2 then it will be trans- Aorta mitted by both discriminators simultaneously. In this (General) The aorta is the main artery supplying oxygenated situation, the coincidence circuit will not produce an blood from the heart to the systemic circulation. It arises from output pulse. the left ventricle of the heart at the aortic root as the ascending aorta. In the chest it is described as the aortic arch with branches which supply the arms and head. The descending aorta is known In this way, the pulse height analyser will only produce an output as the thoracic aorta to the level of the diaphragm. Inferior to the pulse if the input pulse falls between V1 and V2. diaphragm it is known as the abdominal aorta. In the abdomen it finally divides into the common iliac arteries. Antigen targeting Related Article: Aneurysm (Nuclear Medicine) Antigen targeting in nuclear medicine refers to the use of radiolabelled antibodies to target antigens. Antibodies can be molecular engineered to selectively target Aperture tumours which are thus ideal for tumour imaging and therapy. (Ultrasound) The meaning of the word aperture is opening or For diagnostic purposes, small fragments of antibodies are used gap. Theoretically, a plane wave of infinite extent will have the because of their favourable biokinetics, i.e. quick target uptake, same appearance at all locations in space. Transducers are not faster excretion reducing radiation dose and less chance of infinitely large and therefore cannot produce an infinite plane immunological response. For therapeutic purposes antibodies wave, but they can be thought of as a screen which blocks the are labelled with high-energy-emitting radioisotopes (preferably infinite plane wave except at the active surface of the transducer. alpha and beta emitters) which accumulate selectively in tumour Physically, the image of an opening in a screen where an infi- cells. Alpha emitters with a short half-life, like 213Bi and 211At are nite plane wave can pass is equivalent to a transducer with an unsuitable for labelling whole antibodies because the redistribu- active surface that vibrates. The situation is similar in optics, tion time for the antibodies widely exceeds the half-life of the where never an exact geometric shadow of an object is produced. two isotopes, hence giving an unnecessary radiation dose to sur- At the edges fringes are produced, an effect that is aggravated rounding organs prior to accumulation in the target organ. 213Bi as the object gets smaller compared to the wavelength. Christian and 211At are therefore labelled to antibody fragments rather than Huygens visualised this effect as the result of interference from an whole antibodies. infinitesimal number of spherical wave radiators on the surface of Related Articles: Tracer kinetic modelling, Receptor targeting, the aperture. In fact, also the plane wave can be visualised in this Neuroreceptor targeting, DNA targeting, Glycolysis targeting, manner, but then of course also the aperture is of infinite extent. Apoptosis targeting The result of blocking an infinite plane wave except within a cir- Further Reading: Imam, S. K. 2005. Molecular nuclear cular surface can be seen in Figure A.56. What is shown is the imaging: The radiopharmaceuticals (Review). Cancer Biother. intensity in the plane that includes the central axis of the beam. Radiopharm. 20(2):163–172. The aperture width is approximately four wavelengths. The sid- elobes that arise can be reduced by apodization. Anti-aliasing filter (Diagnostic Radiology) See Band limiting Apex (Nuclear Medicine) The apex of the heart is the lowest superficial Anti-idiotype antibody technique part of the heart and is directed downward, forward, and to the left. (Nuclear Medicine) This is a technique designed to minimise the It is overlapped by the left lung and pleura. In a bulls-eye image radiation dose to normal tissue in patients undergoing radioisotope representation the apex is located in the centre of the image. Apodization 52 Apodization A t Hz (a) (t) Hz x Windowing function (b) FIGURE A.56 Theoretically calculated beam profile of a circular trans- = ducer. The plane shown includes the central axis of the circular aperture. The aperture is actually located a small distance to the left of the figure for computational reasons. (c) Hz Apodization (Magnetic Resonance) In filter theory, apodization (Greek de- footing) is the process of reshaping the input signal in order FIGURE A.57 Concept of apodization applied to a SINC pulse: (a) SINC pulse and its Fourier transform, (b) truncated SINC pulse and Fourier to effect desired changes in its Fourier domain representation transform and (c) apodized truncated SINC pulse and Fourier transform. and in particular to reduce ‘ringing’ effects. For a temporal signal or a spatial distribution, apodization is achieved by multiplication by a windowing function. Mathematically this is equivalent to convolution of the Fourier transform of the in MRSI leads to truncation artefact or ‘Gibbs artefact’ when signal/distribution by the Fourier transform of the windowing k-space data are transformed to an image using a Fourier function. transform. As a result, data from one voxel can ‘leak’ to other, In MRI, apodization of sinc (=Sin(πt)/πt) excitation pulses adjacent voxels. Apodization of the k-space data prior to sup- can be used to suppress out of slice sidebands resulting from press outer k-space values prior to Fourier transformation helps the cut-off at the beginning and end of a pulse of finite dura- reduce ringing. The apodization process will also reduce spatial tion and to improve the uniformity of excitation through a slice resolution. profile. A Fourier transform of a sinc pulse of infinite duration Further Reading: Toga, A. and J. Mazziotta. 2002. Brain has a rectangular frequency/amplitude profile. In other words a Mapping, Academic Press, San Diego, CA. sinc pulse contains a finite band of frequencies of equal ampli- tude. This is the ideal profile to uniformly excite a slice cho- Apodization sen by slice selection. However a real sinc pulse is of finite (Ultrasound) Apodization is the distribution of energy across duration and includes sharp discontinuities. In the frequency the aperture of a transducer to control the intensity profile of domain representation these discontinuities appear as a ring- the ultrasound beam and to reduce side and grating lobes. For a ing effect, distorting the ideal sharp profile of the rectangular single transducer this may be achieved by tapering the electric frequency profile. By windowing the finite sinc pulse in the field towards the edges of the transducer, or by attenuating the time domain these sharp discontinuities may be smoothed out, beam in the same way over the surface. In arrays the task is and a frequency profile closer to the ideal obtained. Common simpler, since each element can be controlled individually and windowing functions used for apodization of symmetric sig- thereby the amplitude by which it is excited. Apodization of nals include the Hamming and Hanning weighting functions the receive beam is also used to reduce the effect of side lobes. (Figure A.57). The far field pressure magnitude along a cross section of the In MRS (magnetic resonance spectroscopy) a Fourier trans- beam is essentially the Fourier transform of the aperture function. form is used to generate spectra from FID (free induction decay) For a rectangular aperture the beam pattern will thus be a sinc- signals. The spectra show resonances corresponding to metabo- function, and significant side lobes will appear. A strong reflector lites present in the tissue voxel under examination. Apodization in the direction of a side lobe can then be interpreted as a weak may be used to reshape the FID prior to application of the Fourier reflector in the main lobe. transform. Generally the apodization function is a smoothly vary- By employing a tapering function on the aperture, as described ing function that attenuates signals occurring later in the FID. earlier, the level of the side lobes will decrease. These aperture The apodization process is equivalent to convolving the metabo- functions have different names, such as Hamming, Hanning or lite spectrum with a smoothing curve. Apodization produces a Gaussian, and all have different properties. The reason for using spectrum with an improved SNR but with some broadening of various functions is that there is a trade-off between several prop- spectral lines. erties of the far field and the aperture function. The side lobe

In magnetic resonance spectroscopy imaging (MRSI or level is probably the most important factor, but has to be weighed ‘chemical shift imaging’) spectroscopic methods are used to against width of the main lobe, as well as roll-off of the side lobes generate a low resolution image of the spatial distribution of (how fast their amplitude decrease with distance from the main a metabolite. The relatively small number of k-space samples lobe). Apoptosis targeting 53 Apparent focal spot Apoptosis targeting Related Articles: Source strength, Mass of radium, Contained (Nuclear Medicine) This term refers to the targeting of apoptosis activity, Equivalent mass of radium, Reference air kerma rate A enzyme markers by certain radiopharmaceuticals. Apoptosis is (RAKR), Air kerma strength self-induced cell death which normally prevents mutated cells surviving. However, this programmed cell sacrifice is lost dur- Apparent diffusion coefficient (ADC) ing the oncogenesis. One example is Annexin V, a protein that (Magnetic Resonance) The self-diffusion of water molecules is targets specific enzymes that are exposed during apoptosis. 99mTc a three-dimensional random motion where the speed of the mol- or 18F can be labelled to Annexin for SPECT or PET imaging, ecules depends on their size, the viscosity of the fluid and on the respectively. Annexin plays an important role when measuring temperature. The diffusion coefficient D is measured in area per the apoptosis in different clinical situations, e.g. it could dif- unit time [m2/s] and describes the area sampled by a molecule ferentiate recurrent tumours from necrosis or measure tumour during a specific time period. In diffusion MRI, the diffusion response. coefficient can be determined by measuring the signal for at least Related Articles: Tracer kinetic modelling, Receptor target- two different diffusion sensitivities (b-values) followed by a lin- ing, Antigen targeting, DNA targeting, Glycolysis targeting, ear, bi- or multi-exponential fit to the signal attenuation curve. Neuroreceptor targeting, Hypoxia targeting In vivo the diffusion coefficient is denoted apparent diffusion Further Reading: Imam, S. K. 2005. Molecular nuclear coefficient (ADC) since biological conditions affect the measured imaging: The radiopharmaceuticals (review). Cancer Biother. value. In particular, water molecules are hindered by, e.g. proteins Radiopharma. 20(2):163–172. and barriers consisting of cell membranes, and myelin sheaths. Hence the ADC value of the cerebral water is normally lower in Apparent activity tissue than in free water. Therefore, one important application of (Radiotherapy, Brachytherapy) Calibration of source strength is diffusion-weighted MRI is the detection of acute ischaemia by a very important part of a comprehensive brachytherapy quality lowered ADC as a result restricted motion of water protons. The system. Instruments, ion-chambers and electrometers, used for measured ADC value is also affected by perfusion and in areas source strength determinations, should have calibrations that are with a large perfusion fraction, e.g. in brain grey matter. Therefore traceable to national and international standards. the ADC value is often overestimated, especially if the diffusion Specification of Source Strength for Photon Emitting rate is determined based on measurements including only limited Sources: Source strength for a photon emitting source can be number (e.g. 2) of diffusion weightings and where the low sensi- given as a quantity describing the radioactivity contained in the tivity is below approximately b < 100 s/mm2 (Figure A.58). source or as a quantity describing the output of the source: Related Article: b-value Apparent focal spot 1. Specification of contained activity (Diagnostic Radiology) See Focal spot a. Mass of radium; mg Ra b. Contained activity; Ci, Bq 2. Specification of output Apparent focal spot a. Equivalent mass of radium; mg Ra eq (Radiotherapy) Apparent focal spot or apparent source point b. Apparent activity or position is relevant to electron beams since they have passed c. Reference exposure rate d. Reference air kerma rate e. Air kerma strength Contained activity is a quantity that can be used for all types of brachytherapy sources. Sources are encapsulated, and contained activity is difficult to determine. For brachytherapy dosimetry, the output of the encapsulated source is the quantity of interest, not the contained activity. The quantity apparent activity, an output specification, has been used as an alternative and is still used, especially for radiation protection applications. The apparent activity of an encapsulated photon emitting source is the activity of a hypothetical unfiltered point source of the same nuclide that gives the same air kerma rate or exposure rate at the same distance from the centre of the source. In modern brachytherapy dosimetry, reference air kerma rate or air kerma strength is the quantity used to calculate absorbed dose. Apparent activity is still used in some treatment planning systems to specify source strength. The user of such a system is cautioned to use the same value of the air kerma rate constant as the value used in the treatment planning system, when calculating apparent activity from the measured source output, the reference air kerma rate/air kerma strength. Apparent activity is also used in radiation protection applications. See Source strength for a full description of specification of FIGURE A.58 ADC-map obtained in a healthy volunteer based on a source strength. linear fit with b-values up to b = 1000 s/mm2. Apparent power 54 Applicator (brachytherapy) through a scattering filter. This will change the beam from its spots. In anger logic positioning mode the total PM tube signal A well-defined collimated shape to one which diverges. The degree received in a well-defined time interval is used in centroid esti- of divergence will be energy dependent and will mean that elec- mations, i.e. estimating the point of interaction. If two or more tron beams with different energies will appear to have originated events occur during this interval the resulting event will be placed at different source positions. between the events. This is more likely to happen if the pho- Source position is an important factor when calculating the ton intensity is high. Modern scintillation cameras can separate change in output factor for extended SSD treatments. simultaneous events if they are spatially separated by isolating See also Effective source point, Apparent source position and the involved PM tubes from each other and performing individual Virtual source position centroid estimations. Abbreviation: SSD = Source to surface distance. Abbreviation: PM = Photomultiplier. Related Articles: Apparent source position, Virtual source Related Articles: Anger logic, Photomultiplier (PM) tubes position, Effective source point Apparent source position Apparent power (Radiotherapy) Apparent source position is relevant to electron (General) Apparent power (sometimes referred to as ‘VA’) is the beams since they have passed through a scattering filter. This will value derived from the known AC current into, and AC potential change the beam from its well-defined collimated nature to one across an electrical load, whilst ignoring any phase difference which diverges. The degree of divergence will be energy depen- which may exist between the two parameters (Figure A.59): dant and will mean that electron beams with different energies will appear to have originated at different source positions. Apparent power = Potential in Volts rms * Current in Amps rms Source position is an important factor when calculating the change in output factor for extended FSD (focus to skin) In reality the apparent power is the maximum power that would treatments. be delivered into the load, and could only reach this level if the load was resistive. If the load has inductive or capacitive proper- Applicator ties, then a phase difference would exist between current and volt- (Radiotherapy) In order to provide a useable electron treatment age, resulting in a lower real power. beam it is necessary to attach an electron applicator (sometimes The relative phases of the AC potential and current must also called cone) to the head of the linear accelerator. These applica- be taken into account where a load possesses reactive properties tors typically come in a range of set field sizes (e.g. 6 × 6, 10 × 10, (capacitance or inductance). In such cases the true power deliv- 15 × 15, 20 × 20 and 25 × 25 cm2). ered to the load is given by The applicator is needed because the penumbra produced without it would be clinically unacceptable. This is due to the fact True AC power that while some beam shaping is provided by the secondary col- limators in the head of the linear accelerator, there is a significant = AC voltagerms * AC currentrms * cos (phase angle) amount of scatter both within the linear accelerator and in the air between the accelerator and the patient. Therefore the applicator Abbreviations: AC = Alternating current and RMS = collimates the beam and defines it typically at a distance of 5 cm Root-mean-square. from the patient. Some applicators are also used to provide addi- Related Articles: Alternating voltage, Alternating current, tional electron scatter thus improving the flatness of the beam. Direct current, Direct voltage If fields other than the sizes produced by the applicator set are required, it is common to create an appropriately shaped alloy Apparent source position that can be inserted into the end of the applicator. (Nuclear Medicine) Apparent source position refers to a certain Related Articles: Electron applicator, Collimation. incorrect positioning of photon counts in scintillation detectors using anger logic, occurring at high count rates. The effect is very Applicator (brachytherapy) prominent when the activity is concentrated in two or more hot (Radiotherapy, Brachytherapy) Applicators are used in brachy- therapy to position sources in the correct place relative to the tar- get volume. A wide variety of applicators exist, for intracavitary brachytherapy including all varieties, e.g. intraluminal, endobron- AC current and potential waveforms 400 chial, intravascular, and surface brachytherapy, and for interstitial brachytherapy, e.g. needles and catheters. 200 Applicators for high dose rate afterloading units must be closed in order to maintain source integrity; the source and its 0 drive cable must never come in contact with bodily fluids. Further, care must be taken when cleaning and disinfecting these applica- 1 4 7 10 13 16 19 22 25 28 31 34 37 40 –200 tors to ascertain that they remain dry inside. Pictures of applicators are shown in other articles: Temporary implant – high dose rate needles (closed), Permanent implant – –400 Time in milliseconds low dose rate needles for seed implantations (open), Intracavitary brachytherapy – radium applicator and high dose rate applicators Potential in Volts Current in Amps RMS value for gynaecological treatments, Interstitial implant – high dose rate and low dose rate needles. FIGURE A.59 Typical alternating current and voltage signals where a For safety reasons, it is imperative that vendor-specific con- phase shift of 60° exists between current and potential. nectors and applicators are used in remote controlled afterloading AC potential Apron, lead 55 Arcing of x-ray tubes techniques. But, that is actually not a limitation to the type of shows the three channels and also the fixation steel wire that goes application that can be made. Using flexible applicator tubes, for around the patient’s teeth in the upper left jaw. A instance, it is possible to fabricate custom-designed applicators. Related Articles: Brachytherapy, Temporary implant, To give an example, Figures A.60 and A.61 show a mould appli- Permanent implant, Intracavitary implant, Interstitial implant, cator with three channels, used to treat a patient with a maxil- Image guided brachytherapy lary cancer with high dose rate brachytherapy postoperatively. The three channels must go out through the patient’s mouth for Apron, lead connection to the afterloader. The mould is put into place in the (Radiation Protection) See Lead apron maxillary cavity via the mouth (the patient had undergone surgery previously, as well as preoperative radiotherapy); Figure A.61 Arcing of x-ray tubes (Diagnostic Radiology) The high vacuum inside the X-ray tube assures the undisturbed path of the thermal electrons from the From back-target; cathode filament to the anode target. Internal ionisation of the channels 1, 2, x-ray tube leads to vacuum reduction and internal discharges (arcing sparks) between the two electrodes or between the Channel 3 tube envelope and one electrode (usually the cathode) (Figure A.62 and Figure A.63). The arcing current between the high voltage electrodes of the x-ray tube can reach high values and can damage not only the x-ray tube, but also the x-ray genera- tor. Severe arcing can also lead to implosion of the x-ray tube. Manufacturers have special tests of the x-ray tube metal hous- ings to assure patient safety in case of such a dangerous situ- ation. There

are two main types of arcing – in new tubes (or Lead markers unused ones) and in old tubes. Arcing in New (Unusued) X-Ray Tubes: During x-ray tube manufacturing, the glass envelope is first vacuumed and then sealed. However, with time, the glass, and the other parts of the FIGURE A.60 Mould applicator with three channels seen ‘from the x-ray tube inside the vacuum, emit ions (cold emission). Special back’. The afterloader specific applicators are white. measures are taken for reducing this emission – such as polish- ing and degassing the glass and the metal electrode assemblies. However, this treatment does not eliminate the problem entirely. Cold emission exists, causing internal ionisation of the x-ray tube volume. This leads to discharges in the vacuum (small arcing sparks) increasing the current between cathode and anode. This is effectively a short-circuit and is associated with disruption of the production of x-rays for short periods of time. This can create artefacts or (rarely) damage the x-ray tube and generator. Because of this, new tubes and tubes that have not been used for a long time (or have been stored for a long period) must be ‘degassed’. The method includes slow warming with several low-power expo- sures before regular use. Arcing in Old (Aged) X-Ray Tubes: The arcing inside old x-ray tubes is more powerful and more dangerous. It is most often observed in x-ray tubes with glass envelopes. During each From front exposure, the glass envelope is heated to a very high temperature. This is followed by cooling and heating again during the next FIGURE A.61 The same applicator seen from ‘the front’ with the three exposure. This leads to thermal stress of the material, which can channels going out through the patient’s mouth. cause micro-cracks in the glass after several years of work. This FIGURE A.62 Arcing on an x-ray tube photographed through the glass envelope: (a) normal work of the x-ray tube (glowing-hot anode on the left and cathode cup on the right); (b) small spark (arcing) appears close to the cathode. Arithmetic mean of counts in attenuation correction 56 Artefact A FIGURE A.63 Arcing from Figure A.62 continues, forming a large cluster of sparks (a sparks ball) close to the cathode; (i) the large spark activates the electrical safety circuit, which switches off the high voltage and stops the x-ray tube operation. (All images: a, b, h, i – from Tabakov, 2018.) process increases the release of ions inside the tube. Additionally, The integration of the array elements with the available the evaporation of the cathode filament and the anode during the receiver channels in the MR system varies between system designs exposures leads to metallisation of the glass, which further cre- and has evolved over time. In the simplest concept, a single ele- ates conditions for arcing inside the x-ray tube. ment within a given array coil may be selected and switched into Further Reading: Tabakov, S. 2018. X-ray tube arcing: a single receiver channel. In such a case the elements are used Manifestation and detection during quality control. J. Med. Phys. independently and separately to build up a large field of view. In Int. 6(1):157–161. modern systems elements within the array coil are used simul- taneously. Furthermore elements in multiple array coils may be Arc therapy combined to effectively form a single, large FOV combined array (Radiotherapy) Arc therapy uses a standard radiotherapy linac coil. Each element may have its own dedicated RF channel, or and involves the delivery of the treatment beam in continuous share a channel through multiplexing or electrical combination mode as the treatment gantry is rotated around the patient. The of signals. For example, elements forming part of a pair CP (cir- treatment may be a single arc or several arcs. The treatment plan- cularly polarised) coils share channels in some designs (e.g. an ning system treats the arced beam as a set of discrete static beams 8-element, 4-channel head coil). close together in angle. Array coil designs have moved towards increased density of Abbreviations: IMAT = Intensity modulated arc therapy, receiver elements in order to increase the SNR achievable, evolv- Linac = Linear accelerator and MLC = Multileaf collimator. ing from 2 through 4, 8, 16, 32 and greater numbers of elements. Related Articles: X-ray therapy, Linear accelerator, Equally, the increased number of RF channels available on MRI Stereotactic radiosurgery, Volumetric-intensity-modulated arc systems has allowed dedication of receiver channels to individual therapy coil elements, with less sharing of RF channels. This allows great flexibility in the coverage and set up of individual examinations, Ardran and Crooks cassette with elements from many coils contributing to the final image, (Diagnostic Radiology) A testing cassette developed by GM without penalty in data acquisition speed. Ardran and HE Crooks (1968) that is used to measure the poten- tial applied to an x-ray tube (kVp) and the filtration. It contains ARSAC filters of varying thicknesses and works on the penetrameter prin- (Radiation Protection) See Administration of Radioactive ciple first developed and demonstrated by Roentgen. Substances Advisory Committee The Wisconsin cassette uses a similar principle. Related Articles: Wisconsin cassette, kV meter Artefact (Diagnostic Radiology) The term ‘artefact’ is derived from the Arithmetic mean of counts in attenuation Latin, arte factum, and, in diagnostic imaging, refers to structures correction in SPECT in the image that are not a true representation of the object. (Nuclear Medicine) See Attenuation correction in SPECT using A fuller description of artefacts in CT, and the origin of dif- conjugate counting ferent types of CT artefact, is given in each of the related articles listed in the following. Array coil Related Articles: Beam hardening, Cone beam artefact, (Magnetic Resonance) An array coil consists of a single physical Helical artefact, Image artefact, Metal artefact, Motion artefact, coil comprising several RF receiver elements. The individual ele- Partial volume effect (artefact), Ring artefact ments provide the high SNR of a small coil, while the combina- tion of many elements provides a large field of view (FOV) of the Artefact anatomy of interest. (General) In medical imaging artefacts refer to false signal phe- The development of array coils has been built on the concept nomena caused by a number of reasons. Artefacts can be intro- of the phased array, where the electromagnetic and electronic duced by a number of mechanisms, e.g. faulty equipment (e.g. design of the coil is optimised to avoid coupling between coil ele- broken PM tube in a scintillation camera) or patient movement ments. In this way, individual elements of the coils retain their during acquisition. For a radiologist it is important to recognise desirable small coil, high SNR properties. artefacts and separate them from pathological changes. Artefact 57 Arterial input function (AIF) Artefact artery. The AIF is of relevance in pharmacokinetic modelling, e.g. (Ultrasound) Artefacts are imperfections in an ultrasound image in the calculation of perfusion-related parameters. Theoretically, A caused by variations of the tissue properties, multiple reflections the ideal AIF would be instantaneous, i.e. the dose distribution of the ultrasound pulse and the physical properties of the ultra- over time is described by the Kronecker delta function. However, sound transducer. When forming an ultrasound image an ideal in practice, the arterial input of tracer is unlikely to be instanta- operating medium is assumed as follows: neous and thus extended in time due to the duration of the intra- venous tracer injection and the transport through the circulatory • The speed of sound in the medium is constant system, i.e. from the injection site, via the heart and lungs, to the • The attenuation in the medium is constant tissue of interest. • The beam axis is straight One important quantity in first-pass measurement techniques • Pulses received at the transducer originates only from is the tissue residue function R(t), i.e. the fraction of tracer that the beam axis remains in the tissue at a time (t) after an instantaneous arterial • An infinitesimally thin beam bolus input. Retrieval of the tissue residue function is normally not straightforward due to the fact that the arterial bolus input Variations from these assumptions lead to artefacts in the is extended in time. However, the convolution integral expresses image. These are categorised as speed of sound artefacts, attenu- the relation between the measured tissue tracer concentration C(t), ation artefacts, reflection artefacts and beam shape artefacts. the arterial tracer concentration AIF(t), the tissue residue function Artefacts are very common in diagnostic ultrasound imaging and R(t) and the tissue blood flow F: it is very important that the operator is aware of them. Artefacts can be annoying but they may also give the operator extra infor- t mation about the examined tissue. An example of this is the rever- C (t ) = F éëR (t ) Ä AIF (t )ùû = FòAIF (t) R (t - t)dt beration pattern from plaque in a vessel, Figure A.64. 0 Example of a Speed of Sound Artefact – Range Error: In human tissue we assume that the speed of sound is constant at By monitoring the tracer concentrations in tissue as well as in an 1540 m/s when calculating the distance between transducer and appropriate tissue-feeding artery, the convolution kernel can be reflecting surface using the range equation d = ct/2. Since human obtained by deconvolution. Hence, the blood flow F and the tissue tissue is not homogeneous the speed of sound differs slightly residue function can be determined. between different types of tissues. For example, in fat it could be In brain perfusion imaging by dynamic susceptibility contrast as low as 1420 m/s. If the speed of sound in the medium is less MRI (DSC-MRI), one single AIF site is often assumed to repre- than the assumed 1540 m/s then the echo will arrive later at the sent all arterial input locations of the entire brain (Figure A.65). transducer than expected. This means that the target position will This approach is, however, dubious since any arterial dispersion be projected further away from the transducer than the real target occurring between the site of the AIF registration and the true position. This specific kind of artefact is called Range error. site of arterial input will introduce an error in the deconvolution- Other speed of sound artefacts includes Boundary distortion, based calculation of cerebral blood flow (CBF) and mean transit Size errors and Refraction. time (MTT). In order to reduce such errors, the application of local or regional AIFs has been proposed. Artefact reduction technique Related Articles: Perfusion imaging, Dynamic susceptibility (Diagnostic Radiology) See Metal artefact contrast MRI, Cerebral blood flow, Cerebral blood volume, Mean transit time Arterial input function (AIF) (Magnetic Resonance) The arterial input function (AIF) is the tracer concentration time curve observed in a tissue-feeding 50 40 B-mode-reverberation from calcifications 30 A small calcified plaque causes reverberation with the plaque appearing to protrude further into 20 the lumen than was evident in another plane. (Note too the attenuation in tissue 10 deep to and in line with the plaque) 0 0 20 40 60 80 Position (x, y) = 39.70 FIGURE A.64 Reverberation pattern from plaque in a vessel. (Courtesy FIGURE A.65 Typical arterial input function (AIF) observed in of EMIT project, www .emerald2 .eu) dynamic susceptibility contrast MRI for assessment of cerebral perfusion. Arterial spin labelling (ASL) 58 Artificial intelligence (AI) Further Readings: Calamante, F., D. G. Gadian and A. A Connelly. 2000. Delay and dispersion effects in dynamic sus- ceptibility contrast MRI: Simulations using singular value Inversion decomposition. Magn. Reson. Med. 44:466–473; Calamante, slab F., M. Mørup and L. K. Hansen. 2004. Defining a local arterial Imaging input function for perfusion MRI using independent component slice analysis. Magn. Reson. Med. 52:789–797; Knutsson, L., E. M. Larsson, O. Thilmann, F. Ståhlberg and R. Wirestam. 2006. Calculation of cerebral perfusion parameters using regional arterial input functions identified by factor analysis. J. Magn. Reson. Imaging 23:444–453; Rempp, K. A., G. Brix, F. Wenz, (a) (b) C. R. Becker, F. Gückel and W.J. Lorenz. 1994. Quantification Uninverted inflowing spins Uninverted static spins of regional cerebral blood flow and volume with dynamic Inverted inflowing spins susceptibility contrast-enhanced MR imaging. Radiology 193:637–641. FIGURE A.66 The principle of a modified EPISTAR pulsed ASL label- ling technique. (From Edelman, R. R. et al., Radiology, 192, 513, 1994.) Arterial spin labelling (ASL) Part (a) illustrates the

labelling experiment and part (b) shows the cor- (Magnetic Resonance) Arterial spin labelling (ASL) or arte- responding control experiment. rial spin tagging (AST) is an MRI technique for generating quantitative perfusion maps using the arterial water spins as an endogenous tracer, i.e. without any exogenous contrast agents. for magnetisation transfer effects, the control inversion pulse is For this, the hydrogen nuclei of the inflowing intravascular arte- applied to a slab distal to the imaging slice. Usually a gap between rial water are labelled by applying one or two radiofrequency the inversion slab and the imaging slice is introduced to account pulses. Normally, a 180° RF pulse is applied to achieve an for imperfections in the slice selection profiles. Absolute quantifi- inversion of the arterial spins upstream of the region of interest, cation of perfusion with pASL requires a reliable model describing i.e. the imaging slice. The inversed spins arrive at the imaging the relationship between the measured difference in longitudinal slice and enter the tissue by means of water exchange between magnetisation and the perfusion. Other common pASL labelling the blood capillary system and the tissue. Hence, they contrib- schemes are flow alternated inversion recovery (FAIR) and proxi- ute to a reduced longitudinal magnetisation of the tissue and mal inversion with a control for off resonance effects (PICORE). subsequently to reduced signal in the MR image. Therefore, Dynamic or time-resolved ASL, where the bolus of labelled water this effect is directly related to the local microcirculation (i.e. is followed over time, has also been proposed (Petersen et al., the perfusion or regional cerebral blood flow). ASL perfusion 2006). The weakness of pASL is its lower SNR as compared with maps can be constructed by subtraction of the tagged/labelled cASL and its sensitivity to motion between the two acquisitions. image from a control image (without applying an inversion The advantages are lower SAR and less magnetisation transfer RF-pulse). The intensity of the difference image is directly effects. proportional to blood perfusion. However, the low contrast-to- For a recent ASL review, see Petersen et al. (1994). noise ratio normally requires averaging of a number of ASL Further Readings: Detre, J. A., J. S. Leigh, D. S. Williams images depending on the imaging sequence used (typically of and A. P. Koretsky. 1992. Perfusion imaging. Magn. Reson. Med. the order of 50–100 images). ASL sequences generally fall into 23:37–45; Edelman, R. R., B. Siewert, D. G. Darby, V. Thangaraj, two main groups of measurement techniques, i.e. continuous or A. C. Nobre, M. M. Mesulam and S. Warach. 1994. Qualitative pulsed ASL: mapping of cerebral blood flow and functional localisation with Continuous ASL or ASL Steady-State Techniques: echo-planar MR imaging and signal targeting with alternating Continuous ASL (cASL) is the original technique of ASL pro- radio frequency, Radiology 192:513–520; Petersen, E. T., T. Lim posed by Detre et al. (1992). The blood is continuously labelled and X. Golay. February 2006. Model-free arterial spin labelling as it passes through a plane proximal to the imaging slice using quantification approach for perfusion MRI. Magn. Reson. Med. a train of RF pulses. If an equilibrium is reached, the perfusion 55:219–232; Petersen, E. T., I. Zimine, Y. O. Ho and X. Golay. parameters are calculated by comparing with the signal from the 1994. Non-invasive measurement of perfusion: A critical review same, unlabelled slice. The drawback with this technique relates of arterial spin labelling techniques. Br. J. Radiol. August, to the amount of RF energy (i.e. relatively high specific absorption 79(944):688–701. rate) delivered, long transit time and considerable magnetisation transfer effects. Artificial intelligence (AI) Pulsed ASL: The primary difference between pulsed ASL (General) Artificial intelligence (AI) is a contemporary set of (pASL) and cASL is the labelling technique. In pASL, the arterial hardware and software tools simulating the human intellectual blood is labelled in a slab using one short RF pulse, which creates process. The main objective of AI is to support the decision- a ‘bolus’ of labelled spins. As an example of a labelling technique, making process in various disciplines through the application of the principle of echo-planar imaging and signal targeting with specialised computerised systems and powerful algorithms. The alternating radio frequency (EPISTAR) (Edelman et al. 1994). main phases of the AI process follow human intelligent behaviour This technique consists of two acquisitions (Figure A.66). In the and include learning, reasoning and self-correction. first acquisition (Figure A.66a) all spins are inverted in inversion AI has a variety of applications in healthcare – administra- slab below the imaging slice. After a time delay the inverted blood tive and management, diagnostics, therapeutics, decision support, spins (label) will flow into the imaging slice. In the second acqui- data analysis and research. Medical physicists and biomedical sition (Figure A.66b) no inversion of the slab below the imaging engineers play a key role in the introduction and progress of AI slice is performed as a control experiment. In order to account technologies in contemporary medicine: Artificial neural networks 59 As low as reasonably practicable (ALARP) • Electronic Health Records – data management, analysis Artificial optical radiation (AOR) and prediction (Non-Ionising Radiation) Within the Control of Artificial Optical A • Telemedicine – remote monitoring, data exchange and Radiation at Work Regulations, 2010, artificial optical radiation expert advise is defined as ‘any electromagnetic radiation in the wavelength • Diagnostic imaging – analyse and interpret imaging range between 100 nm and 1 mm which is emitted by non-natural results sources’. • Radiation oncology – data analysis and treatment Related Articles: Control of Artificial Optical Radiation at planning Work Regulations • Brain–computer interfaces – assist humans with Further Reading: Health and Safety Executive, Control of disabilities Artificial Optical Radiation at Work Regulations 2010, S.I no. 1140. www .l egisl ation .gov. uk /uk si /20 10 /11 40 /pd fs /uk si _20 10114 Further Readings: European Commission High-Level Expert 0 _en. pdf. Group on Artificial Intelligence, https :/ /ec .euro pa .eu /digi tal -s ingle -mark et /en /high -leve l -exp ert -g roup- artifi cial -inte llige nce; AORD (Artificial Optical Radiation Directive) The ITU/WHO Focus Group on Artificial Intelligence for Health (Non-Ionising Radiation) AORD is the commonly used acronym (FG-AI4H), www .i tu .in t /en/ ITU -T /focu sgrou ps /ai 4h /Pa ges /d for the EU Directive 2006/25/EC on artificial optical radiation, efaul t .asp x; United Nations Activities on Artificial Intelligence adopted by the European Parliament in April 2006. The Directive (AI) http: / /han dle .i tu .in t /11. 1002/ pub /8 12 0d5 d5 -en . is aimed at protecting all employees working with artificial opti- cal radiation sources, and it refers to exposure limits established Artificial neural networks by the International Commission on Non-Ionising Radiation (General) An artificial neural network is a mathematical or com- Protection (ICNIRP). All EU states were required to implement it putational model conceptually based on the physical and chemi- by the end of April 2010. cal interconnectivity of neurons in a living biological organism. Artificial optical radiation sources in the healthcare environ- The biological neuron is a single cell which is organised with ment range from general lighting (usually harmless) to special potentially many inputs and one single output and this is repli- treatment sources such as ultraviolet phototherapy equipment cated by the artificial neuron as a mathematical function capable which are high risk. of receiving many inputs and summing them, often with some Related Articles: ICNIRP, Exposure limit values, weighting factor applied and filtered by a non-linear activation Phototherapy function, to produce an output. The artificial neurons are then Further Readings: Council Directive 2006/25/EC on the organised in layers to form a neural network where the output of minimum health and safety requirements regarding the exposure one layer acts as an input into neurons in the next layer or even of workers to risks arising from physical agents (artificial optical feeding back into the input of previous layers. radiation) (19th individual Directive within the meaning of Article A collection of such relatively simple mathematical pro- 16(1) of Directive 89/391/EEC) [2006] OJ L 114; A Non-Binding cessing elements has the ability to express complex behaviour Guide to the Artificial Optical Radiation Directive 2006/25/EC, through the connections between the elements and the connec- Radiation Protection Division, Health Protection Agency. tion parameters. An artificial neural network is an adaptive sys- https :/ /os ha .eu ropa. eu /en /legi slati on /gu ideli nes /n on -bi nding tem that can learn by changing its connection structure based on -guid e -to- good- pract ice -f or -im pleme nting -dire ctive -2006 -25 -e c external or internal information that flows through the network -201 aarti ficia l -opt ical- radia tion2 019, last accessed January 2020. during the learning phase. It therefore can be used in non-linear statistical data modelling or to find and recognise patterns in data or images. As low as reasonably achievable (ALARA) principle Modern artificial neural networks can be used in radiology to (Radiation Protection) A basic principle of radiation protec- assess images, extract specific features and automatically differ- tion set down by the International Commission for Radiological entiate between various pathological findings. Protection in their 1977 Recommendations (Report 26). This is the principle that exposure to radiation should be kept as low as possible to provide an effective diagnostic image while limiting Input Artificial neural network Output risk to the patient. Since changed to ALARP (as low as reasonably practicable) to reflect the cost factors (economic and social) that may be considered by the employer. For more information, see Optimisation. Related Article: Optimisation As low as reasonably practicable (ALARP) (Radiation Protection) A basic principle of radiation protec- tion set down by the International Commission for Radiological Protection in their 1990 Recommendations (Report 26). It super- seded the ALARA Principle (as low as reasonably achievable) to reflect the cost factors (economic and social) that may be consid- ered by the employer. For more information, see Optimisation. Related Articles: Optimisation, As low as reasonably achiev- able (ALARA) principle ASA 60 Asymmetric fields ASA institution or private company. Usually, healthcare facilities A (Diagnostic Radiology) The ASA (American Standards differentiate capital assets from operating expenses based on the Association) is an old system for photographic film speed (pho- combination of two main factors: the purchase cost of the assets tographic exposure), which became the basis for the current ISO (establishing a minimum threshold for considering the item as a film speed system (used worldwide). capital asset) and the useful life of the item (if it spans more than ASA has been introduced by the American National one year it is considered a capital asset, while items consumed Standards Institute (ANSI) – a private non-profit organisation during the same fiscal year are treated as operating expenses). overseeing various standards for products, services, systems, etc. Medical equipment managed by clinical engineering depart- in the United States. ANSI also coordinates US standards with ments are usually considered as capital assets; therefore, asset international standards. depreciation has different uses in clinical engineering: The most widely used system for film speed currently is the one produced by the International Organisation for Standardisation • To support the hospital finance management during (known as ISO). capital assets budgeting Another old system for film speed DIN has been introduced by • To establish the residual value of equipment (as opposed the Deutsches Institut für Normung e.V. (known as DIN), which to other values often utilised, such as acquisition cost or in English stands for German Institute for Standardisation. replacement value) In principle the photographic-related standards of these sys- • To establish maximum maintenance expenditure lim- tems measure the photographic film sensitivity to light. This way its and evaluate for each repair if it is worthwhile to the film speed can be determined from the characteristic curve of proceed with the related repair costs or if it is better to the film. While this measure is relatively simple for black/white dismiss the equipment and replace it with a new one films (as x-ray films), the colour films require separate curves for • To compare different equipment in terms of life cycle blue, green, and red. cost, assessing the yearly cost for each device (including The current standard ISO 5800 (from 1987) defines both an the

need of consumables, maintenance costs and asset arithmetic scale and a logarithmic scale for measuring colour- depreciation as here defined) negative film speed. The ISO arithmetic scale corresponds to the old ASA scale. Asymmetric energy window The ISO logarithmic scale corresponds to the old DIN scale. (Nuclear Medicine) In an asymmetric energy window the pho- For example topeak is located off centre in the pulse height analyser window. An example of an asymmetric energy window is one where the ASA 100 = DIN 21° upper threshold of the window is located at the centre of the pho- ASA 400 = DIN 27° topeak, a so-called asymmetric low window (or off peak low). ASA1600 = DIN 33° The opposite situation, i.e. lower threshold at the centre of the photopeak produces an asymmetric high window (or off peak Due to this reason the film speed is listed as ISO 100/21°; ISO high). Asymmetric energy windows are used to test a NaI (Tl) 400/27°, etc. crystal for possible hydration, to tune the PM tubes or to reveal Related Article: Characteristic curve electronic problems (Figure A.67). Hyperlinks: http://en .wikipedia .org /wiki /Film _speed Hyperlinks: www .IAEA .org Further Reading: Graham, L. S., A. Todd-Pokropek and E. Asset depreciation B. Sokole. 2003. IAEA Quality Control Atlas for Scintillation (General) The American Hospital Association publishes a book Camera Systems, pp. 18–19. titled ‘Estimated useful lives of depreciable hospital assets’. The last revision of the book was published in 2018 and is a useful Asymmetric fields reference for planning the replacement of capital assets based on (Radiotherapy) Traditional beam collimation uses four field-defin- their obsolescence compared to a fixed useful life that is specific ing jaws or collimators. One of the two sets of opposing jaws move for each type of equipment. In the introduction to the book, there concurrently to define the field width and the other set defines the is a list of definitions, and we find the following: ‘Depreciation: field length, resulting in a square or rectangular field which has The cost assigned to an asset for a given period based on the esti- the field centre coincident with the collimator axis. Independent mated useful life of the asset. This cost is usually determined movements of the jaws allow the generation of rectangular beams based on one of several depreciation methods. For Medicare pur- which are offset from the central axis (asymmetric beams). It is poses, health care providers most commonly use the straight line method, in which depreciation is applied evenly over the asset’s useful life’. With this approach, a capital asset with a purchase cost of X and an estimated useful life of Y years will have a yearly asset depreciation of X/Y. Other formulas are available (e.g. in the Italian accounting rules, capital assets are assigned a fixed depreciation period based on the typology of item, and the depre- ciation is X/2Y the first and last year of depreciation, while X/Y is applied for the remaining years), but the concept is basically the same. This concept of asset depreciation is utilised for tax and Energy, keV Energy, keV accounting purposes, to differentiate the cost of capital assets from current year expenses, and properly account for capital FIGURE A.67 Schematic representation of a low and high asymmetric assets in the yearly balance sheet of a hospital or other healthcare energy window. Counts Counts Asymmetric jaws 61 Asymmetric screen film A True central axis Asymmetric jaw Independent Central ray of jaw asymmetric field SAD SAD x r FIGURE A.69 Geometry of asymmetric jaws. FIGURE A.68 Geometry of asymmetric field. concurrently to define the width and the length of the irradia- tion field. The resulting square or rectangular field has the centre important to distinguish between the central axis as defined earlier coinciding with the beam axis. Modern linear accelerators have and the beam axis which represents the projection of the centre of collimating jaws that can be moved independently of the cor- a defined collimator opening. For symmetric beams they are coin- responding opposed jaws and this permits to block a portion of cident but they differ for asymmetric beams (Figure A.68). the field from one side without affecting the opposite jaw setting. The asymmetric collimation produces a nontrivial alteration Some linacs have one independent jaw, others have two or four in average beam energy, absolute radiation output, depth dose independent pairs. The independent jaw option is interlocked to and beam profiles. The dose due to the radiation scatter can be avoid errors in the setting of symmetric fields, in which case the separated into collimator and phantom components. The collima- opposite jaws open or close symmetrically. When one of the jaws tor scatter for an asymmetric field is almost the same as for a is closed the resultant asymmetric field has a smaller dimension symmetric field of identical dimension while the phantom scat- than the original symmetric field and its centre does not coin- ter components of symmetric and asymmetric fields are differ- cide with the beam axis of the symmetric field (Figure A.69). ent. The asymmetric beam collimation changes the off-axis beam Asymmetric jaws are sometimes used to block off part of the field quality as a consequence of using a flattening filter to reduce the without changing the position of the isocentre. dose rate at the photon beam centre. In fact the unfiltered pho- The uses of asymmetric jaws in clinic are to ton beam gives a sharply peaked dose distribution because of the angular distribution of the bremsstrahlung radiation and the pres- • Eliminate the beam divergence at the junction of adja- ence of the filter to modify the beam intensity results in a greater cent fields by blocking one half of the field along the beam hardening close to the central axis than in off axis regions. central axis beam, thus improving the dose distribution This effect causes a hot spot near the edge of an asymmetric beam over the junction far from the central axis. The hot spot reduces in magnitude with • Simplify the matching of two opposite tangential fields depth as the relatively softer beam near the edge of large field is • Irradiate in an arc therapy a target volume that sur- attenuated faster than the harder beam near the central axis. This rounds a critical organ effect, if not taken into account, introduces a dose error of about • Keep the same treatment centre as the original field 5% for large fields. Asymmetric beams require suitable calcula- • Keep the same treatment centre as the original field in tion techniques that consider the variation of dose with off-axis the set up of a boost field distance. Various dose computation methods have been proposed Although these functions have traditionally been performed by based on the use of appropriate correction factors. beam splitters or secondary blocking on a shadow tray, the use Further Readings: Khan, F. M., B. J. Gerbi and F. C. Deibel. of asymmetric jaws reduces the needed time for the set up and 1986. Dosimetry of asymmetric x-ray collimator. Med. Phys. avoids handling the massive shielding block. 13(6):936–941; Loshek, D. D. and K. A. Keller. 1988. Beam pro- Normally, the calculation of dose and isodose distribution in a file generator for asymmetric fields. Med. Phys. 15(4):604–610; patient requires basic data which are measured with symmetrical Marinello, G. and A. Dutreix. 1992. A general method to per- collimators. An asymmetric jaw setting produces changes in the form dose calculations along the axis of symmetrical and asym- depth dose that must be taken into account by the algorithms used metrical photon beams. Med. Phys. 19:275–281; Palta, J. P., K. to calculate dose and dose distribution. M. Ayyangar and N. Suntharalingam. 1988. Dosimetric charac- teristics of a 6 MV photon beam from a linear accelerator with Asymmetric screen film asymmetric collimator jaws. Int. J. Radiat. Oncol. Biol. Phys. (Diagnostic Radiology) Radiography cassettes generally have 14:383–387; Thomas, S. J. and R. J. Thomas. 1990. A beam gen- identical intensifying screens on each side of a double-emulsion eration algorithm for linear accelerators with independent colli- film. However, in some designs, the two screens have different mators. Phys. Med. Biol. 35:325–332. characteristics. These might be used with a film that has differ- ent emulsion designs on each side to produce two different film- Asymmetric jaws screen combinations within one cassette. An example is where (Radiotherapy) Conventional linear accelerators have four jaws to one combination produces a desired contrast characteristic and collimate the photon beam and usually two opposite jaws move the other enhances detail within certain exposure ranges. Atom 62 Atomic excitation Atom The optical and x-ray emissions arise from electrons decaying A (General) An atom is a fundamental unit of matter composed of from an excited state to a lower energy state with the energy of a positively charged nucleus with negatively charged electrons the radiation corresponding to the difference in the energy levels orbiting around it. The nucleus is composed of both positively of the two levels. charged protons and electrically neutral neutrons. Atoms have Atoms in an excited state can be generated by a variety of equal numbers of protons and electrons; therefore, they are elec- means – such as heating, electrical discharge or stimulated by trically neutral. external radiation, e.g. a laser beam. Elements: Atoms are classified into elements by the number The emission of an electron occurs when the energy imparted of protons they contain. Elements are represented symbolically to it (by any of the aforementioned mechanisms) exceeds the by A Z X, where Z is the atomic number (number of protons), X is binding energy of the electron. When an atom emits an electron, the chemical symbol, and A is the atomic mass number which it becomes ionised. represents the number of particles (nucleons) in the nucleus. For A rather special situation is when the emissions arise example, the element carbon has six protons and six neutrons and from internal conversion or electron capture during a nuclear therefore has the symbol 12 6C. transformation. Isotopes: An element may be composed of atoms that all have Related Article: Nuclear transformation the same number of protons, i.e. have the same atomic number Z, but have a different number of neutrons, i.e. have different atomic Atomic Energy Commission (AEC) mass numbers A. Such atoms of identical Z but differing A are (General) Many countries have or have had an Atomic Energy called isotopes of a given element. Commission (AEC), in order to foster and develop peaceful devel- The term isotope is often misused to designate nuclear spe- opment of atomic science and technology, promote world peace cies. For example, cobalt-60, caesium-137 and radium-226 are not and improve public welfare. In some countries the commission is isotopes, since they do not belong to the same element. Rather still active while in Australia and United States it has been closed than isotopes, they should be referred to as nuclides. On the other or transformed. Australia (1958–1981), France (1945–present), hand, it is correct to state that deuterium (with nucleus called deu- Japan (1955–present), India (1948–present), Pakistan (1964–pres- teron) and tritium (with nucleus called triton) are heavy isotopes ent), United States (1946–1974). of hydrogen or that cobalt-59 and cobalt-60 are isotopes of cobalt. The United Nation Atomic Energy Commission (UNAEC) Thus, the term radionuclide should be used to designate radioac- was founded in 1946, and the first resolution adopted was calling tive species; however, the term radioisotope is often used for this for the peaceful use of atomic energy and elimination of weapons purpose. of mass destruction. The term nuclide refers to all atomic forms of all elements. Effort was put on the adoption of a resolution giving the The term isotope is narrower and only refers to various atomic power to the United Nations (UN) to impose controls on atomic forms of a single chemical element. development that would not be subject to UN Security Council In addition to being classified into isotopic groups (common veto. These controls would allow only the peaceful use of atomic atomic number Z), nuclides are also classified into groups with energy. Agreement was not reached on this resolution, mainly common atomic mass number A (isobars) and common number because the Soviet Union abstained on the proposal. Debate on of neutrons

(isotones). For example, cobalt-60 and nickel-60 are the plan continued until 1948, although it was clear that agree- isobars with 60 nucleons each (A = 60); hydrogen-3 (tritium) and ment was unlikely to be reached. The commission adjourned helium-4 are isotones with two neutrons each (A − Z = 2). indefinitely the debate. In this way ended in practice the activity If a nucleus exists in an excited state for some time, it is said to of the commission in 1948. be in an isomeric (metastable) state. Isomers thus are nuclear spe- In the same year 1946, the US Atomic Energy Commission cies that have common atomic number Z and atomic mass number (US-AEC) was founded. At the beginning US-AEC was given A. For example, technetium-99 m is an isomeric state of techne- extraordinary power and independency to carry out its mis- tium-99 and cobalt-60 m is an isomeric state of cobalt-60. sion, including freedom in hiring scientists and professionals. Related Articles: Atomic mass, Atomic mass unit, Atomic The National Laboratory system was established from the facili- number, Atomic weight, Electron, Elementary particles, Isotope, ties created under the Manhattan Project. The Argonne National Isotones, Neutrons, Nucleus, Proton, Radioactivity Laboratory was one of the first laboratories authorised under this legislation as a contractor. At the beginning, before the estab- Atomic attenuation coefficient lishment of the Nuclear Regulatory Commission (NRC), nuclear (Radiation Protection) The average factor by which individual regulation was also the responsibility of US-AEC. The US-AEC’s atoms within a medium attenuate the incident radiation is called regulatory programme sought to ensure public health and safety the atomic attenuation coefficient. from the hazard of nuclear power, without imposing excessive See further details in the articles on Attenuation and requirements that would inhibit the growth of industry. This was a Attenuation coefficient. difficult goal to achieve and an increasing number of critics pointed Related Articles: Attenuation, Linear attenuation coefficient, out the insufficiency of the programme in several important areas, Mass attenuation coefficient including radiation protection standards, nuclear reactor safety and environmental protection. In 1974 the programme was under such Atomic emissions pressure that the Congress decided to abolish the Commission. (General) The term ‘atomic emissions’ covers all electromagnetic Related Article: International atomic energy agency (IAEA) radiation and particles emitted from an atom apart from those Hyperlinks: AEC: www .aec .gov; NRC: www .nrc .gov emitted from the nucleus when undergoing a nuclear transforma- tion. The only particles emitted are electrons and the electromag- Atomic excitation netic radiation is in the optical (ultraviolet, visible and infrared) (Nuclear Medicine) Atomic excitation is the elevation of orbital and x-ray region of the electromagnetic spectrum. electrons from a lower energy state to a higher energy state. It is Atomic mass 63 Attenuation described by the Bohr atomic model in which electrons are dis- The quantity μ is the linear attenuation coefficient of the absorber tributed in shells with different orbital distance from the atomic for the particular photon energy and it has the unit of cm−1. μ is the A nucleus. In an unexcited atom the electrons are situated in the sum of the photoelectric coefficient, the Compton coefficient and lower energy states, i.e. the shells closest to the nucleus. Such the pair production coefficient. a system is referred to as the ground state. An electron can be The mass attenuation coefficient μm is obtained by dividing elevated to a higher energy state by absorption of energy, e.g. pho- the linear attenuation coefficient by the density of the absorber ρ: tons with appropriate energy; hence the atom is excited. The atom may return to the ground state by emitting a photon or an Auger m electron with a characteristic energy. mm = (A.25) r Atomic mass This has the unit of cm2/g. As the density effect is now factored (Nuclear Medicine) The atomic mass (ma) is defined as the sum out, this coefficient depends on the absorber atomic number Z and of the proton, neutron and electron rest masses in a specific atom. the photon energy E. The atomic mass should not be confused with relative atomic An important concept in the design of shielding for radiation mass, average atomic mass and atomic weight. The atomic or protection is the half value layer (HVL). This is defined as the molecular mass is often expressed in unified atomic mass units thickness of an absorber required to reduce the intensity of the (u). 1 u is equal to approximately 1.66 × 10−27 kg (Table A.4). photon beam to half of its original value. A nuclear medicine image is affected greatly by attenuation. Atomic mass unit Areas of radioactivity deeper within the body will appear to have (Nuclear Medicine) The unified atomic mass unit (abbreviated a lower count density than those nearer to the surface. In SPECT as u) is a unit used to express atomic and molecular masses. 1 the images will appear to have higher activity towards the periph- u equals 1.66 × 10−27 kg. ery of the patient. Attenuation correction is difficult to apply Related Articles: Atomic mass, Atomic number because the human body is composed of structures of different densities (e.g. lungs, soft tissue and bone). In modern SPECT sys- tems attenuation correction is achieved through the use of trans- Atomic number mission sources or a CT scanner. (Nuclear Medicine) The atomic number (or the proton number) is the number of protons in the atom nucleus. The atomic num- ber is typically denoted as Z. Each element has a specific num- Attenuation ber of protons, i.e. different atomic numbers. The mass number (Radiation Protection) When a beam of mono-energetic x- or of an atom is determined by the sum of protons and neutrons in γ-rays (i.e. photonic ionising radiation) passes through a mate- the nucleus. Atoms of the same element but with different mass rial (attenuator), it is found that equal thicknesses of the material number are referred to as different isotopes of the same element. reduce the incident beam by the same fraction of the initial value. The beam intensity is thus never reduced to zero. This process is known as attenuation. Attenuation of photonic radiation is an ATP (adenosine triphosphate) exponential process that can be represented by the attenuation (Magnetic Resonance) See Adenosine triphosphate (ATP) equation: Attenuation I = I e-mx 0 (Nuclear Medicine) When a beam of photons travels through matter, a proportion of these photons will be transmitted. Others where will be removed from the beam by photoelectric absorption, I0 is the initial beam intensity Compton scattering and pair production. This process is known I is the final beam intensity as attenuation. x is the distance travelled The extent of attenuation depends on the energy of the photon μ is the linear attenuation coefficient of the material (absorber) and the density and thickness of the absorbing matter. Consider the setup where a narrow beam of mono-energetic photons of initial The process by which the energy of a beam of ionising radia- intensity I0 passes through an absorber of thickness x. The transmit- tion is attenuated whilst propagating through a material has two ted beam has intensity I which is given by the following equation: components, absorption and scatter. The most important interac- tion process leading to absorption in the diagnostic energy range I = I e-mx 0 (A.24) (25–125 kV) is the photoelectric effect and the most important scattering process is Compton (inelastic) scatter. The linear attenuation coefficient can therefore be defined as the fraction of the incident beam that is attenuated per unit length TABLE A.4 of the material, or more correctively the probability that any par- Atomic Mass of the Particles (When the Atom Is ticular photon in the beam will interact with the atoms or mol- Motionless) That Constitute an Atom ecules of the material per unit length. The amount of attenuation in an absorber may be related to mr, Rest Mass Unified Atomic Mass Unit Kilogram MeV/c2 the mass or density of the absorbing material in terms of its mass Protons 1.007276 1.6726 × 10−27 938,272 attenuation coefficient, given by Neutrons 1.008 664 1.6749 × 10−27 939,573 Electron 5.485 799 × 10−4 9.1096 × 10−31 0.511 m r Attenuation 64 Attenuation coefficient where A μ is the linear attenuation coefficient, and ρ is the density of the material. I2 = 1020/10 = 100 I1 Finally, a factor can be defined which is the average attenua- The ultrasound intensity is reduced 100 times through 5 tion of incident radiation by a single atom within the absorbing cm of liver at the frequency 10 MHz. material – it is called the atomic attenuation coefficient. Related Articles: Linear attenuation coefficient, Mass attenu- ation coefficient, Atomic attenuation coefficient Related Articles: Absorption, Divergence, Damping, Intensity, Scatter Attenuation (Ultrasound) The intensity of a propagating ultrasound wave is Attenuation coefficient decreased by distance. This phenomenon is called attenuation and (Radiation Protection) When a beam of ionising radiation is inci- is due to scattering, reflection, divergence and absorption of the dent on a material, a number of interaction processes can occur ultrasound beam energy. that remove energy from the beam (attenuation). The fraction of The decrease in ultrasound intensity depends on the tissue prop- energy removed can be related to the distance or length travelled erties, the frequency and the distance. If the attenuation is expressed through the material, the mass or density of the material, or the in dB, it has been shown that the attenuation is almost proportional atomic density of the material. The terms used to describe this to frequency for most kind of tissues, Figures A.70 and A.71. The lost fraction are, respectively, the linear attenuation coefficient, attenuation coefficient for a specific tissue can be expressed in dB/ mass attenuation coefficient and atomic attenuation coefficient. cm/MHz. Some measured values of the attenuation coefficient for More generally the amount of energy removed from the beam a number of common tissues are shown in the table. may be called the attenuation factor. Related Article: Attenuation Tissue α (dB/cm/MHz) Water 0.02 Blood 0.15 Liver 0.4 Muscle 0.57 Bone 22 EXAMPLE How much will the ultrasound intensity be reduced when traversing 5 cm at 10 MHz through liver tissue? Attenuation = (0.40 dB/cm/MHz) ´ 5 cm ´10 MHz (a) (b) I =20 dB = 10 log 2 FIGURE A.71 Effect of attenuation in an ultrasound phantom. The I1 image at 3.5 MHz (a) shows better penetration than at 5 MHz (b). (Courtesy of EMIT project, www .emerald2 .eu) Attenuation Causes loss of intensity with depth Increases with increased frequency Measured in dB/cm/MHz 14 12 10 Muscle 8 6 Liver 4 2 Blood 0 1 2 3 4 5 6 7 8 9 10 Frequency MHz FIGURE A.70 Attenuation increase with increasing frequency. (Courtesy of EMIT project, www .emerald2 .eu) Attenuation dB/cm Attenuation correction in SPECT 65 Attenuation correction in SPECT Attenuation correction with events from the rod source. Since the spatial position of the (Nuclear Medicine) In planar imaging using a scintillation rod source is known it is also known which detectors are irra- A camera and in SPECT and PET, images are created by measur- diated. An increase in count rate in the irradiated detectors can ing photons emitted from a radionuclide distribution inside a therefore be measured. The advantage of this method is that it patient. The photon penetrates the patient and interacts within saves time, thereby decreasing the risk of patient movement. The the detector, generating a signal that can be measured and used last approach is to acquire both the transmission and the exam- to calculate a position and hereby an image. However, some ination image at the same time. This approach is used to save photons will also interact with the tissue in the patient. The valuable time. A rod transmission source can contribute to the interaction can be either absorption or scattering with or with- scattered and random coincidences in the acquired images and out an energy loss. Thus, the number of photons will be less that is a serious disadvantage. Modern PET scanners are usually as compared to the case assuming the radionuclide in air. This combined with a CT scanner (PET/CT). Attenuation correction is reduction of measured photons is caused by attenuation. The therefore applied using the CT data. attenuation is exponential and depends on

the tissue composi- Related Article: PET tion, density and the photon energy. In a complex geometry Further Reading: Cherry, S. R., J. A. Sorenson and M. E. such as the patient the attenuation is heterogeneous and the Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, magnitude depends on the angle for which the detector is posi- Philadelphia, PA, pp. 355–357. tioned. Since it is assumed that the image reflects the activity of the radionuclide in a particular volume within the patient Attenuation correction in SPECT using conjugate counting the presence of attenuation will change this activity estimate. (Nuclear Medicine) In SPECT imaging the number of photons This can result in a false defect or be interpreted as a reduction attenuated depends on the source depth (i.e. the thickness of tis- in metabolism. It is therefore generally desired to correct for sue the emitted photon travels through) and the attenuation prop- this effect by applying an attenuation correction. In practice, erties of the tissue. The transmission T of photons is described by this is not easy since in many cases the information of attenu- ating tissues is unknown. Modern scintillation cameras and N SPECT systems can have computed tomography capability so T = = e-mx (A.28) N that measurements of the anatomy (and related tissue composi- 0 tion and density) can be made accurately. where μ is the tissue attenuation coefficient Attenuation correction in PET x is the source depth (Nuclear Medicine) The attenuation correction is the most exten- N0 is the number of photons emitted from the source sive correction in PET. For an annihilation event at a depth x N is the number of photons transmitted through the tissue inside an object with thickness T to be registered as an event, both photons must pass through the object and be detected by the Without attenuation correction, sources with identical emis- detector. If assumed that they are both emitted in a direction that sion rates but at different tissue depths will register different allows them to reach the detector, the probability that they will count rates. One approach to deal with attenuation is to use con- reach the detector is the product of their individual probabilities: jugate counting. The technique involves the use of two opposite detectors acquiring simultaneously. Conjugate counting over a P = -mx ´ -m(T - x) -mT full 360° is combined to create a data set equivalent to 180° data det e e = e (A.26) acquired with a single detector. A source located close to one of the detectors will give a high and narrow response profile and μ is the linear attenuation coefficient and it varies with different low attenuation in the first detector and larger in the second. The tissue types (e.g. bone, lung). The probability of detection is only combination of conjugate counting data is performed using two dependent on the thickness of the object and not the position along methods: arithmetic mean and the geometric mean. the line of response (LOR). In PET an attenuation correction fac- Arithmetic Mean: The arithmetic mean is the summation of tor Ai,j for detector pair i and j can be calculated using a blank counts from along a line of response in the respective detector: image and a transmission image. The blank image is acquired using a rod source (68Ga) along the axis direction without a subject (I in the scanner. The rod source is captured using the gamma cam- 1 + I2 ) IA = 2 (A.29) era at projections from many angles. To produce the transmission image the same procedure is then repeated with a subject inside This simple summation provides a more accurate representation the scanner. The correction factor is given by of the activation distribution but there are still some residual posi- tion-dependency, i.e. sources in the centre of the object will have blank A i, j i, j = trans (A.27) a lower and broader response profile than an identical superficial i, j source. A better way to correct for photon attenuation is the geo- metric mean. where blanki,j and transi,j are the counts for ith and jth detector in Geometric Mean: Using the geometric mean the effects of each projection scan. The transmission scan should be performed source depths are practically eliminated. The combining of the prior to the injection of the radiotracer and it is also very impor- conjugation counts into a geometric mean is described by tant that the patient does not move between the transmission scan and the examination scan. Patient movement can lead to serious artefacts which can manifest as a high or low regional uptake. IG = I1 ´ I2 (A.30) Another approach is to acquire the transmission image after the examination scan. In such a case the residual activity can interfere Consider the arrangement seen in Figure A.72. Attenuation correction in SPECT using transmission 66 Attenuation correction in SPECT using transmission scans D thickness. Another approach is to use the contours of an initial A reconstructed image f′(x,y) to estimate the path lengths for each pixel for all projections. The ACF is then calculated for each pixel according to a b ACF ( 1 x, y) = N (A.34) (1 / N )å e-mdi Detector 1 Detector 2 i=1 FIGURE A.72 A point source at source depths a and b inside an attenu- where di is the attenuation path length for a pixel in projection ating material with thickness D. The data are acquired using two detectors i (total of N projections) and μ is the tissue attenuation coeffi- (conjugate counting). To compensate for attenuation the collected data can cient (again considered constant). A new corrected image f(x,y) be combined into a single data set using the arithmetic or geometric mean. is created by multiplying the reconstructed image f′(x,y) by the ACF(x,y) on a pixel-by-pixel basis: For photons directed towards detector 1 (upper) or 2 (lower) f ( x, y) = f ¢(x, y)´ ACF (A.35) the attenuation probability is This technique is known as Chang’s multiplicative method. I = I ´ e-ma 1 01 (A.31a) A more complex implementation for the Chang method is to do a forward projection of the f(x,y) image (see article Filtered I = I ´ e-mb 2 02 (A.31b) back-projection) in order to attain ‘attenuation’ projections. These projections are subtracted from the original measured projection profiles to form a set of error projections. The error projections The geometric mean of the counts registered by each detector is are reconstructed using filtered back projection to form an error image ferror(x,y). With this error image the final attenuation cor- I ´ I = I ´ I e-mD /2 1 2 01 02 (A.32) rected image is As seen in (A.31) the geometric mean depends on the object thick- f ( x, y) = f ¢(x, y)´ ACF + ferror ´ ACF ( x, y) (A.36) ness D and not on source depths a and b. The geometric mean is accurate when correcting for attenuation from a single radioac- The Chang method is relatively successful in regions where the tive source. In clinical cases administered activity is often accu- attenuation coefficient is relatively constant, e.g. head and abdo- mulated in a number of volumes (e.g. different organs). For such men. But this method has serious limitations when imaging vol- cases the attenuation correction is favourably performed with the umes with high density gradients, e.g. lungs and pelvic region. Chang correction method (see separate article). For these regions it is more suitable to use transmission scans and Further Reading: Cherry, S. R., J. A. Sorenson and M. E. attenuation maps (see separate article) in the Chang method. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Related Articles: Attenuation correction in SPECT using con- Philadelphia, PA. jugate counting, Attenuation correction in SPECT using transmis- sion scans, Filtered back projection, SPECT Attenuation correction in SPECT using the Chang method Further Readings: Chang, L. T. 1978. A method for attenu- (Nuclear Medicine) A problem in SPECT imaging is the fact that ation correction in radionuclide computed tomography. IEEE the probability of a photon reaching a detector depends on the dis- Trans. Nucl. Sci. 25:638–643; Cherry, S. R., J. A. Sorenson tance travelled in an attenuating object. Consider two sources, a and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., superficial and a deeper lying source. Even if the two sources have Saunders, Philadelphia, PA, p. 312. identical emission rates, the number of registered events will be higher from the shallow source because of the extensive attenua- Attenuation correction in SPECT using transmission scans tion of photons from the deep source. (Nuclear Medicine) A problem in SPECT imaging is the fact that For a single source, attenuation is corrected using conjugate the probability of a photon reaching a detector depends on the dis- counting (see separate article). The technique uses two opposite tance travelled in an attenuating object. Consider two sources, a scintillation cameras and by combining the two views into one superficial and a deeper lying source. Even if the two sources have (geometric mean), the effects of attenuation are minimised. This identical emission rates, the number of registered events will be technique is successful when imaging a point source, but in clini- higher from the shallow source because of the extensive attenua- cal imaging the source is seldom (or never) a point source. To tion of photons from the deep source. attain quantitative accuracy in clinical imaging a simple method Numerous approaches to deal with the decrease in count rate could be used which involves an attenuation correction factor for deep lying sources have been suggested; conjugate counting (ACF). The attenuation correction is given by multiplying the pro- and Chang method are both discussed in separate articles. Both jection profiles by an ACF using the geometric mean: these approaches have limitations, e.g. the Chang method fails to correct for attenuation effects in volumes with high density gra- 1 = mD /2 dients. Using transmissions scans can lead to a satisfying correc- ACF e-mD /2 = e (A.33) tion where other methods fail. The transmission scan is an additional scan, typically per- where μ is the linear attenuation coefficient, which is assumed to formed on the detector system intended for the emission scan. be constant for the entire object. D is the object diameter or tissue Projections are acquired with an external source, typically a flood Attenuation depth 67 Attenuation steps or a line source and in some cases with an x-ray tube and a detec- range). For electromagnetic ionising radiation attenuation follows tor mounted next to the scintillation detectors on the camera. an exponential law and attenuation depth is not a useful concept. A Data acquisition from two scans are acquired; one without an Related Article: Electron attenuation object (blank or reference scan) and one with an object. The rela- tionship between the blank (Iblank) and transmission counts (Itrans) Attenuation equation depends on the exponential behaviour of the photon attenuation: (Radiation Protection) Attenuation of photonic radiation through a material is an exponential process that can be represented by the I I e x trans = -m blank ´ (A.37) attenuation equation: I = I e-mx where μ is the linear attenuation coefficient. The natural loga- 0 rithm of the quotient between the two counts is where I0 is the initial beam intensity æ I ln blank ö ç ÷ = mx (A.38) I is the final beam intensity è Itrans ø x is the distance travelled μ is the attenuation coefficient of the material (absorber) These projection profiles represent the sum of all attenuation coefficients along a line of response Related Articles: Attenuation, Attenuation factor, Radiation protection mx = åmiDxi (A.39) i Attenuation factor (Radiation Protection) where Δxi represents the portion of the line of response that runs The attenuation factor is the factor by which the intensity of through the ith pixel and μ is the linear attenuation coefficient in a beam of ionising radiation is reduced in travelling through a the i:th pixel. A map with a pixel-specific attenuation coefficient particular medium. Given an incident radiation beam of energy E, is called an attenuation map. The attenuation map could be used for an attenuating element, m, of thickness t, its attenuation factor, in the Chang

method to get a more accurate correction. But the fat(m,t,E) can be defined by the following relation: most common use of the attenuation map is in iterative recon- struction methods. Iout = fat (m,t,E)* Iin As previously mentioned the transmission scan is acquired using a flood source, line source, multiple line sources, a moving where Iin, and Iout are, respectively, the incident and the exiting line source or an x-ray tube. An important source characteristic radiation intensities. is the energy of the emitted photons. The line source emission Related Articles: Attenuation, Attenuation coefficient, energy must differ from the energy used for imaging if the two Attenuation equation projections are to be acquired simultaneously, using two energy windows. A radionuclide with long half-life as a line source is Attenuation steps preferred or else the source must be frequently replaced. Two suit- (Diagnostic Radiology) The test of radiographic contrast ability of able radionuclides are 153Gd (T1/2 = 242 days, Egamma = 97 and 103 one x-ray system requires tools with specific attenuation steps. A keV) and 123mTe (T1/2 = 120 days, Egamma = 159 keV). typical tool is an aluminium (or copper) step-wedge (Figure A.73). Consider simultaneous acquisition of the transmission scan, The absorption and thickness of each step is calculated to produce using a 123mTe rod source and emission scan, using 99mTc (E specific attenuation of the x-ray beam. The radiograph of these gamma = 140 keV) with two energy windows over the respective photo attenuation steps produces specific optical densities of the x-ray peaks. A number of counts in the transmission scan will inevi- film – a contrast scale. Such tools are used for quality control of tably be Compton scattered photons emitted by a 99mTc radionu- x-ray radiographic and fluoroscopic equipment. clide. This contribution to the transmission counts is referred to as downscatter. Downscatter can be avoided by acquiring the two scans sequentially instead of simultaneously. Abbreviation: SPECT = Single photon emission computed tomography. Related Articles: Attenuation correction in SPECT, Attenuation correction in SPECT using conjugate counting, Attenuation correction in SPECT using the Chang method, Iterative reconstruction methods, Downscatter Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA, pp. 313–315. Attenuation depth (Radiation Protection) The attenuation depth in a given medium is the thickness of matter which is needed to absorb completely all the energy of the incident radiation. The attenuation depth is a FIGURE A.73 Aluminium step-wedge with 14 attenuation steps made finite thickness for charged particles such as electrons (called the for radiographic quality control. (Courtesy of Ing. A Litchev.) Attenuator 68 Auto-contouring Auger effect A (Radiation Protection) The Auger effect describes the process in which Auger electrons are produced. Related Article: Auger electron Auger electron (Radiation Protection) If an inner (K-shell) electron is removed from an atom through some form of ionisation process, the vacancy will be filled by an electron from one of the outer elec- tron shells. The excess energy will be released from the atom in the form of a photon whose energy is equivalent to the difference in the binding energies between the K- and L- (or other) shell electrons. This is the origin of characteristic x-rays from an x-ray FIGURE A.74 Contrast scale from fluoroscopic tool with attenuation tube. steps. (Courtesy of EMERALD project, www .emerald2 .eu) In a smaller number of ionising events the K-shell vacancy will be filled by an outer shell electron but instead of the excess energy being released as a photon, an outer shell electron is emit- One tool used for testing the contrast scale in fluoroscopy is the ted. This is known as the Auger effect and the emitted electron Leeds Test Object TO GS2. This tool uses very small attenuation is an Auger electron. The Auger effect was named after Pierre steps to assess precisely the greyscale (contrast scale) of image inten- Victor Auger who explained the effect in 1925. sifiers. Figure A.74 shows contrast scale produced with TO GS2 The ratio between the number of x-rays per de-excitation and from a digital fluoroscopic system allowing a profile of the contrast Auger electrons per de-excitation is called fluorescent yield. The steps (density profile) to be presented on the screen. The lack of vis- fluorescent yield depends on the atomic number of the disintegrat- ible contrast steps in the dark part of the greyscale (on left) shows ing nucleus and the yield is low for elements with high atomic that this system has intrinsic contrast limitation, which cannot be number and high for elements with low atomic number. improved with the ‘contrast’ and ‘brightness’ settings of the monitor. Auger electrons are low energy electrons. Auger spectroscopy Related Articles: Attenuation, Step wedge, Quality control, is a useful analytical tool to determine the composition of materi- Radiography, Fluoroscopy, Image intensifier als. In medicine a number of radiopharmaceuticals, most notably Hyperlinks: EMERALD: www .emerald2 .eu technetium-99 m, emit auger electrons. These contribute to the absorbed dose in a way that is not well understood. Attenuator Related Articles: Auger effect, Photoelectric effect, (Radiation Protection) Any material that absorbs or scatters the Characteristic x-rays energy incident in a beam of radiation is called an attenuator. Related Article: Attenuation Augmented reality (General) Augmented reality (AR) is a powerful method of con- Audit temporary learning process (in particular, e-learning). It enhances (General) An audit is a process of inspection and the checking the study of real methods and equipment by adding to these vir- of biomedical equipment services, such as maintenance, process, tual objects. AR devices such as smart glasses, head-mounted dis- program or system conformity with well-defined quality require- plays and others already have a place in medical education. Areas ments. Conformity must be documentable and recorded (docu- such as modern training in minimally invasive interventions use ments or reports analysis, process actions analysis, interviews AR, and there is ongoing research on including AR in medical with healthcare staff, scientific evidence). An audit is necessary to physics fields as radiation protection, imaging and radiotherapy. evaluate efficacy and efficiency and then output and performance Augmented reality devices and methods differ from virtual of a medical device, process or system. Conformity to prefixed reality (VR) devices and methods, as the latter does not have the standards is the result of focused strategies and organisational component of real visualisation. Instead, VR is completely based actions. An audit can be periodic: the inspectional checks are on virtual information and adds an imaginative educational layer repeated to guarantee high performance level and quality assur- to the learning process. ance of medical devices and services every day. Related Articles: e-Learning Further Reading: Peters, T. M., C. A. Linte, Z. Yaniv and J. Audit, quality audit Williams. 2018. Mixed and Augmented Reality in Medicine, CRC Press, Series in Medical Physics and Biomedical Engineering, (Radiation Protection) An assessment of the systems and proce- ISBN 9781138068636 Skin exposure to solar or artificial light is dures and the adherence to those procedures within a department usually expressed as an irradiance. using radiation. An assessment of systems and procedures would check compliance against legal requirements (e.g. national ionis- ing radiation regulations) as well as accepted codes of practice. Auto-contouring Assessing adherence to local protocol is achieved through vari- (Radiotherapy) This is a computer-assisted contouring technique ous quality control metrics, such as auditing patient administered in which the structure of normal tissue organs and treatment tar- doses, image quality and equipment maintenance. gets can be delineated automatically by means of computer soft- There must be appropriate mechanisms in place to review the ware tools. The principle of the technique is based on differences results of quality audits, to feed back those results and apply cor- in CT numbers of different tissue structures when contouring in rective actions, and to ensure the actions are completed. CT images. When using MR imaging, the technique is based on Related Articles: Quality control, Quality assurance intensity level of MR signals emitted from different tissues. Autocorrelation 69 A utomatic brightness control Autocorrelation The system for automatic brightness control (ABC) is used (Ultrasound) Autocorrelation is a measure of how well a signal in fluoroscopy and assures consistent brightness of the image A matches a time-shifted version of itself. This is used in signal produced by the image intensifier (II). The system is also called processing to find periodicity in a signal, e.g. A more strict brightness control. definition is that the autocorrelation R(t1,t2) of a signal x(t), is the The ABC produces x-ray images with constant brightness, expected value of the product x(t1)x(t2), i.e. E{x(t1)x(t2)}. Of special irrespective of the material which is being x-rayed (i.e. patient interest is the value along the diagonal t1 = t2 = t which is the thickness or absorption). It consists of a detector (sensing the II average power of x(t). output luminance of the II) and a feedback system, which changes In diagnostic ultrasound autocorrelation is mostly commonly the fluoroscopic parameters (kV, mA or both – depending on the known as the name of an algorithm used to estimate blood flow type of the system) so that optimal brightness of the fluoroscopic velocity. The estimate is not the result of a correlation per se, but image is maintained. The most often used sensors are photodi- the calculation of the discrete autocorrelation is used to give an odes or other photo sensor, or a photo-multiplier coupled at the estimate of the phase shift between two successively transmitted output of the II. pulses, reflected off a moving target. In some systems the video signal can also be used to drive the In the derivation of this estimator, the starting point is an feedback (this should not be confused with the video automatic expression that relates the estimated velocity along the ultrasound gain control, which is related only to the imaging system). propagation direction to the phase change between two successive The ABC changes the x-ray parameters (kV and mA) accord- pulses. The end result is an expression that includes the arcus tan- ing to the transparency of the object and most often the operating gent of a quotient between sample values for the imaginary and panel shows patient with various thickness (Figure A.75). real parts of the two pulses. It turns out that this can be more com- Figure A.76 shows a block diagram of a typical ABC system pactly written as the autocorrelation evaluated at lag 1, namely as for x-ray fluoroscopy. In principle it has similarities with the automatic exposure control (AEC) system used in radiography. The ABC includes feedback block C1 (most often including an c æ Im (R (1)) ö vz = - arctan ç ÷ integrator and comparator), which can take signal either from 4pf 0Tprf ç è Re(R (1)) ÷ ø a dosimeter (ionisation chamber – D2) placed between the anti scatter grid (AS) and the image intensifier, or from a photo sen- sor (D1) which measures the overall brightness at the output of where the II (before the TV camera). Although both systems would vz is the estimated velocity along the ultrasound propagation have very similar effect, the system using D2 effectively moni- direction tors the dose rate, while the system using D1 monitors the image c is the sound speed brightness. The ABC will vary the kV and mA of the equip- f0 is the transducer centre frequency ment until the measured signal equals a preset value. The block Tprf is the pulse repetition time diagram shows x-ray equipment with a high frequency genera- Im() denotes the imaginary part tor, meaning that the feedback will change the frequency of the Re() is the real part DC–AC converter. R(m) is the discrete autocorrelation for lag m The change of the mA and kV can be very simple (i.e. step change of kV plus internal step changes of the mA), but contem- Using N lines as an estimate for the autocorrelation function, porary fluoroscopic equipment use microprocessor-controlled R(1) becomes operational characteristics (linear on non-linear dependence of kV and mA). These characteristics change the kV and mA in a N -2 1 way to achieve the required brightness and at the same time to Rˆ (1) = r* (i)r (i +1) N å influence other parameters (contrast, dose, etc.) Various manu- i=0 facturers apply

different ABC operational characteristics. For where r(i) is the complex (quadrature demodulated) received sig- nal, x(i) + jy(i). The asterisk denotes complex conjugation. Autofluoroscope (Nuclear Medicine) Autofluoroscope is an early (1963) concept of multiple-crystal scintillation camera, introduced by Bender and Blau. This early device has used 294 discrete NaI crystals arranged in a matrix of 21 × 14 (each crystal 8 × 8 cm and 4 cm thick). Related Article: Scintillation camera Further Reading: Hendee, W. and E. R. Ritenour. 2002. Medical Imaging Physics, Wiley-Liss, New York. Automatic brightness control (Diagnostic Radiology) The two main factors in determining the amount of x-ray output required to produce a diagnostically acceptable image are the size and therefore attenuation of the patient, and the sensitivity of the detector in terms of quantum efficiency – i.e. how well the detector converts x-ray photons to a FIGURE A.75 Typical control panel of an image intensifier. Note the signal in the image whilst keeping noise minimised. ABC panel in the upper right corner (medium patient thickness selected). Automatic circuit breaker 70 Automatic exposure control (AEC) Set A P C1 value HF x-ray generator AS D2 D1 DC–AC Image converter intensifier TV Patient FIGURE A.76 Block diagram of fluoroscopic equipment with ABC. The feedback C1 can have two main options – taking signal from D1 (dosimeter) or D2 (photometer). example, Siemens names some of its curves as Isowatt (uses high includes detector, negative feedback and regulator. In the case dose), Anti-isowatt (makes compromise between dose and image of ABC, these are, e.g. photo detector, comparator and kV/mA contrast), minimal radiation (used for paediatric fluoroscopy), regulator. high-image-contrast (used in interventional radiology), etc. Related Article: Automatic brightness control The values of the kV and mA are directly linked to the x-ray transparency of the examined object, because the ABC is adjusted Automatic dose rate control to maintain a certain value of the II input dose rate (e.g. 0.3 (Diagnostic Radiology) The system for automatic dose rate con- μGy/s). Decreasing the II light output – i.e. II dose input – imme- trol is used to control the entrance dose rate in front of an image diately leads to increase of the kV, mA or both in order to keep intensifier (II). It comprises of an ionisation chamber (as detec- the same II input dose rate. However this is also linked with the tor) and feedback system controlling the kV and mA. Most often II field sizes. When a smaller II field size is chosen (e.g. using the the system is used to control brightness of the II output image. magnification 15 inch field of a 30 inch II), then the II light output Typical values of dose rate controls (in front of the II) are between decreases and the ABC boosts the dose up. Each reducing of the 0.15 and 0.30 μGy/s. For more details see the block diagram in the visible II field size (magnification, or zoom mode of the II) leads article on Brightness control. to increase of the dose to the patient in the observed field. This Related Articles: Brightness control, Image intensifier increase depends on the ABC system, but can reach more than five times. Related Articles: Automatic exposure control (AEC) Fluoroscopy, Image intensifier, Automatic (Diagnostic Radiology) The system for automatic exposure con- dose rate control, Automatic exposure control, Brightness control Further Reading: trol (AEC) has been developed during the time of x-ray films, Oppelt, A. (ed.). 2005. Imaging Systems for which have been very sensitive to the exact radiation exposure Medical Diagnostics, Siemens, Erlangen, Germany. necessary for producing optimum contrast of the radiograph. The system continues to be used with digital detectors, deliver- Automatic circuit breaker ing optimal dose to the detector (despite the fact that the contrast (General) See Circuit breaker of the digital radiograph depends mainly on the correct use of window parameters for visualisation). At the time of X-ray films, Automatic collimation control the AEC system was used predominantly for chest screening (Diagnostic Radiology) Some collimators (beam restrictors) used (specifically for tuberculosis). For this reason, the system uses in radiography have a sensor which detects the size of the detector three radiation measuring detectors, positioned in a way to facili- (film cassette) and automatically restricts the beam to this size. tate the visualisation of lungs. The description of AEC will use This is achieved by exact movement of the lead jaws of the dia- the term film, as a synonym of the detector of any type (digital phragm (most often the Bucky diaphragm). This system is also or analogous). known as positive beam limitation device (PBL). Most contemporary radiographic equipment is equipped with The x-ray fluoroscopic systems use collimator with circular an automatic exposure control (AEC) system. This system con- shutter. All these systems have automatic collimation control sists of a radiation detector and feedback loop, which interrupts which changes the diameter of the x-ray field to match the image the exposure when certain predetermined x-ray exposure (dose intensifier field size. level) is reached. The main purpose of the AEC is to ensure con- Related Articles: Beam restrictors, Filter compensating, sistent optical density (darkness) of the x-ray film, independent of Diaphragm, Collimator, Image intensifier the overall x-ray absorption of the x-rayed object, by varying the x-ray exposure. In digital systems, which do not produce film, the Automatic control system AEC ensures a relatively constant detector dose and noise, and (General) All imaging equipment apply various automatic con- reduces the likelihood of dose creep. Sometimes this system is trol systems. One such system controls specific parameter of the called auto-timer. imaging chain by varying another linked parameter. For example There are many different types of AEC systems. Often the – controlling the brightness of the diagnostic image by varying difference relates to the type of detector used (ionisation cham- the output dose of an x-ray fluoroscopic equipment (automatic ber, photo-timer or solid-state detector). The AEC with ionisa- brightness control, ABC). In general an automatic control system tion chamber detectors are widely introduced in the radiographic Automatic exposure control (AEC) 71 A utomatic exposure control (AEC) practice. These often consist of a stand with several detectors, today are made using AEC – i.e. there is no other control of the known also as ‘dominants’ or ‘cells’, placed in the patient table exposure parameters than this automatic system. A or stand, immediately in front of the film cassette. Due to the The AEC systems sense the dose (dose rate) of the exposure, fact that AEC have been most often used for chest exposures, the compare it with the pre-set value, take into consideration the sen- widely used configuration is with three dominants, the left and the sitivity of the film/screen combination and on this basis interrupt right approximately at the position of the middle parts of the lungs the exposure after a certain period of time. Due to the fact that the and the central dominant below them, at the position of the spine. AEC measures all radiation, which reaches it, it also measures This type of AEC is mounted at the vertical (chest) radiographic scatter radiation from the patient. In order to compensate this, stand (Figure A.77). an optical density correction system is used. This system is often The AEC is controlled through the operator panel connected with a ±D switch at the AEC control panel (often D (Figure A.78). It allows choosing different combinations of stands for ‘dunkel’ – ‘darkening’ in German). In the case of mam- active detectors; mean optical density (darkening) of the film, mography such compensation is very difficult. This is due to the etc. Sometimes the AEC operates through the so-called ‘organ- fact that the AEC detector is placed behind the film (to minimise automatic’ or ‘anatomic programming’. This AEC system uses a the absorption of the low energy radiation used). In this case the microprocessor with stored data for the most effective exposure transparency of the object (the ratio between Nex – the number of parameters for radiography of each anatomical organ. Quality quanta exiting from the object, and Nin – the number of quanta control of AEC is of great importance, as most of the radiographs entering the object) depends solely on the absorption of the object. This is so, because of the compression device, which compresses the size of the breast tissue to an averaged thickness to reduce scatter, and improve uniformity of absorption. In this case a spe- cial microprocessor system is used to automatically correct the optical density of the film. Figure A.79 shows two typical AEC systems. The upper AEC feedback (through C1) is the most often used system for chest stands. Its detector D is most often a thin ionisation chamber (or several chambers) placed between the anti scatter grid (AS) and the film-screen combination (S/F). The signal from D passed through C1, which is comprised of an integrator and comparator. This way C1 integrates the summary signal from D (i.e. measures dose per time – dose rate) and compares it with a set value. The set value depends on the sensitivity of the film, the necessary darken- ing, etc. When the measured summary value (equal to the overall x-ray exposure) reaches the set value C1 signals the x-ray genera- tor and interrupts the exposure. The block diagram shows x-ray equipment with high frequency generator, but in simple x-ray FIGURE A.77 Vertical radiographic stand with three AEC chambers (dominants). The drawer for film cassette is open – cassette loading. equipment this can just be an interrupting switch. When this AEC type uses semiconductor detector, it should be placed behind S/F, as otherwise its shadow will be seen on the x-ray film. The lower AEC feedback (through C2) is the system most often used in mammography. Its detector D can be either semiconduc- tor or a thin ionisation chamber. Usually there are two detectors D with a special absorbent filter F placed between them. This way the AEC also takes into account the energy spectrum of the x-rays (very important for the contrast in mammography). The two sig- nals from D are analysed by a special processor P, which sup- plies C2 with signal corresponding with the beam quality (x-ray Set C1 value HF x-ray generator D AS S/F DC–AC converter ~ F D D Patient C2 P Set value FIGURE A.78 Typical AEC control panel. Manual setting of AEC parameters: chamber (dominant) selection and number; chamber (domi- FIGURE A.79 Block diagram showing two typical AEC types. C1 is nant) sensitivity and darkening. used for chest radiography and C2 used for mammography. Automatic film processor 72 Automatic timer spectrum). This signal, together with the signal from the foremost light output can be taken from the II output), as in this case it can A detector, is analysed by C2 (which again comprises of an integra- also play the role of brightness control. tor and comparator). This way C2 integrates the summary signal However the amplification of the video signal is also linked from D and the signal from P, and compares these with a set value. with the image noise (hence signal to noise ratio, SNR), as The set value depends on the sensitivity of the film, the necessary increasing the sensitivity or the gain (amplification coefficient) darkening, etc. When the measured integrated signal (equal to the leads to unwanted increase of the noise amplitude. This way the overall x-ray exposure) reaches the set value C2 signals, the x-ray AGC system is not directly linked to the patient dose, but is very generator interrupts the exposure. important for the final image quality. Normally the system moni- Related Articles: Automatic dose rate control, Chest radiog- tors the signal from the central part of the visible area (dominant) raphy, Mammography of the II and reacts to Further Reading: Oppelt, A. (ed.). 2005. Imaging Systems for Medical Diagnostics, Siemens, Erlangen, Germany. • The maximal (peak) value of the video signal • The minimal value of the signal Automatic film processor • The mean value of the signal (Diagnostic Radiology) An automatic film processor is a device that processes radiographic film by transporting it through four The AGC has special importance for digital fluoroscopic systems steps: development, fixing, washing, and drying as illustrated in (and DSA), which can be with either peak-sensitive or mean-sen- Figure

A.80. sitive AGC. One problem with monitoring the mean (summary) The several factors that affect the processing are automatically value of the video signal is related to the fact that some specific controlled. These include the time in the developer, temperature contrast differences can be omitted. Monitoring the minimal and replenishment of the developer solution. value of the video signal can lead to areas of the image with over saturation (too bright). Monitoring the peak (maximal) level of the Automatic frequency control signal will prevent the monitor from saturation, but some darker (Diagnostic Radiology) Automatic frequency control is a sys- parts of the image will become even darker, therefore affecting tem in high frequency x-ray generators, which controls the fre- the overall contrast. Due to this reason most contemporary AGC quency of the DC to AC converter, as this frequency is directly monitor several sensitive regions of the image. related to the kV and other parameters of the exposure. If one Related Articles: Fluoroscopy, Image intensifier, Automatic of these parameters changes during the x-ray exposure, a nega- dose rate control, Automatic exposure control tive feedback triggers the automatic frequency control, which Further Reading: Oppelt, A. (ed.). 2005. Imaging Systems for changes the frequency accordingly in order to keep this param- Medical Diagnostics, Siemens, Erlangen, Germany. eter constant. Related Article: High frequency generator Automatic kV reduction (Diagnostic Radiology) See Radiographic kV control Automatic gain control (Diagnostic Radiology) Apart from the brightness control and Automatic line voltage regulation dose rate control systems, which keep the image intensifier (II) (Diagnostic Radiology) Automatic line voltage regulation is nec- light output constant, a second circuit is incorporated in the TV essary to compensate the voltage drop in most x-ray radiographic system of fluoroscopic equipment. This system (automatic gain systems (see the article about Voltage drop). Usually the system control, AGC) controls the video gain of the Video amplifier. This regulates the autotransformer output voltage (in classical x-ray amplifier is immediately after the TV camera and assures good generators), or the frequency of the DC–AC converter (of high amplitude of the video signal to the TV monitor of the equipment. frequency x-ray generators). The AGC system is of special importance when the image inten- Related Articles: Voltage drop, High voltage generator, High sifier is connected to the TV camera with fibre optics (hence no frequency generator Automatic multiple-sample systems of NaI(Tl) well counters (Nuclear Medicine) For measurements of multiple radioactive Film path samples as tissues, biopsies and blood from patients or experi- mental animals a multiple sample system is typically used. The system is usually based on a NaI(Tl) scintillation crystal (how- ever, also semiconductor systems are in use) coupled to a single or multiple channel analyser. One or several isotopes with different photon energies can be analysed. The system is equipped with a mechanical sample changing system in that several samples can be loaded in the machine and then counted in sequence. The results can then be fed into a computer with analysing program. Related Articles: Sodium iodide crystal, Scintillation detec- tor, Radioactive sample, Well counter Developer Fixer Wash Dryer Automatic timer FIGURE A.80 An automatic film processor showing how the film is (Diagnostic Radiology) Automatic timer is a term which is transported through the four steps. In a typical processor the total pro- sometimes used instead of automatic exposure control (AEC). cess requires approximately 90 s. (Courtesy of Sprawls Foundation, www Contemporary AEC systems use microprocessors which vary the .sprawls .org) x-ray exposure in a specific way; however, the early AEC systems Automatic tube current modulation 73 Automatic tube current modulation control only the duration of the exposure aiming to achieve spe- reduced attenuation. However, in order to achieve dose optimi- cific optical density (darkening) of the x-ray film. These systems sation, the user must select an appropriate image quality level. A have used pre-set mA and kV and have only varied the time of the Different methods are used to specify image quality. Some sys- exposure, hence their name ‘automatic timer’. tems require an input of noise level (standard deviation of CT Related Article: Automatic exposure control number), whereas others require a reference mAs relating to a standard patient. Automatic tube current modulation (Diagnostic Radiology) Automatic tube current modulation (ATCM) is the method used in CT to automatically adjust tube current for variations in patient attenuation. It is also referred to as ‘mA modulation’, ‘dose modulation’, ‘automatic tube current control’ or ‘automatic exposure control (AEC) in CT’. The pur- pose of ATCM in CT is to achieve the desired image quality for all patients and at the same time optimise radiation dose. Different levels of ATCM are available on CT scanners. Firstly, the tube current (mA) may be adjusted for overall patient size and kept constant throughout the scan (Figure A.81a). Secondly, the mA may be adjusted with z-axis position, for each x-ray tube rota- tion (Figure A.81b). Thirdly, it may be adjusted to account for variations in attenuation throughout a rotation, particularly for differences between the anterior–posterior and lateral (left-right) patient dimensions (Figure A.81c). Most commonly, all three lev- els of ATCM are combined for greatest effect (Figure A.81d). Prior to performing a scan with ATCM, information must be available relating to patient attenuation. This is usually obtained from the scan projection radiographs (SPRs), usually referred to by their trade names, such as scanogram, scout view or topogram (Figure A.82). For rotational modulation (Figure A.81c) some systems employ so-called ‘on-line’ modulation where the attenu- ation in the first 180° of rotation is used to adjust the mA in the Attenuation subsequent 180°. ATCM not only results in more uniform image quality FIGURE A.82 Patient attenuation variation along z-axis obtained from from patient to patient and along the patient, but it is also a scan projection radiograph. (Courtesy of ImPACT, UK, www .impact dose optimisation tool as a lower mA is applied in regions of scan .org) (a) z (b) z (c) z (d) z-axis position FIGURE A.81 Different levels of ATCM (a) patient size, (b) z-axis, (c) rotational and (d) all three combined. (Courtesy of ImPACT, UK, www .impactscan .org) mA mA mA mA z-axis position Autoradiogram 74 Avalanche ionisation in Geiger–Müller counter ATCM systems are usually referred to by their trade A names such as SmartmA/AutomA, DoseRight, CAREDose or SureExposure. α Autoradiogram (Nuclear Medicine) An image made from placing an object con- taining a radioactive substance or substances on a photographic plate or film, or by coating the object with photographic emulsion. The image is formed by exposure of the plate, film or emulsion to Electric field radiation emitted from the object. Digital autoradiography is performed by using solid-state detec- FIGURE A.83 Scheme of the first Townsend coefficient for the gas as tors or scintillation materials. For scintillators CCDs or image inten- function of the electric field. sifiers are used for detecting the emitted light from the scintillator. High resolution autoradiography can be obtained by electron microscopy autoradiography. Usually a cylindrical geometry of electrodes is used in gas pro- For detection beta particles, conversion electrons or alpha par- portional counters. In this case the electric field increases in the ticles are commonly used. same direction as the avalanche progresses. For that reason the Abbreviations: ARG = Autoradiography and CCD = Charge avalanche growth is stronger than in uniform electric field. coupled device. The charge amplification resulting from the avalanche process results in better signal to noise characteristics in proportional Autoradiography counters than in ionisation chambers. (Diagnostic Radiology) A process which produces image record Related Articles: Gas-filled radiation detectors, Proportional (on film) of material or tissue which is radioactive. The process can counter be applied both at macroscopic and microscopic examinations, Further Reading: Knoll, G. F. 2000. Radiation Detection and usually through the use of radioactive isotopes. An example of Measurement, 3rd edn., John Wiley & Sons, Inc., New York, pp. autoradiography use is to record (in vivo or in vitro) radiolabelled 159–160. tissue. Autoradiography (or digital autoradiography) is rarely used. Avalanche ionisation in Geiger–Müller counter Autotimer (Radiation Protection) In a Geiger–Müller gas-filled detector, an (Diagnostic Radiology) See Automatic timer ‘avalanche’ is caused when the accelerated electrons strike the anode causing the emission of UV (ultraviolet) radiation. These Autotransformer UV photons produce photoelectrons as a result of interacting (Diagnostic Radiology) See High-voltage generator with the gas molecules and the detector walls. The photoelectrons strike against the anode and produce more UV photons. In this Avalanche gain way the avalanche is extended and covers the whole length of the (Radiation Protection) The avalanche is formed in gas propor- anode (Figure A.84). Each avalanche can create new avalanches tional counters when the electric field is over a threshold value. with a multiplication factor of 1010. In order to detect any new This threshold value depends on the gas and its pressure, e.g. at radiation entering the GM counter, this continuous discharge must atmospheric pressure it is about 106 V/m. In this case the electrons be stopped. This process is called quenching. There are various and ions created in gas by the ionising radiation are accelerated by methods of quenching. One of them is lowering the DC voltage. the electric field. The kinetic energy of electrons can be greater A second method is adding a small amount of quenching gas e.g. than the ionisation energy of the gas and the secondary ionisa- ethyl alcohol in vapour form to the inert gas, e.g. argon. During tion process will occur. The gas gain (multiplication) process is in collisions between the counting gas ions and the quenching gas the form of a cascade which is called a Townsend avalanche. The molecules, the ionisation is transferred to the latter. The electron relative increase in the number of electrons (dn/n) per unit path energy is used up in dissociating the molecules instead of causing (dx) can be estimated using the Townsend equation: further ionisation. Even without using a quenching gas, the propa- gation of avalanches is eventually terminated by the build-up of a cloud of positive charge around the anode wire that consists of dn = a dx n where α is the first Townsend coefficient for the gas (α = 0 for the electric field below threshold value) (Figure A.83). In a uniform electric field (parallel plate geometry) α is con- stant and the number of electrons (n) per unit path increases expo- + nentially with x corresponding to the avalanche progress: V Current n(x) = n(0)eax where R α is the first Townsend coefficient for the gas n(0) is the number of electrons per unit path at x = 0 FIGURE A.84 Example of avalanches spreading along the anode in a n(x) is the number of electrons per unit path at x Geiger–Müller detector. Avalanche photodiode 75 A xial (transverse) plane the positive ions that were also formed during the avalanches. discussed in terms of the arithmetic mean, often simply called This positive space charge reduces the electric field in the critical the mean. A multiplication region below the strength required for additional Radioactive decay is governed by probabilities. For any given avalanches to form. radioactive atom, there is a constant chance or probability that it Related Articles: Gas-filled radiation detectors, Geiger–Müller will decay in a stated period of time. In a collection of nuclei of (GM) counters, Ionisation chamber, Proportional counter the same radionuclide, each nucleus will be completely indepen- Further Readings: Knoll, G. F. 2000. Radiation Detection dent of the other atoms as to when it decays. and Measurement, 3rd edn., John Wiley & Sons, Inc., New Each radionuclide has its own probability – for a radioio- York, Chichester, Weinheim, Brisbane, Toronto, Singapore, pp. dine-131 (131I) nucleus it is one chance in a million of decaying 201–202; Saha, G. P. 2001. Physics and Radiobiology of Nuclear per second (1 in 997,779), whilst the probability for a technetium- Medicine, 2nd edn., Springer–Verlag, New York, p. 68. 99m (99Tcm) nucleus it is one in thirty thousand (I in 31,155). The mean life, T–, of a nucleus before it decays is therefore directly Avalanche photodiode related to the probability – for an 131I nucleus it is just under a mil- (Nuclear Medicine) An avalanche photodiode is the solid-state lion seconds (997,778 s = 11.55 days) and for a 99Tcm nucleus it is equivalent to a photomultiplier. It

is a solid-state detector which is 31,155 s (= 8.65 h). able to multiply the charge of the deposited energy, hence increas- The mean life is related to the half-life by T– = 1.44T1/2. In the ing the signal output. Scintillation light enters the diode and inter- previous examples, the half-life of 131I is 11.57/1.44 = 8.01 days actions with the material create electron–hole pairs. The increase and for 99Tcm, 8.65/1.44 = 6.01 h. in gain is achieved by accelerating these primary electrons in an Related Article: Half-life electric field and allowing them to excite and create secondary electron–hole pairs (an avalanche of electrons) prior to read out. Average life time of atoms It is possible to achieve gains of up to several hundreds in read (Nuclear Medicine) The average life time of an atom is defined out signal. The gain is strongly dependent on the temperature and by taking the mean life time of a large population of radioactive applied voltage. The electron multiplication is a function of the nuclides. The radionuclide average life time is not equal to the interaction position hence introducing a statistical variance in the half-life since the fraction of radionuclides that ‘live’ longer than gain. This variance leads to a lower energy resolution relative to the half-life can live for several half-lives before decaying. The low noise silicon detectors. Avalanche photodiodes are preferably average lifetime τ is defined as the reciprocal of the decay con- used for detecting low energy photons. stant λ. τ is related to the half-life T½ according to Abbreviation: APD = Avalanche photodiode. Further Reading: Knoll, G. F. 2000. Radiation Detection 1 T t = = 1/2 and Measurement, 3rd edn., John Wiley & Sons, New York, pp. = 1.443 ×T l 1/2 ln2 291–292, 489–491. Related Article: Half-life of radionuclides Average absorbed dose (Nuclear Medicine) The average absorbed dose is the total energy Average mass energy absorption coefficient deposited (J) in an organ by ionising radiation divided by the -tar- (Radiation Protection) Mass energy absorption coefficients get organ mass, mt (kg). The average absorbed dose can be used depend on the energy of the radiation and on the atomic number to calculate an effective dose to patients using specific weight fac- of the medium. When the interactions occur between a radiation tors to determine a future risk of developing cancer. The absorbed beam which has a spectrum of energies, and/or a medium com- dose contribution to an organ stems from self-absorption from posed of a number of different elements, calculations may be sim- radiation originating from the target organ itself and radiation plified by the use of an average mass energy absorption coefficient originating in other organs. which takes into account an effective mean energy of the beam, Further Readings: Loevinger, R. F., T. Budinger and E. E. and the different atomic numbers of the elements in the medium. Watson. 1988. MIRD Primer for Absorbed Dose Calculations, Related Article: Mass energy absorption coefficient The Society of Nuclear Medicine Inc., New York; Snyder, W. S., M. R. Ford, G. G. Warner et al. 1975. S-value Absorbed Dose per Unit Cumulated Activity for Selected Radionuclides and Organs, Avogadro’s number Society of Nuclear Medicine, Reston, VA. (Nuclear Medicine) Avogadro’s number, NA, or constant is the number of atoms in 12 g of 12C. Since the number of atoms in 12 g of 12C is also a mole, Avogadro’s constant correlates the number Average dose of atoms (or molecules) in one mole of any substance. (Radiotherapy) The average dose or mean dose is determined by calculating the dose at a large number of discrete points uni- formly distributed within the specified volume (planning target NA = 6.02 *10-23 /mol volume PTV, critical structure) and calculating the mean of these dose values. Further Reading: Benson, H. 1996. University Physics, Related Articles: Maximum target absorbed dose, Minimum revised edition, Wiley-VCH, New York, p. 369. target absorbed dose, Mean target absorbed dose, Median target Related Article: Mole absorbed dose, Modal target absorbed dose, Hot spot Axial (transverse) plane Average life (General) To describe anatomical planes imagine a person stand- (Nuclear Medicine) The average life of a radionuclide (or a par- ing in an upright position and dividing this person with imaginary ticular excited electronic or nuclear level of an atom) is usually vertical and horizontal planes. Anatomical planes can be used to Axial resolution 76 Azimuthal describe a body part or an entire body. The axial or transverse Pulse length and axial resolution Interfaces A plane is a horizontal plane that divides the body into the upper 1 2 (superior) and lower (inferior) regions by running through the midsection of the body. Pulse length i Related Article: Anatomical body planes a b a b Axial resolution (Ultrasound) Axial resolution can be defined as the smallest sep- aration of a pair of point targets on the beam axis. i The separation between echoes from adjacent scatterers is a b shown diagrammatically in Figure A.85. a b If the pulse duration is tp then the interval between the end of the echo from the first scatterer and the beginning of that from the Transmitted pulse Reflected pulses second scatterer is given by δt is the interval between the start of the echoes from the two Incident pulse I scatterers − tp. Reflected pulse a-from 1st interface, b-from 2nd interface δt = 2d/c − tp. When tp = 2d/c then δt = 0 and the echoes are unseparated. The Reflected pulses are separated with short pulse (upper) distance d = c(tp /2) is half the pulse length. not separated with long pulse (lower) In practice the axial resolution is dependent on the pulse enve- lope, the signal/noise ratio (SNR) and signal processing within FIGURE A.86 A shorter pulse may resolve echoes from adjacent inter- faces that in a longer pulse merge together. the scanner. Generally, for conventional B-mode systems, the axial resolution is approximately half the pulse length. The implications for this can be seen in Figure A.86. A wavelength, results in longer echoes which may not be separated longer pulse, either from using an increased number of cycles, from adjacent tissue interfaces. e.g. in colour flow imaging, for a lower frequency, increasing Axial resolution may be measured using pairs of wires or filaments at progressively decreasing separation distances or by examining the image from a single wire or filament to measure Transmitted pulse the axial length from a reflector. Transducer Tissue Scatterers Related Articles: Bandwidth, Backing Azimuthal d (Ultrasound) The word is of Arabic origin and means ‘the ways’, that is the ways, or directions, a person faces. In navigation it is the tp horizontal angle in degrees, from true north to the present course, or direction (known as bearing). In the context of ultrasound fields, the term seems to have been used in somewhat differing ways. Received signal Usually the azimuthal plane refers to the image plane, and the azi- Time muth steering angle is the angle the ultrasound beam is steered off tp δt the normal to the transducer face (as for instance as in a phased array). This resembles the navigation analogy, as to what ‘bearing’ the ultrasound beam has, relative to the normal direction. FIGURE A.85 Axial resolution depends on the ability of the scanner to separate echoes from scatterers in the beam direction. This in turn Others refer to the azimuthal direction as the elevation direc- depends on the length of the first echo. If the echo from the second scat- tion, that is the direction perpendicular to the imaging plane. terer arrives before the end of the first echo, then there is no separation of Evidently the word should be used together with a definition of its echoes and the scatterers are not resolved. usage in the present context. B B B0 in frequency and phase are the basis of slice selection, frequency (Magnetic Resonance) The static magnetic field strength in an encoding and phase encoding, the three principal methods of MRI system is conventionally indicated with B0. It is usually spatial localisation in MRI which are described in detail in spe- expressed in units of Tesla (1 T = 104 G). The z-direction is com- cific articles. monly chosen along the direction of B0, although exceptions may In addition to their role in spatial encoding, magnetic field occur in open MRI systems. In current MRI systems, B0 has a gradients are also used to eliminate unwanted magnetisation constant value over time, ranging from 0.02 to 3 T. Currently, (‘spoiling’) in order to avoid generation of spurious echoes. In dif- experimental MRI systems have a field strength of up to 11 T fusion-weighted imaging, intense magnetic field gradient is used and in MRS, field strength is usually indicated by the proton fre- to introduce diffusion sensitisation into a pulse sequence. quency, for example 600 MHz for 14 T. Physically, gradient fields are generated by passing currents through gradient coils built into the scanner (Figure B.2). These B are resistive wire coils, with a separate coil designed to generate a 0 gradients (Magnetic Resonance) The potential of NMR as the basis of linearly varying field along each of the three Cartesian axes inside a medical imaging technique was realised by Kudravcev in the scanner bore. In the early days of MRI, gradient coil designs 1960 and notably by Damadian in 1971, who patented the idea. drawn from classical electromagnetism were used: Golay coils for However, no efficient means existed of mapping NMR signals the x and y gradients and a Maxwell pair for z. Today, gradients spatially, an obvious prerequisite for imaging. with superior performance in the context of modern scanners with In 1952, Carr had demonstrated use of a magnetic field vary- wide bores and shorter magnets are designed by computational ing linearly in space in the context of NMR spectroscopy. This methods. concept of a magnetic field gradient was to form the basis of spa- Gradients along any arbitrary oblique axis can be generated by tial encoding, allowing NMR to be transformed into a powerful passing currents through two or three of the gradient coils simul- imaging technique. The technique was developed in the 1970s by taneously, with amplitudes chosen to yield the required net gradi- Lauterbur and Mansfield, who in 2003 shared the Nobel Prize in ent direction and amplitude. Medicine for their work. The gradient set should ideally generate linear field variation The resonance frequency of nuclear spins in a static magnetic along the appropriate Cartesian axis, without any field compo- field of amplitude B0 is given by the expression ω0 = γB0. Imagine nents in orthogonal directions. However, electromagnetic theory now that the amplitude of the field is made to vary linearly along, shows that this is not possible, and there are inevitably orthogo- for example the x-axis. It follows that the resonance frequency nal fields known as Maxwell terms. These sources of image arte- will also vary linearly with x, according to the following equation: facts are minimised by careful design, but cannot be completely eliminated. Gradient performance is specified in terms of the maximum w(x) = g(B0 + Gxx) achievable amplitude (the ‘gradient strength’) and the speed with which the gradient can be switched on and off (expressed as ‘slew where Gx is a measure of the steepness of the field variation in rate’ in mT/m/s). These are limited for safety reasons, as electrical x-direction (the ‘gradient strength’), usually expressed in units of currents induced in the body can lead to peripheral nerve stimula- mT/m. tion, an unpleasant and, in extreme forms, painful phenomenon. A magnetic field is a vector quantity, and it is important to Another important parameter is the linearity of the gradient over note that the direction of the field remains that of the main static the desired imaging volume, which ensures correct geometrical field, B0, usually along the z-axis (the cranio-caudal direction representation of the imaged object. Linearity is more difficult to with respect to the patient). Figure B.1 illustrates magnetic field achieve in more compact magnets and particularly in open scan- vectors along the x-axis without (a) and with (b) imposition of a ners where flatbed gradient sets are required. gradient. In the latter case, the field remains oriented along the Related Articles: Diffusion imaging, Frequency encoding, z-axis but

varies linearly in amplitude along x. The term gradient Phase encoding, Slice selection, Spoiling, Spurious echoes direction indicates the direction of field variation (the x-axis in this example). B0 homogeneity Magnetic field gradients are applied for short intervals (Magnetic Resonance) B0 refers to the main magnetic field of the of time, on the order of milliseconds, during an MRI pulse MR system, typically varying from 0.1 to 3.0 T for clinical MR sequence. If a gradient is applied while there is magnetisation systems. The homogeneity of the static magnetic field is an impor- in the transverse plane, the precessional frequency of this mag- tant measure of the quality of the magnet. Inhomogeneities can netisation varies linearly with position along the gradient direc- be produced by the scanner and the magnetic susceptibility of the tion. When the gradient is switched off, magnetisation returns to object being imaged. a common precessional frequency determined by B0. However, The homogeneity of the main magnetic field is measured in the period of differential precession results in linear variation parts per million (ppm) within a given spherical volume in the in phase along the gradient direction, and this variation persists centre of the magnet, the size of this volume given as the diam- until removed or modified by another gradient. These variations eter of the spherical volume (DSV). Homogeneity requirements 77 B0 inhomogeneity 78 Back pointer B1 homogeneity (Magnetic Resonance) B1 refers to the field produced by the MR RF system. RF system homogeneity can vary greatly as it mainly depends upon the RF coil used. There are two main factors which affect B1 homogeneity, the first is the interaction between the RF B field and the object being imaged and the second is the inherent x x inhomogeneities of the RF coils used in the system. Volume coils such as the birdcage and saddle coils, produce FIGURE B.1 Effect of a magnetic field gradient of static field amplitude. relatively good B1 homogeneity at their centre. The saddle coil will produce very good B1 homogeneity in the direction of its long axis and the bird cage coil produces the highest B1 homogeneity over most of the coil volume, giving excellent image uniformity. Current ports Surface coils are renowned for their inhomogeneity, producing a Outer loops high SNR at the surface of the object, which rapidly decreases with depth in the object. Flip angles in conventional RF pulses are sensitive to B1 inho- mogeneities; therefore, to remove the effects of B1 inhomogeneity, adiabatic RF pulses can be used. They are amplitude- and fre- quency-modulated pulses that are mostly insensitive to the effects of B1 inhomogeneity. The disadvantage of these pulses is that they (a) Inner loops (b) can require longer scanning times. Abbreviations: RF = Radio frequency and SNR = Signal-to- FIGURE B.2 Gradient coil designs for the z-axis (a) and for x and y-axis (b). noise ratio. Related Articles: RF uniformity, B1 inhomogeneity for imaging are generally lower than for MR spectroscopy, but B1 inhomogeneity for most imaging techniques, homogeneity must be maintained (Magnetic Resonance) B1 refers to the RF system used in MRI. over a large region for good quality imaging. The field should be B1 inhomogeneity is one of the most important causes of image homogeneous in the range of a few ppm over a 30–50 cm DSV. non-uniformity. There are two causes of B1 inhomogeneity: firstly, Abbreviations: DSV = Diameter of a spherical volume, ppm the interaction between the RF field and the object being imaged = Parts per million, SNR = Signal-to-noise ratio and T = Tesla. and secondly, inherent inhomogeneities of the RF coils used in Related Article: B0 inhomogeneity the system. More detail is provided in the article B1 homogeneity. B0 inhomogeneity Abbreviations: NMR = Nuclear magnetic resonance and RF (Magnetic Resonance) B0 inhomogeneity is the degree of inho- = Radio frequency. mogeneity of the main, static magnetic field (B0). Manufacturers Related Articles: RF uniformity, B1 homogeneity try to make the magnetic field as homogeneous as possible, in particular at the isocentre of the magnet, but it is not possible to b-factor have a perfectly homogeneous magnet. Homogeneity of the main (Magnetic Resonance Imaging) See b-value. magnetic field is measured in parts per million (ppm). Sometimes, b-values are referred to as b-factors. The b-factor Abbreviations: ppm = Parts per million and SNR = Signal- terminology was frequently used in the first papers about diffu- to-noise ratio. sion-weighted imaging (DWI) and diffusion tensor imaging (DTI). Related Article: B0 homogeneity B-lines B1 (Ultrasound) A B-mode (brightness of greyscale) image is pro- (Magnetic Resonance) In an MR system, the radio frequency duced by the data from a large number of B-lines, where each field strength is conventionally indicated by the symbol B1. To B-line is echo amplitude data from one pulse-echo sequence. affect a nutation, it is applied at Larmor frequency, usually in a The width of each line will determine the lateral resolution of the plane transverse to B0. If the strength of the B0 field increases, image. The number of lines used to produce an ultrasound image also the frequency of the B1 field increases linearly. For example, affects the frame rate, Figure B.3. the frequency of the B1 field for a 4 T system is 171 MHz for Related Articles: B-mode, B-scanner, Frame rate proton imaging. At such high frequency, the interaction between the B1 field and the human body can no longer be neglected. This interaction is caused by the dielectric resonance since the effec- B-scan tive wavelength of the B1 field is now comparable with the dimen- (Ultrasound) A B-mode image is produced by the data from a sion of the human body or some body structure. This interaction large number of B-lines. The word scanning is used for the method degrades the B1 field homogeneity and subsequently the quality to send and receive pulses sequentially along parallel beam lines of images. This interaction can also produce an increase of tem- starting at one end of the array transducer, Figure B.4. perature in some part of the body as brain or eyes because the Related Article: B-mode electric field associated with the B1 field increases with the B1 inhomogeneity and consequently increases the specific absorp- Back pointer tion rate. (Radiotherapy) The back pointer of a radiotherapy treatment Related Article: Specific absorption rate machine, such as linear accelerator, is a device used to assist beam B(x) B(x) Background 79 Background signal Frame rate and scan parameters-I Background equivalent radiation time (BERT) (Radiation Protection) The background radiation equivalent time e frame rate is dependent on: (BERT) is a unit of measurement of radiation dose. Sometimes depth of scan, it is important to give an idea about the magnitude of the doses width of scan, related, for example to diagnostic radiological investigations, in line density comparison to the value of background irradiation to which popu- B lation is exposed. BERT values are, of course, only a comparative indication, n-number since there is no evaluation of the specific methods and param- d of scan lines eters used for the x-ray investigation, as well as of the local–indi- vidual background values. I-line density Typically, BERT values given vary from 2 BERT for a dental x-ray to 400 BERT for a barium enema investigation. Further Reading: UNSCEAR. 2000. Sources and effects of w ionizing radiation. United Nations Scientific Committee on the Effects of Atomic Radiation Report, New York. FIGURE B.3 Frame rate and scan parameters. (Courtesy of EMIT proj- Background radiation ect, www .emerald2 .eu) (Radiation Protection) There are two kinds of ionising radia- tion sources: natural and man-made. The natural sources pro- duce the background radiation. The natural background consists of the cosmic radiation, the terrestrial radiation and the internal Beam profile, focus and output display radiation. Cosmic Radiation: All living things on the Earth are con- stantly bombarded by radiation from the sky. Charged particles from the sun and the stars interact with the atmosphere and the magnetic field and produce a shower of radiation, typically beta and gamma radiation. Elevation and magnetic field clearly influ- ence the cosmic radiation; therefore, there are different values in different parts of the world. It is evaluated that cosmic radiation contributes 8% of the total radiation, including the man-made As beam moves along array, echoes detected depend on beam component. width-affects spatial and contrast resolution Terrestrial Radiation: Radioactive substances are in nature, namely, soil, water and vegetation. The major contribution to terrestrial radiation is given by uranium and its decay products: thorium, radium and radon. Some of these materials might be ingested with food while radon is inhaled. Locations with higher concentration of uranium and thorium in the soil will have higher terrestrial radiation; therefore, different parts of the world have FIGURE B.4 The principle of scanning and how beam width affects different values. It is evaluated that terrestrial radiation contrib- spatial and contrast resolution. (Courtesy of EMIT project, www . utes 8% in addition to the radon that contributes 55% of the total emerald2 .eu) radiation, including the man-made component. Internal Radiation: All people have radioactive potassium-40, carbon-14, lead-210 and other isotopes inside their body from birth. set-up. The front pointer indicates the centre of the radiation beam Again, the individual quantities can vary but the variation is minor at the point of entry into the patient and the back pointer indicates compared to cosmic and terrestrial radiation. It is evaluated that centre of the radiation beam at the point of exit from the patient. internal radiation contributes 11% of the total radiation including The back pointer can be in the form of a mechanical pointer or an the man-made component. optical pointer. The background radiation accounts for about 81% of all pub- lic exposure, while the man-made part accounts for about 19%. Background Given the earlier-described differences, the average, individual, (Nuclear Medicine) In nuclear medicine imaging, background total radiation dose is evaluated to be circa 3.6 mSv/year. Natural refers to the signal generated by radioactive sources other than and artificial radiation doses produce the same kind of effects. the source of interest. Background signal is also referred to as Further Readings: NUREG-series publications; UNSCEAR. (detector) noise. For example, signals originating from detector 2000. Sources and effects of ionizing radiation. United Nations electrical noise and background radiation are considered to be Scientific Committee on the Effects of Atomic Radiation Report, ‘background’. New York. www .n rc .go v /rea ding- rm /do c -col lecti ons /n uregs , Background signal decreases the spatial resolution and image accessed 31 July 2012. contrast of the imaging system and introduces a source of error Hyperlinks: NRC: www .nrc .gov; IAEA: www .iaea .org in quantitative measurements. For special noise-sensitive applica- tions such as dynamic studies, the background contribution can be subtracted from the accumulated image. Background signal Related Article: Signal-to-noise ratio (SNR) (Nuclear Medicine) See Background Backing material 80 Back-projection reconstruction Backing material Before describing how these reconstructions can be done, (Ultrasound) PZT is used as transducer material because it can be a remark on the test images: The classical test object for testing moulded in different shapes and can efficiently convert electrical projection reconstructions is the Shepp–Logan phantom. This is energy to mechanical energy and vice versa. However, the mis- a synthetic image made of ellipses. The projection transform and match in acoustic impedance between PZT (30 × 106 kg/m2/s) and Fourier transform of ellipses can be computed analytically; thus, B soft tissue (1.6 × 106 kg/m2/s) is a significant disadvantage for the the entire transform of the Shepp–Logan phantom can be computed transmission of ultrasound energy into tissue and for the ability of to arbitrary precision. transducers to produce short acoustic pulses necessary for high- Back-Projection: The plain back-projection does not give resolution imaging. The mismatch will cause internal reverbera- good reconstructions but helps understand what happens for the tions within the transducer element. more correct ones. The idea is very simple: for each projection A damping backing material, mounted on the backside of angle, we smear back the values on the image domain. The full the transducer element, with an acoustic impedance adjacent to back-projection is the average of these images. the acoustic impedance of PZT, will reduce the Q-value (ring- Consider the following 4 × 3 image s, with an

object in the ing) effectively. Unfortunately, this will also reduce the sensitiv- middle Figure B.6. We also show the projections: ity (Figure B.5). The most efficient way of reducing the Q-value without reducing the sensitivity is to use a matching layer at the pq = P(s,q = 0,90) front. A common compromise is to use a backing material with acoustic impedance slightly lower than that for PZT. Fourier Transforms of Projection, the Fourier Slice Theorem The backing material should ideally damp out, or absorb, all (FST): The Radon transform, i.e. the projection operation, is a the ultrasound energy so that none of the ultrasound energy will transform from image space to a space of lines in the sense that be reflected back into the transducer element. we get one value per line. A line is specified by a direction θ and In continuous wave applications, as CW-doppler, high Q-value a distance r from origin. It can be related to the Fourier trans- transducers are required and therefore air backing is commonly form, the ‘Swiss army knife’ of image processing. It is possible to used. Air will guarantee total reflection on the backside of the transducer disc, which will minimise the ultrasound energy loss. Abbreviation: PZT = Lead zirconate titanate. Related Articles: Transducer, Matching layer, PZT 0 0 0 0 Back-projection (Diagnostic Radiology) See Back-projection reconstruction 0 1 0 1 (s, 0) = p0 e projected values are smeared back to generate Back-projection reconstruction 1 bp0 = (p0); 0 1 0 bp90 = (p90) (Diagnostic Radiology) Many medical imaging modalities pro- duce projection data: CT is the archetypal example, and PET is also a source of projection data. The first MR image from P. 0 0 0 0 Lauterbur was also a projection (MR radial sampling). The great insight for the revolution of medical imaging in the 1970s was (s, 90) = p90 the realisation that it is possible to reconstruct images from such projections. 0 2 0 bp0 = (p0) bp90 = (p90) Transducer elements Pulse excitation 0 2 0 0 0 0 Displacement 0 2 0 1 1 1 0 2 0 1 1 1 Long pulse length Undamped transducer element 0 2 0 0 0 0 Poor axial resolution Pulse excitation Displacement 0 1 0 Backing 0.5 1.5 0.5 0.5 1.5 0.5 Damped transducer element Short pulse length Good axial resolution 0 1 0 FIGURE B.5 The role of backing material to produce short ultrasound FIGURE B.6 Example of plain back-projection. , projection operator; pulses. (Courtesy of EMIT project, www .emerald2 .eu) , back-projection operator. Backscatter 81 Backscatter factor Iterative (See Iterative algorithms): The Radon transform, and related projection transforms, are at least approximately lin- Slice: 2D 2D fourier 1D ear operations. This implies that rather than using specific dis- cretised version of the continuous inverse formula, we can try to use techniques from the inversion of linear operators. The mod- ern toolbox of linear inversion contains a wide range of iterative B techniques. 1D fourier A very simple iterative idea for the solution of linear systems DFT1 Slice 2D is the following: every row of a linear system of equation specifies 1D: PS Projection P a linear space (or hyperspace in higher dimensions). We can pick any starting point, find the nearest point on the first space, then the nearest of this on the second space, and iterate. This algorithm (Kaczmarcz) is well suited for reconstruction of projection. For projection data, the ‘nearest’ involves a back-projection, and the FIGURE B.7 Example of FST – see explanation in the text. corresponding algorithm belongs to the class of algebraic recon- struction technique (ART). Related Article: Iterative algorithms perform a Fourier transform on the distance variable. The Fourier Further Readings: Kak, A. C. and M. Slaney. 2001. Principles slice theorem, or projection slice theorem, relates the 1D Fourier of Computerized Tomographic Imaging, Classics in applied math- transform of a projection to slices of the 2D Fourier transform of ematics 33, Society for Industrial and Applied Mathematics, the image. Philadelphia, PA; Sureshbabu, W. and O. Mawlawi. 2005. PET/ The FST claims that in the following figure (Figure B.7), CT imaging artifacts. J. Nucl. Med. Technol. 33:156–161. we get the same result by following the upper path (2D Fourier Hyperlink: http://www .mat hs4m edic alimaging .co .uk followed by slice extraction) or the lower path (projection, slice extraction, then 1D Fourier). Backscatter Filtered Back-Projection: The filtered back-projection (Radiotherapy) At kilovoltage energies, the dose at the surface algorithm addresses the drawback of back-projection: the back- of an irradiated phantom (or patient) is greater than the dose at projection blurs the image Figure B.8. This is because the back- the same point if no phantom is present. This is particularly true projection is an intuitive operator, but is ‘not’ the mathematical at energies below 400 kV and arises from the fact that phantom inverse of the projection. A more careful investigation of the materials scatter a fraction of the primary radiation back to the transform using the Fourier approach shows that the correct surface. This contribution to the dose at the point in the phantom inverse involves a filtering, then a back-projection. Indeed, as is known as backscatter. the FST tells us what the Fourier transform of projections are, Related Article: Backscatter factor we can use inverse Fourier transformation to recover the image Further Readings: Podgorsak, E.B. 2005. Radiation Oncology (Figure B.7 shows this clearly, to invert the projection, the arrow Physics, IAEA, Vienna, Austria, available at: http: / /www -nawe b P, we can follow the dotted arrows: S1->DFT1->IS2->IDFT2). .iae a .org /nahu /dmrp /syll abus. shtm; Walter, M. 2003. Textbook However, remember that the variable is a ‘distance’ as in a polar/ of Radiotherapy Radiation Physics, Therapy and Oncology, 6th spherical coordinate system and the formula for integration in edn., Churchill Livingstone, Edinburgh, New York. polar coordinates involves an integrating factor proportional to that distance, which acts as a filter. Theoretically exact inver- sion would thus require computing for all values of this distance, Backscatter but finiteness and discretisation requirements result in different (Ultrasound) Ultrasound is reflected off large surfaces and choices of filtering. tissue interfaces. Within tissue, ultrasound is scattered from Note that the projection is called Radon transform in math- small-scale variations in acoustic impedance. In this context, ematics, and we thus want to compute the inverse Radon trans- small means smaller than the wavelength. In this case, the form. Equivalent formulas for projections from lines in space sound is not re-directed in a specific direction but rather in are even more complex. Recently, Katsevitch developed 3D many directions. The energy scattered in various directions is formulas for helical reconstruction, which generalise the Radon given by the expression for the scattering cross section, which is formulas. a function of angle, the angle relative to the propagation direc- tion of the incident. The sound waves reflected at an angle of 180° are known as backscatter. The backscattered sound is what is detected in diagnostic ultrasound and contributes to the ultra- 1 sound image. Backscatter factor (Radiotherapy) The backscatter factor is defined as the ratio of 2 an appropriate radiation quantity (e.g. dose or exposure) at the reference point, i.e. on the surface of the patient or phantom on the central axis of the beam, to the equivalent quantity at the same position in the absence of the patient or phantom. This fac- tor increases with the area of the field irradiated and the thickness FIGURE B.8 Filtered backprojection example – Line 1 shows projec- of the underlying tissues. In general, the increase is more marked tion data for one angle (Shepp – Logan, angle θ = 0) and Line 2 shows the for smaller field sizes and tails off with increasing field size. filtered data, which would then be backprojected. Related Article: Backscatter Bad pixel 82 Band limiting Further Readings: Podgorsak, E.B. 2005. Radiation Ball bearing Oncology Physics, IAEA, Vienna, Austria, p. 175, available at: (Diagnostic Radiology) See Bearing http: / /www -nawe b .iae a .org /nahu /dmrp /syll abus. shtm; Walter, M. 2003. Textbook of Radiotherapy Radiation Physics, Therapy and Band gap energy Oncology, 6th edn., Churchill Livingstone, Edinburgh, New York; (General) The band gap energy (or energy gap) is the energy B Williams, J. R. and D. Thwaites. 2000. Radiotherapy Physics in range between the top of the valence band and the bottom of the Practice, Oxford Medical Publications, Oxford, UK. conduction band. No electron states exist in this energy range. The width of the band gap energy determines the type of material Bad pixel – conductor, semiconductor, insulator. (Diagnostic Radiology) Bad pixel is a term related to flat LCD For semiconductors, the band gap is a band with forbidden monitors and large-area solid-state detectors in digital radiogra- energy states, in which the electrons are forbidden to propagate. phy. Bad pixels (or dead pixels) result from problems in the micro- In order for an electron to get excited from the valance band to electronic thin film technology (TFT) used in these monitors and the conduction band, the electron must receive enough energy to detectors. Such pixels present either a black (or white) pixel on the transcend the band gap. The width of the band gap determines the monitor screen or the lack of detection (no response to input pho- conductivity of the semiconductor (Figure B.9). tons) from the detector. Each manufactured monitor or detector Hyperlink: Wikipedia, http://en .wikipedia .org /wiki /Band is generally tested to determine bad pixels. A discrete number of _gap_ energy bad pixels can be corrected by software referring to a correction map and replacing them by values interpolated with surrounding Band limiting pixels. Depending on the type of data readout, it is possible for a (General) All electrical signals can be described in terms of their bad pixel to result in loss of information in a whole detector row frequency content, and band limiting is the restricting or filter- or column. ing of these signals in some way to a limited band of frequencies (Figure B.10). Baird-atomic system multi-crystal camera The main purpose of band limiting is to preserve the desired (Nuclear Medicine) The Baird-atomic system multi-crystal camera information whilst minimising unwanted information (noise) was an early scintillation camera developed in the 1970s to avoid where the two can be separated due to their different frequency image artefacts at high count rates and to get a better spatial reso- content. lution. It consisted of several small NaI(Tl) columnar crystals that A second common use is to prevent unwanted signals from were glued together to form a large detector block. The camera was passing further through the chain of processes where their pres- then operated as an ordinary scintillation camera with Anger posi- ence could inadvertently cause problems or indeed be processed tioning logic. This system was primarily designed for high count in such a way that they transmute into frequencies within the rate scenarios and for clinical applications in cardiac studies of bandwidth of the signal of interest. high activity bolus passage. A very high photon fluence rate would impinge on the crystal, causing pile-up effects both in the crystal from the scintillation light and also in the electronics. The former could then be reduced by altering the crystal geometry. E Related Articles: Scintillation camera, High count rates, High photon fluence rates, Cardiac studies Conduction band Balanced FFE Band gap (Magnetic Resonance) See Fast imaging with steady state preces- Valence band sion (FISP) Balanced gradients (Magnetic Resonance) A magnetic-field gradient with amplitude G in direction i, varying as a function of time t, is balanced at the point T in time if the zeroth moment of the gradient with respect FIGURE B.9 Band gap. to time is 0, i.e. if the following condition is fulfilled: T Lowpass filter òGi (t)dt = 0 Original signal 0 The condition for refocusing of phase dispersion induced by field gradients is fulfilled if this condition is fulfilled. Note: After a 180° radio frequency (RF) pulse, a sign change of the gradient must be applied due to the inversion of phase induced by the RF pulse. Related Article: Gradient motion rephasing (GMR) Band limited signal Further Reading: Haacke, E. M., R. W. Brown, M. R. Thompson and R. Venkatesan. 1999. Magnetic Resonance Log frequency Imaging: Physical Principles

and Sequence Design, John Wiley & Sons, New York. FIGURE B.10 Illustration of band limiting. Log amplitude Bandwidth 83 Bar phantom Typical examples include lowpass, highpass, bandpass and bandstop filters, which are usually specified by the frequen- 1 cies where the amplitude drops to 1/√2 or 3 dB of the passband amplitude. One specific form of band limiting (lowpass) is required when 0.8 analogue signals are to be digitised by an analogue to digital con- B verter (ADC). Commonly referred to as an ‘anti-aliasing filter’, this filter prevents any signal greater than half the ADC sampling 0.6 frequency from entering the conversion process, which could oth- erwise result in the higher frequencies aliasing and being trans- muted into signals inseparable from the original data in the lower 0.4 frequencies. 0.2 Bandwidth (Ultrasound) Bandwidth in ultrasound applications is usually a measure of spectral width of a pulse. It is most commonly defined 0 as the frequency band over which the amplitudes in the ampli- 0 1 2 3 4 5 6 7 8 9 10 tude spectrum is over 70% (−3 dB bandwidth) of the maximum Frequency (MHz) amplitude (or 50%, −6 dB bandwidth). For a power spectrum, the power level is 50% (−3 dB) and 75% (−6 dB), because of the FIGURE B.12 The bandwidth of the pulse in Figure B.10, defined as −3 decibel definition. The definition is the same for the frequency dB bandwidth (approximately 0.8 MHz) and as −6 dB bandwidth (approx- characteristics of an amplifier, for example although in this case imately 1.2 MHz). the measurement is not usually taken on a pulse (Figures B.11 and B.12). Related Articles: Axial resolution, Absorptive backing Resolution test pattern Bar pattern (Diagnostic Radiology) See Bar phantom Bar phantom (Diagnostic Radiology) Test devices consisting of parallel lines of an x-ray absorbing material separated by spaces of equal thick- ness to form line pairs (one line and the adjacent space). The width of a line and space is expressed in terms of line pairs per 1 2 3 4 5 6 7 8 unit of length (lp/mm or lp/cm). This is in the spatial frequency Spatial frequency (LP/MM) domain. The test pattern is designed with sections containing dif- ferent spatial frequencies as illustrated in Figure B.13. FIGURE B.13 Typical bar phantom pattern with the bar spatial fre- Bar phantoms are used to evaluate imaging system blurring quency. (Courtesy of Sprawls Foundation, www .sprawls .org) by determining the maximum spatial frequency at which the lines can still be ‘resolved’. The maximum frequency that can be resolved has a reciprocal relationship to the blur dimension. Bar phantom (Magnetic Resonance) Bar phantoms are used in MRI quality control to qualitatively assess the spatial resolution of the imag- 0.015 ing system. Several groups of thin parallel plates are spaced in a test object. Each group of plates are placed a set distance apart 0.01 from each other, which varies between each group. The distances used for the Eurospin test object are 2, 1.5, 1, 0.5 and 0.3 mm (Figure B.14). 0.005 To evaluate the spatial resolution, it is necessary to determine the smallest distance between plates that can be resolved. A con- 0 ventional MR system should have a spatial resolution of 1 mm, but this can vary greatly with the pulse sequence and coil used as –0.005 well as a number of other factors. Related Article: Eurospin test object –0.01 Bar phantom –0.015 (Nuclear Medicine) A bar phantom is a common name for physi- 0 50 100 150 200 250 300 350 400 cal and software phantoms used to evaluate the spatial resolution Time (μs) of an imaging system. A typical bar phantom is made of lead, and has four quadrants, each with a fixed (and varying) bar width FIGURE B.11 A typical ultrasound pulse. and bar pitch. The ability of the imaging system to separate and Amplitude Relative amplitude Barium 84 Barn earth metal whose name derives from the Greek word ‘bary’, meaning ‘heavy’. Barium metal reacts readily with oxygen in the air and with water. It is commonly found in nature as barite (bar- ium sulphate) or witherite (barium carbonate). Medical Applications: X-ray contrast agent – Barium has B become the most widely used contrast agent for studies of the digestive tract due to its high Z (=56) and low toxicity. Typically, barium is administered to the patient orally as a ‘barium meal’, or through their rectum as a barium enema. Internally, the bar- ium gradually coats the colon, improving image contrast due to its x-ray absorbing properties (barium is ‘opaque’ to diagnostic x-rays). Barium Fluoride Detectors: Barium fluoride is an inorganic crystal with a very fast decay time (0.8 ns), sometimes used in time-of-flight PET detectors. Unfortunately, its photon yield is relatively small. Related Articles: Contrast agent, Contrast enhancement, Contrast media, Detector PET, Time-of-flight techniques in PET Barium fluoride (BaF) (Nuclear Medicine) Barium fluoride (BaF) is an inactivated (not FIGURE B.14 Example of a Eurospin bar phantom. doped) inorganic scintillator. The scintillation decay can be described by two components; a slow (630 ns) and a fast com- ponent (0.6 ns). Together with the high atomic number, BaF is suitable for detector applications that require high detection effi- ciency per unit volume and a fast detector response. Further Reading: Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, New York, pp. 241–242. Barn (General) The barn, b, is a special unit of cross section: FIGURE B.15 A typical four quadrant bar phantom (left), and the cor- 1 barn = -28 responding gamma camera image (right). 10 m2 = 100 fm2 Cross Section: The concept of cross section, symbol, Φ, is isolate the bars gives a measure of the system’s performance in used in discussing the strength of the interaction of particles, regards of spatial resolution. An example of a bar phantom, and charged or uncharged (including photons) with nuclei or other the corresponding gamma camera image is shown in Figure B.15. ‘target entity’. If a beam is interacting with a target, then the cross section is the average area perpendicular to the direc- Barium tion of the radiation that has to be assigned to each nucleus in (General) order to account geometrically for the total number of interac- tions. The unit of cross section is therefore in terms of area. It can be thought of as the ‘area of influence or interaction’ and is Symbol Ba completely independent from its physical dimensions. The cross Element category Alkaline earth metal section depends not only on the type of the target but on the Mass number A of stable 130 (0.106%), 132 (0.101%), 134 type and energy of the particles. Cross section is not limited to isotopes (2.417%), 135 (6.592%), 136 beams – it can be expressed in more general terms and the term (7.854%), 137 (11.232%), and 138 is defined by the International Commission on Radiation Units (71.698%) and Measurements as Atomic number Z 56 Atomic weight 137.33 The cross section, σ, of a target entity, for particular inter- Electronic configuration 1s2 2s2 2p6 3s2 3p6 3d10 4s2 4p6 5s2 action produced by incident charged or uncharged par- 4d10 5p6 6s2 ticles, is the quotient of P by Φ, where P is the probability Melting point 1000 K of that interaction for a single target entity when subjected Boiling point 2170 K to the particle fluence, Φ, thus Density near room temperature 3510 kg/m3 P s = F History: Barium was first identified by Carl Scheele in 1774 and extracted by Sir Humphry Davy in 1808. It is a soft alkaline Unit: m2 Barrel distortion 85 Barten model Fluence: The fluence, Φ, is the quotient of dN by da, chemical substances and bacteria into the brain whilst allowing where dN is the number of particles incident on a sphere of passage for oxygen and vital nutrients. cross-sectional area da, thus Related Article: Compartment models dN Barten contrast sensitivity model F = da (Diagnostic Radiology) See Barten model B Unit: m−2 Barten model (Diagnostic Radiology) In medical imaging, it is important to Related Article: Cross section have a visual consistency in how a given digital image appears on Further Reading: ICRU. 1998. Fundamental quantities different monitors (CRT or LCD), taking into account also con- and units for ionizing radiation. ICRU Report 60, International trast sensitivity properties of human visual perception. National Commission on Radiation Units and Measurements, Bethesda, Electrical Manufacturers Association (NEMA) has prepared and MD. kept under constant review a standard DICOM 3.14 (NEMA, 2011) to provide an objective, quantitative mechanism for map- Barrel distortion ping digital image values (pixel values or grey levels) into a given (Nuclear Medicine) A barrel image distortion is an effect of range of luminance of a medical imaging display (or of a light camera non-linearity. Non-linearity effects occur when the sig- box). The relationship that the DICOM standard defines between nal in the photomultiplier tubes (PM tubes) does not change lin- pixel values and displayed luminance is based upon contrast sen- early with the displacement of a source across the detector face. sitivity measurements and Barten model (Barten, 1999) of human Consider a point source moving from the edge to the centre of the visual perception. By this model, human contrast sensitivity is PM tube; if the increase in PM tube counting efficiency is lower distinctly non-linear within a wide range of luminance. Barten’s than the increase expected from a linear correlation, the x and model considers neural noise, lateral inhibition, photon noise, y-position signals will also change in a non-linear way. This leads external noise, limited integration capability, the optical modula- to an image distortion known as barrel distortion. The typical bar- tion transfer function, orientation and temporal filtering. rel distortion of a straight-line source is seen in Figure B.16. The The human eye is relatively less sensitive in the dark areas of opposite distortion is the pincushion distortion and it is discussed an image than it is in the bright areas of an image. This variation in a separate article. Barrel distortion can be found at other imag- in sensitivity makes it much easier to see small relative changes ing modalities – i.e. at the monitor of a fluoroscopic x-ray system in luminance in the bright areas of the image than in the dark (image intensifier). areas of the image. Based on Barten model, a unit called just Related Articles: Photomultiplier (PM) tubes, Pincushion noticeable difference (JND) was defined. JND is the luminance distortion difference of a given target under given viewing conditions that Further Reading: Cherry, S. R., J. A. Sorenson and M. E. the average human observer can just perceive. The grayscale Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, standard display function (GSDF) is the mathematically defined Philadelphia, PA, pp. 234–235. mapping of an input JND index to luminance values. GSDF of DICOM 3.14 covers the luminance range from 0.05 to 4000 cd/ Barrier m2 – this range corresponds to 1023 JNDs. The minimum lumi- (Nuclear Medicine) In nuclear medicine, a barrier can refer to nance corresponds to the lowest practically useful luminance of both a physical tissue boundary defining a compartment and/or cathode-ray-tube (CRT) monitors and the maximum exceeds the a physiological boundary preventing distribution of radiotrac- highest luminance of very bright light boxes used in mammog- ers. An example of a physical and physiological boundary is the raphy (Figure B.17). blood–brain barrier (BBB). This restricts the passage of various 10,000 1,000 100 10 1 0.1 0.01 1 201 401 601 801 1001 JND FIGURE B.16 Distortion of four line sources caused by nonlinearity in the PM tubes response. This particular distortion is referred to as the FIGURE B.17 The DICOM 3.14 grayscale standard display function barrel distortion. (GSDF). Luminance (cd/m2) Basal cell carcinoma 86 Baseline correction Monitors that are calibrated to the DICOM 3.14 standard should always convert the same digital input values into the same luminance values. AAPM has provided the standard guidelines for the performance evaluation of electronic displays (AAPM, 2005). The AAPM document describes how to make quality B assessments on medical display systems and also gives some acceptance criteria for the display systems. Related Articles: Image display, Digital display, Brightness, Image perception, DICOM, Human visual response function, Greyscale standard display function, Medical image display Further Readings: AAPM. 2005.

Assessment of display per- formance for medical imaging systems, AAPM On-line Report No. 03 by Task Group 18, American Association of Physicists in Medicine, College Park, MD; Barten, P. G. J. 1999. Contrast Sensitivity of the Human Eye and Its Effects on Image Quality, SPIE Press, Bellingham, WA; NEMA. 2011. Digital Imaging and Communications in Medicine (DICOM) - Part 14: Grayscale standard display function. PS 3.14–2011. National Electrical Manufacturers Association, Rosslyn, VA. Hyperlinks: NEMA DICOM; AAPM Basal cell carcinoma FIGURE B.18 The sonogram range is from approximately ±58 cm/s. (Non-Ionising Radiation) Basal cell carcinoma is the most com- The peak systolic flows show aliasing with the peaks misrepresented. mon type of skin cancer (80%). It originates from cells in the basal layer of the skin, and therefore, is mainly linked to longer UV wavelength that penetrates at those depths, in particular, to expo- sure to the sun and sunbeds (which are primarily UVA sources). Related Articles: AORD, Melanoma, UV light hazard, UV dosimetry Further Reading: Blumenberg, Miroslav. 2018. Human Skin Cancers: Pathways, Mechanisms, Targets and Treatments, London. Base layer (Diagnostic Radiology) The base of a typical radiographic film, also known as film base (see the eponymous article). Baseline (Ultrasound) In this context, the baseline refers to the horizon- tal line in a sonogram that defines zero velocity or frequency. This line can be shifted in most pulsed Doppler systems, thereby changing the useful velocity/frequency range in each direction. This may aid in reducing the effect of aliasing. If, for instance, the velocity range normally is displayed as −x to x cm/s (x dependent on the pulse repetition frequency), the baseline can be shifted to show velocities between 0 and 2x cm/s (Figures B.18 and B.19). This can be done when the direction of flow under investigation FIGURE B.19 The baseline has been changed, extending the velocity is known and the sonogram is uncontaminated by spectra from range in the direction of flow to 100 cm/s. This allows the sonogram to be displayed without aliasing and permits accurate measurement of peak other vessels. velocities. Baseline correction (Magnetic Resonance) In vivo NMR spectra frequently contain domain signal by a linear ramp that rises from zero at the begin- broad baseline features, due either to instrumental imperfections ning of signal acquisition to unity after a few data points. This is or to the presence of less mobile nuclei. This latter type of base- a fairly crude technique that can result in undesirable line shape line is most significant in 31P NMR, where it originates from phos- distortion. pholipid molecules in cell membranes (Figure B.20). Removal of Convolution difference – Here, strong exponential line broad- the baseline is frequently a prerequisite of peak assignment and ening of the time domain signal (see Spectral analysis) is applied peak area -measurement, and this may be achieved either in the to yield a spectrum in which all of the resonance peaks have time domain or in the frequency domain. been broadened to the extent that they are effectively eliminated Time Domain Methods: Apodisation – On the assumption from the spectrum and only the broad baseline remains. This is that the baseline is due to short T2 components that decay rap- then subtracted from the original spectrum to achieve high-pass idly in the FID, they may be suppressed by multiplying the time filtering. Basic safety standards (BSS) 87 Bateman equation for transient equilibrium Secular equilibrium 1.2 1 0.8 0.6 B 0.4 Parent 0.2 Daughter 0 0 5 10 15 20 Number of daughter half lifes FIGURE B.21 The graph shows the build-up to a secular equilibrium. 25.000 20.000 15.000 10.000 5.0000 .00000 –5.0000 –10.000 –15.000 –20.000 –25.000 The half-life of the daughter is negligible compared to the half-life of the (ppm) parent. FIGURE B.20 Phosphorus (31P) NMR spectrum of the brain showing broad baseline feature. if the activity of the daughter is zero at t = 0. Ad and AP are the activity of daughter and parent nuclei, respectively. λd is the decay constant of the daughter nuclides. BR is the branching ratio for Frequency Domain Methods: Curve fitting – Least squares decay to the daughter when the parent has several decay paths fitting of one or more polynomial functions, or some other math- available. After a time equivalent to 5 daughter half-lives, the ematical function, to the frequency domain spectrum is a popular activity of the daughter almost equals that of the parent nucleus means of baseline correction. (Ad = 0.98Ap), as seen in Figure B.21. Related Articles: Peak areas, Peak assignment, Spectral Related Articles: Transient equilibrium, Secular equilibrium, analysis Bateman equation, Bateman equation for transient equilibrium Basic safety standards (BSS) Bateman equation for transient equilibrium (Radiation Protection and General) (Nuclear Medicine) The Bateman equation for transient equilib- There are two key international organisations that write rium is a special case of the Bateman equation for parent–daugh- basic safety standards for the protection of workers, patients ter relationship. The equation is applicable when the half-life of and the public from the uses of ionising radiation. These are the the parent is longer than the daughter’s half-life, but not long International Atomic Energy Agency (IAEA) and the European enough for λP to be approximated to zero (which will yield a Atomic Energy Community (EURATOM). secular equilibrium). One situation with a transient equilibrium Both organisations take the recommendations set out in the is 99Mo (T½ = 66 h) and 99mTc (T½ = 6 h). The original Bateman reports of the ICRP and present them as standards to be used equation cannot be simplified unless λP = 0, so in the case of a by employers for protecting persons from the risks arising from transient equilibrium, the equation is exposure to ionising radiation. The BSS produced by EURATOM are then adopted as a Directive (i.e. the Basic Safety Standards Directive) published ìé A = ï l A (0) d ´(e-l ù ü pt - e dt ï BR A e d by the European Commission, which requires all member states d í -l ê p - )ú ´ ý + (0) -l t d îïë ld l p û þï of the EU to enact national radiation safety legislation to ensure compliance with the requirements of the Directive. The BSS produced by the IAEA are considered to be best where practice – they are not compulsory, although many countries Ad and AP are the activities of the daughter and parent, around the world outside Europe do adopt the framework and respectively standards set out in the IAEA BSS into their national radiation λd and λP are the decay constants for daughter and parent safety legislation. nuclides, respectively Related Articles: IAEA, EURATOM, ICRP BR is the branching ratio for decay to the daughter when the parent has several decay paths available Bateman equation for secular equilibrium Figure B.22 describes a hypothetical parent–daughter pair with (Nuclear Medicine) The Bateman equation for secular equilibrium T is a special case of the Bateman equation for parent–daughter rela- P = 10 Td and B.R. = 1. When the ratio of the parent and daughter activities is constant tionship. The equation is applicable when the half-life of the parent – the parent and daughter are said to be in transient equilibrium. In is much longer than that of the daughter. One such parent daughter some situations, it is interesting to know when the daughter activ- couple is 226Ra (Tp = 1620 years) and 222Rn (Td = 4.8 days). In such ity has reached its peak, i.e. when dAd/dt = 0. One example is the a situation, λP is close to zero and the original Bateman equation 99Mo–99mTc generator where 99mTc is eluted. The maximum activity (see article Bateman equation) can be abbreviated to is attained if the generator is eluted every 23 h. The time at which the daughter activity reaches its maximum, tmax, can be calculated A t A e dt d ( ) = - p(0)(1- l )´ BR using Activity (a.u.) Bateman equation in parent–daughter decay 88 Beam attenuation Transient equilibrium on the beam entrance surface due to leakage radiation from the 1 collimator and also from geometrical factors such as the source 0.8 dimension and the relative distance between the source and the collimator jaws. The magnitude of the penumbra increases with 0.6 distance from the source and with the depth in phantom/patient. B The usual statement of the beam size refers to the lateral distance 0.4 between the 50% isodose lines at a reference depth in a phantom. Parent An indication of the field size entering the patient is given by the 0.2 Daughter field-defining light. Beam alignment is the procedure in which the field size defining light is made to coincide with the 50% isodose 0 0 2 4 6 8 10 12 14 lines of the radiation beam projected on a plane perpendicular to the beam axis and at a standard SSD or SAD. Number of daughter half lifes FIGURE B.22 The graph illustrates the build-up phase to a transient Beam area equilibrium for a parent–daughter pair (where T (Ultrasound) The ultrasound beam area is defined as the area in P = 10 Td). The dotted line and solid line represent the activity of the parent and daughter nucleus, a specified surface consisting of all points at which the pulse- respectively. A transient equilibrium occurs when the half-life of the par- pressure-squared integral, ppsi, is greater than a specified frac- ent nucleus is approximately 10 times the daughter half-life. tion of the maximum value of the ppsi in that surface. A common specified level is −6 dB. Units are metre squared, m2. Related Articles: Beam width, Pulse-pressure-squared é ù integral 1.44T t pTd æ Tp ö max = ê ( ú ln Further Reading: Report IEC 61157. Ultrasonics-standard T T ç ÷ ëê p - d ) ûú è Td ø means for the reporting of the acoustic output of medical diagnos- tic ultrasonic equipment, standard number IEC/TR2 61157-2007, where TP and Td is the half-life of parent and daughter nuclides, 2nd edn., International Electrotechnical Commission, Geneva, respectively. Switzerland, 2007. Related Articles: Transient equilibrium, Secular equilibrium, Bateman equation, Bateman equation for secular equilibrium Beam arrangement (Radiotherapy) The beam and shielding device geometries are Bateman equation in parent–daughter decay determined taking into account the 3D determination of the tar- (Nuclear Medicine) The Bateman equations are used to describe get position and the critical organ location. Generally, a single parent–daughter relationships, i.e. situations where a radioac- photon beam is of limited use in the treatment of deeply located tive parent nucleus decays to a radioactive daughter nucleus. The tumours as it gives a higher dose at the depth of the maximum activity of the daughter nucleus is described by than at the tumour depth. Single fields are used for palliative treat- ments or for relatively superficial lesions with depth <5–10 cm, ì depending on the photon beam energy. The use of a single photon ïé l Ad = êA d í p(0) ´(e-l l )ù ü pt - e- dt ú ´ ï BRý + A (0)e-ldt beam requires a reasonably uniform dose to the target (±5%), a d - l d îïë l p û þï low maximum dose outside the target and no organs exceeding their tolerance dose. For deep tumours, an arrangement of two Ap(t), Ad(t) are the activities of the parent and the daughter, respec- or more photon beams is required to focus the dose in the target tively, at time t, and λp, λd are their respective decay constants. BR volume, sparing as much as possible the tissues surrounding the is the branching ratio for situations where there is more than one target. Standard beam geometries such as four field box, paral- decay mode for the parent. The last in the equation is the residual lel opposed pair and lateral oblique beams can be used together daughter-product activity that might be present at time t. with conformal shielding to increase the healthy tissue sparing. The Bateman equations can be applied to different situations, In the event that a critical organ or structure is in the path of a and the resulting equation depends on the relationship between beam, more unconventional beam combinations can be used to the parent and daughter decay constants λp and

λd. maximise healthy tissue sparing. It is essential when choosing the Related Articles: Transient equilibrium, Secular equilibrium, beam arrangement to consider the prospective dose distributions. Bateman equation for secular equilibrium, Bateman equation for In some cases, such as in non-coplanar field arrangements, the transient equilibrium physical limitations of the treatment unit and its accessories with respect to patient position must be considered to ensure that no Bayonet catch collisions occur between the gantry and the patient or table. (General) Bayonet style of fitting of an electrical device (as a bulb) is used in some countries instead of screw fitting. The base Beam attenuation of the device uses keyways (and not screw) to connect the device (Radiotherapy) Attenuation is defined as those processes by to the fixture base (bayonet mount). The device is locked in place which a beam of x-ray or gamma radiation is reduced in intensity by pushing it down and twisting to lock firmly. passing through some material. Since ionising photon interaction does not always occur, the interaction processes are statistical in Beam alignment nature. The probability that a photon will interact with matter in (Radiotherapy) The object of beam collimation is to produce one of several possible ways is a function of its energy and of the a radiation beam of some required size and shape. In practice, composition of the interacting material. It is generally assumed the edge of the beam is not sharp as there is always a penumbra that the attenuation of gamma photons is exponential. When an Activity (a.u.) Beam collimator 89 B eam divergence absorber is placed in a gamma beam, the amount of radiations detected on the downbeam side of the absorber depends on the TABLE B.1 absorber thickness, atomic number, density and the amount of Attenuation coefficients scatter radiation that reaches the detector. The amount of scat- ter radiation that can reach the detector depends on the radiation Linear µ cm−1 Mass µ/ρ cm2 beam cross section. When no scatter radiation reaches the detec- /g B tor, the geometry condition is called narrow beam or good geom- Atomic (µ/ρ)(A/N0) cm2/atom etry, whereas when a maximum amount of scattered radiation Electronic (µ/ρ)(A/N0Z) cm2/electron reaches the detector, the geometry condition is called broad beam Notes: ρ is the absorber density, A is the atomic weight, Z is the atomic or poor geometry condition. number and N0 is Avogadro’s number (6.02 × 1023 atoms/gram In narrow beam geometry, a very small area of the absorber is atom). irradiated and the intensity of radiation that reaches the detector is given by an exponential function of the following form: I(x) = I e-mx 0 a primary and a secondary collimator. The primary collimator where defines the largest circular field size which is obtainable by the I(x) is the intensity of the transmitted beam at thickness x bremsstrahlung process of electrons accelerated into the linac I0 is the intensity of the incident beam accelerating tube and it consists in a conical opening into a tung- x is the absorber thickness sten block shaped on the transparent anode, which is used for μ is the linear attenuation coefficient the bremsstrahlung photon production. The photons, emerging from the primary collimator, struck a flattening filter to obtain The intensity I is defined as the rate flow of radiant energy the requested beam homogeneity. The secondary collimator con- across a unit area. The attenuation coefficient is a probability per sists of four movable blocks, two forming the upper secondary photon per unit path length that a photon interaction will occur. collimator and two the lower secondary collimator. The photon In a broad beam geometry, the measured intensity is given by field size ranges from a few millimetres square to 40 cm × 40 cm at the linac isocentre distance. In modern linacs, independent I(x) = BI e-mx 0 jaw movements can provide asymmetric fields whose beam edge where is coincident with the beam central axis. Recently, multileaf col- B is the build-up factor limators (MLCs) have been introduced consisting of up to 120 I(x), I0, x and μ have the same meaning collimating leaves (60 leaf pairs) whose movement motors are controlled by computer. The introduction of MLCs has permit- ted intensity-modulated fields in conformal radiotherapy either in The build-up factor B is introduced to correct for scattered the step-and-shot mode or in a continuous dynamic mode. The radiation. MLCs are covering field size up to 40 cm × 40 cm at the linac The linear attenuation coefficient is made up of three compo- isocentre. A miniature version of the MLC (microMLC) has been nents (τ, κ, σ) to take into account the main interaction processes also introduced, projecting 1.5–6 mm leaf width up to 10 cm × of the photon interaction: τ the photoelectric linear absorption 10 cm at the isocentre distance. The microMLC is used in head coefficient, κ the pair production linear absorption coefficient and σ and neck treatments and in special radiotherapy techniques as the Compton linear absorption coefficient. radiosurgery. In addition to primary and secondary collimators, The total linear attenuation coefficient μ is therefore given by the electron beam collimation is also obtained by adding differ- the sum ent cone applicators. The thickness of the primary and secondary collimators usually corresponds to 3 ten value layers (TVL) in m = tphotoelectric + sCompton + kpairproduction order to attenuate the average primary photon beam to less than 0.1% of its initial value. According to IEC recommendations, the The term ‘linear’ means that these coefficients are measured in maximum leakage should not exceed 0.2% of the open field value. units of inverse length, i.e. cm−1. Sometimes, it is useful to express distances in terms of the mass thickness, which is the product of Beam divergence the absorber density ρ and the thickness x. (Radiotherapy) Photon beams, produced using either a linear The beam attenuation can then be rewritten as accelerator (linac) or an x-ray tube, are emitted in a cone pattern. Beam divergence refers to the increase in the circular cross-sec- I(x) = I e-(m/r)xr 0 tion of a photon beam with increasing distance from the source. The angle of divergence becomes more acute with increasing where μ/ρ is called mass attenuation coefficient, compared with lateral distance from the central axis and as the SSD decreases. the linear attenuation coefficient that does not depend on the den- The maximum beam divergence is limited by the presence of a sity of the absorber. primary collimator in a conventional linear accelerator and by the Sometimes it is convenient to utilise other coefficients that anode angle in an x-ray tube: can be derived from the linear attenuation coefficient μ. The most common coefficients are indicated in Table B.1. A / B = ( f 2 a / fb ) Beam collimator When A and B are the cross-sectional areas at distance fa and fb, (Radiotherapy) The treatment head of linear accelerators incor- respectively, the cross-sectional area is proportional to the square porate a fixed collimator and mobile jaws to achieve the radiation of the distance from the source (IAEA, 2005), as can be seen in beam collimation. The photon beam collimation is achieved by Figure B.23. Beam edge 90 Beam energy Photon source S B Area A = a2 fa Beam edge fb a β Penumbra Area B = b2 20% 80% 50% Field size 50% b FIGURE B.24 Profile through a photon beam identifying penumbra and beam edge. The most obvious specification is the beam spectrum, but this Central axis is extremely difficult to measure directly or model using Monte Carlo. Therefore, alternative specifications have been used and FIGURE B.23 Illustration of beam divergence, from a photon point these are different for different beams – kilovoltage x-ray pho- source. At distance fa from the source S, the field size is A = a2. At a tons, megavoltage x-ray photons and megavoltage electrons. distance fb, the field size is B = b2. (From Podgorsak, E. ed., Review of Kilovoltage X-Ray Photons: The criterion of choice for this Radiation Oncology Physics: A Handbook for Teachers and Students, range is HVL – half value layer, which is the thickness of an atten- International Atomic Energy Agency, Vienna, Austria, 2005.) uating material required to reduce the air kerma rate in air to one half of its original value. HVL is typically quoted in millimetres of aluminium for beams up to 100 kVp, and in millimetres of Abbreviation: SSD = Source to surface distance. copper for kilovoltage beams greater than 100 kVp. HVL is a rea- Related Article: Beam edge sonable guide to allow the user to assess tissue penetration and is Further Reading: Podgorsak, E. ed. 2005. Review of used for the determination of properties in kilovoltage dosimetry Radiation Oncology Physics a Handbook for Teachers and protocols. Students, International Atomic Energy Agency, Vienna, Austria. While the beams have a spectrum of energies (heterogeneous), it is also possible to calculate an ‘effective energy’ of a single Beam edge energy (monoenergetic) beam that would produce the same HVL. (Radiotherapy) An ideal radiation field is when the cross-sectional Megavoltage X-Ray Photons: At this higher energy range, area of the field is constant and non-zero inside the beam aperture the HVL specification does not vary significantly with energy; and zero outside, in which case, the beam edge is defined by the therefore, it is not suitable and other criteria have been developed. field collimators. Practically, this cannot be achieved, as there is One such qualifier is the nominal accelerating potential, a physical penumbra at the beam edge due to a combination of which is simply the energy of the accelerated electron beam as geometrical penumbra, transmission through the collimators and it impacts the target in the head of the linear accelerator. This scattered radiation from the irradiated volume. This penumbra can was used when dosimetry protocols were based on air kerma in be defined as the lateral distance between the 80% and 20% values air measurements, but with the change to absorbed dose to water on a measured profile, relative to the central axis dose. The beam protocols, the penetrating power of the beam as it is attenuated by edge is defined as the 50% point on this penumbra and the field water or tissue has become more prevalent. size is usually defined as the distance between the two 50% points The most common attenuation-based specification is the qual- on the dose profile (Figure B.24). ity index (QI or TPR 20/10) and is a special case of the tissue– Related Articles: Beam divergence, Divergent beam edge, phantom ratio (TPR) for depths of 10 and 20 cm. One of the main Secondary collimator, Dose profile, Penumbra benefits of this is that it is independent of any electron contamina- Further Reading: International Electrotechnical Commission tion of the incident x-ray photon beam. (IEC). 1989. Medical electrical equipment. Medical Electron Other criteria that have been used tended to be related to the Accelerators – functional performance characteristics, depth of maximum dose (dmax), but this makes them susceptible to International Standard IEC 976. any electron contamination at this relatively shallow depth. Another specification is PDD(10), the percentage depth-dose value at 10 cm deep in water, which again raises the problem of Beam energy electron contamination at the depth of maximum dose. It is pos- (Radiotherapy) It is essential that the energy of a beam is known sible to remove this by placing a 1 mm lead foil in the beam, and since the response of most methods employed to measure the there are then additional corrections that must be made to take absorbed dose to water varies with energy. There are a number account of this. of different specifications that can be employed to indicate the Megavoltage Electrons: As with the x-ray photons, the beam energy of the x-ray beam produced by a piece of equipment. of electrons that impacts the patient (or phantom) contains a Beam flatness 91 B eam hardening spread (spectrum) of energies, and therefore the specification of If the flatness or symmetry of the beam is out of the toler- choice has been the mean energy at the patient surface (E0). This ance values, then this may suggest the use of an incorrect filter, or is derived from knowing the depth of 50% dose (R50,D) and

the an error in positioning. Alternatively, it may be that the monitor following relationship: chambers are malfunctioning. The monitor chambers are multi- compartment ion chambers situated in the head of the linac to monitor the delivered dose. However, they are also a vital link in E0 = CR50,D B the feedback circuits that control the path of the beam using steer- ing and bending magnets, in order to maintain the flatness and where symmetry within tolerance. C is a constant (= 2.33 MeV/cm) Cobalt units tend to have flatter profiles than those produced R50,D is quoted in centimetres by linacs, but with wider penumbras due to the large physical source size. The depth of 50% dose (R50,D) is commonly used for the selec- Abbreviations: IPSM = Institute of physical sciences in medi- tion of stopping power ratios and reference depths (zref) in dosim- cine and linac = Linear accelerator. etry protocols and is a similar idea to the penetrating power of the Related Articles: Linear accelerator, Penumbra beam used in x-ray photon protocols. Further Reading: Mayles, W. P. M. et al. 1988. Commissioning Abbreviations: HVL = Half value layer, PDD = Percentage and quality assurance of linear accelerators, IPSM Report No. 54, depth dose, QI = Quality index and TPR = Tissue–phantom ratio. Institute of Physical Sciences in Medicine, York, UK. Related Articles: Calculation of absorbed dose, Mean electron energy, Tissue–phantom ratio, Beam quality, Quality index Beam former (Diagnostic Radiology) The beam former is a metal filter used Beam flatness to produce an x-ray beam with specific hardness and intensity. (Radiotherapy) Beam flatness is a measure of the homogeneity It is used as a compensator while imaging objects with irregular of the dose profile within a certain area of the central axis. The shape. Typical examples of beam formers are beam restrictors and Institute of Physical Sciences in Medicine (IPSM) defines beam wedges used to image the heart (see the article on Beam restric- flatness as the ratio of the maximum dose within the beam by tor) or the bow-tie filter used in computed tomography (see the the minimum dose within the flattened area, i.e. MAX/MIN. eponymous article). The flattened area for a 40 cm field is the portion from the centre Related Articles: Beam restrictor, Bow-tie filter out to 3 cm from the edge of the 50% dose level, as shown in Figure B.25. Beam hardening Similarly, beam symmetry is defined as the greatest value of (Diagnostic Radiology) Beam hardening is the term used to equidistant points from the central axis within the flattened area, refer to the increase in average x-ray beam energy when it passes i.e. A1/A2 or B1/B2 from Figure B.25. through a material (Figure B.26). This occurs in a heteroge- Linear accelerators (linacs) will be designed to give a ‘flat’ neous x-ray beam because the lower energies in the spectrum beam at one particular depth, often 10 cm. To achieve this, flat- are absorbed preferentially. The flat and bow-tie filter present tening filters are placed in the way of the beam, to alter the photon on computed tomography (CT) scanners will remove the low- distribution that exits the head. The photon beam produced by the est energy x-rays from the beam and reduce the effects of beam interactions at the target is sharply forward peaked, which would hardening. produce significantly inhomogeneous dose profiles. The flatten- In CT, when scanning objects with a circular or elliptical cross ing filter is normally circularly symmetric and energy specific, as section, such as a body, the central rays pass through thicker parts thicker filters will be needed for higher energy, and substantially of the object and therefore will undergo a greater amount of beam reduces the dose rate at the centre. However, it also acts as a radi- hardening than those at the periphery. The central channels in the ation filter and changes the energy spectrum across the profile, detector bank will therefore ‘see’ apparently less attenuation than causing beam hardening in the centre of the beam as the low- the outer ones, as a ‘harder’ beam undergoes fewer interactions energy photons are preferentially absorbed. Hence the penetrating (Figure B.27). The effect of this beam hardening can be seen as power of the beam varies across the width, leading to unflatness ‘cupping’ in the image (Figure B.28). This effect can be removed in profiles at depths other than 10 cm. Dose Tube spectrum After 20 cm water After 30 cm water A1 A2 Max Min B1 B2 50% Flattened area 3 cm 3 cm Position across beam keV FIGURE B.25 A diagram to show flatness and symmetry definitions as FIGURE B.26 Increase in average x-ray energy after passage through in IPSM Report 54. Dose is normalised to field centre. water. Relative intensity Beam hardening 92 Beam kernel Uniform cylinder B Ideal projection Projection with beam hardening (a) (b) FIGURE B.29 CT images of human skull phantom (a) without and (b) with iterative bone correction. (Courtesy of ImPACT, UK, www .impactscan .org) Detector channel to the initial electron energy. If the bremsstrahlung photon beam FIGURE B.27 Effect of beam hardening on attenuation profile (CT). enters into a material, a higher attenuation is obtained for the lower energy photons than for higher energy photons. This pro- duces a change in the beam quality and the effect is called ‘beam hardening’. The energy degradation of the photon beam is also related, to a lesser extent, to the Compton scattering of the pri- mary photons. In this case, the consequence is a beam softening. The net photon beam hardening alters the accuracy of the isodose distribution calculated by a treatment planning system, especially at large depths. Beam kernel (Radiotherapy) The dose D(x,y,d) at a given depth d in a flat, homogeneous phantom calculated by the first principle, i.e. con- Without correction With correction volving the relative primary fluence distribution with the profile of the pencil beam distribution at the same depth is given by FIGURE B.28 CT images of water phantom without and with calibra- tion correction for beam hardening. (Courtesy of ImPACT, UK, www D(x, y,d) = òòF(a,b)K(x - a, y - b,d)da db .impactscan .org) where x, y, a and b are lateral distance from the central axis by applying calibration corrections to the attenuation profiles Φ represents the relative fluence distribution for either an open prior to reconstruction. or modified photon radiation field The extent of beam hardening is greater if highly attenuat- K is the two-dimensional cross-sectional profile of the pencil ing materials such as bone or contrast agent are present in the beam dose distribution at the depth d path of the x-ray beam. This can lead to streaks and shading in K is also called the convolution beam kernel the image (as the dark shadows in the middle of the CT scan in Figure B.29a). In these circumstances, the calibration corrections The pencil beam kernel K incorporates the transport of scat- will not be sufficient to remove beam hardening and further soft- tered photons and secondary electrons in the phantom material. It ware corrections, usually iterative bone corrections, are necessary is equal to the cross-sectional profile of a pencil beam at the spec- to reduce these artefacts (Figure B.29). ified depth. Pencil beam dose distributions in water or another tis- Related Articles: Artefact, Bow-tie filter, Cone beam artefact, sue-equivalent material can easily be obtained with Monte Carlo Helical artefact, Image artefact, Metal artefact, Motion artefact, calculation. The cylindrical geometry used for scoring the pencil Partial volume effect (artefact), Ring artefact beam dose distribution is shown in Figure B.30. An 18 MV pencil beam profile at a depth of 5 cm is shown in Beam hardening Figure B.31. The representation is for a kernel at a depth of 5 cm, (Radiotherapy) The production of a high-energy photon beam generated using an 18 MV photon spectrum with the EGS Monte is based on a bremsstrahlung process in which a high-energy Carlo programme. monoenergetic electron beam interacts with the nucleus of a The pencil beam kernel of a 10 MV photon beam at a depth of material. The result of the interactions is a partial or complete loss 10 cm is shown in Figure B.32. of the electron energy and the consequent production of photons. An important consideration in the valuation of the convolu- The resulting photon bremsstrahlung spectra may have energy up tion integral is that the kernel must be spatially invariant, i.e. the Attenuation Beam limiting device 93 Beam modulation Incident pencil beam dose distribution must be invariant as a function pencil beam of the lateral position of the pencil relative to the central axis Depths of the beam. This is not true if, for example the energy spec- trum and the angular distributions of photons vary radially. The dose distributions of pencils at the central axis are therefore not Radii 0 r5 r4 r3 r2 r1 0 exactly the same as those for pencils off axis. However, pencil B d1 beam dose distributions are relatively insensitive to such spec- d tral variations. 2 Further Readings: Mohan, R. and C. S. Chui. 1987. Use of d3 fast Fourier transforms in calculating dose distributions for irreg- d4 ularly shaped fields for three-dimensional treatment planning. Med. Phys. 14:70–77; Storchi, P. R. M. et al. 1999. Calculation d5 of a pencil beam kernel from measured photon beam data. Phys. d6 Med. Biol. 44:2917. Beam limiting device FIGURE B.30 The pencil beam scoring geometry. (Courtesy of Med. (Radiotherapy) A beam limiting device is intended for limiting Phys.) and shaping, or collimating the radiation field. All clinical radia- tion beams used in teletherapy must be collimated. Beam limiting device consists of primary collimator, which defines the largest available circular field, and a secondary moveable collimator, which consists of four blocks (two forming the upper and two forming the lower jaws of the collimator) and which can provide rectangular or square fields. Modern linacs incorporate indepen- dent (asymmetric) jaws that can provide asymmetric fields. Modern linacs are usually provided with multileaf collimators (MLC), which enable arbitrary shaping of the field. MLCs work either in static mode (step and shoot) or in dynamic mode (sliding window). MLCs are added to the secondary collimator or substi- tute the lower jaws of the secondary collimator according to the manufacturer. For very small fields, a micro-MLC can be used. For electron beams, electron beam applicators are used to cre- ate a clinically useful beam. Beam applicators are also used in superficial therapy (with low energy kilovoltage x-ray beams) and orthovoltage therapy (with medium energy kilovoltage x-ray beams). FIGURE B.31 Two-dimensional representation of a typical pencil beam Related Articles: Collimation, Multileaf collimator, photon kernel. (Courtesy of Med. Phys.) MicroMLC, Electron applicators, Asymmetric jaws, Dynamic jaw collimation Beam modulation (Radiotherapy) The pristine Bragg peak of a mono-energetic pro- ton pencil beam would be too narrow to cover the dimensions of most clinical tumours. Therefore, it is necessary to spread out the 101 proton beam both in depth and laterally. This technique is called beam modulation (IAEA TRS-398, 2000). To achieve depth mod- 100 ulation, proton beams of various energies are combined to gener- ate a spread-out Bragg peak (SOBP). A high-energy proton beam 10–1 is used to cover the distal edge of the tumour and a superposition of decreasing energy proton beams is used to cover all depths 10–2 through to the proximal side (Khan and Gibbons, 2014). The two main modulation techniques for spreading out the Bragg peak are passive scattering and active scanning. 10–3 In passive scattering, a range modulator wheel is used to spread the Bragg peak in range or depth of penetration. Two scat- 10–4 tering foils are used to spread the lateral dimensions of the proton beam. A conformal dose distribution to cover the tumour can be 10–5 achieved by using compensators, which shape the beam in the –25 –20 –15 –10 –5 0 5 10 15 20 25 depth direction, and collimators, which shape the beam in the lat- Distance (cm) from interaction point eral direction. Active scanning uses magnets to scan the beam laterally. The FIGURE B.32 Beam kernel of 10 MV photon beam at a depth of 10 cm. depth of penetration is adjusted by either changing the energy of (From Storchi, P.R.M. et al., Phys. Med. Biol., 44, 2917,

1999.) the beam or by inserting material in the beam with a range shifter. Dose deposition kernel Beam quality 94 Beam quality The reader is referred to IPEM Report 75 for schematic dia- where I10 and I20 are the ionisation measurements in a water grams showing passive scattering and active scanning delivery phantom at the depths of 10 and 20 cm for a 10 cm × 10 cm field systems. at the nominal focus surface distance (SSD) of 100 cm. The ratio Related Articles: Bragg peak spreading, Modulation wheel, I10/I20 (Equations B.1 and B.2) is related to the photon attenu- Range modulation, Spread-out Bragg peak (SOBP) ation in the exponential part of the depth-dose curve in water B Further Readings: Horton, P. and D. Eaton. 2017. Design and under reference conditions: Shielding of Radiotherapy Treatment Facilities, IPEM Report 75, 2nd edn., IOP Publishing; IAEA TRS-398. 2000. Absorbed Dose ( 2 é 100 +10) ù Determination in External Beam Radiotherapy: An International I I - 10 2 20 = 10 ê ú e ( ln /HAT) (B.1) Code of Practice for Dosimetry based on Standards of Absorbed ëê 100 + 20 ûú Dose to Water, International Atomic Energy Agency, Austria; Khan, F. M. and J. P. Gibbons. 2014. Khan’s the Physics of where HAT is the half-attenuation thickness of water required Radiation Therapy, 5th edn., Wolters Kluwer Health. to attenuate the photon beam of a factor of 2 in a broad-beam geometry: Beam quality I10 = 1.19e6.93/HAT (Radiotherapy) The radiation beam spectrum is the best descrip- (B.2) I20 tion of the beam quality but it is very difficult to be measured and therefore it cannot be used in the radiotherapy clinical applica- Dosimetric protocols have also suggested the use of tissue–phan- tions. Various other parameters based on beam measurements are tom ratio (TPR) or percentage depth dose at a depth of 10 and 20 utilised as quality indices for x-ray, photon and electron beams. cm in a water phantom as photon beam quality index. X-Ray and Photon Beam Quality: In the case of a monoener- Electron Beam Quality: The beam quality of an electron getic photon beam, the energy of the beam is the only index for beam differs from that in a photon beam since in an electron the beam quality but in a bremsstrahlung continuous spectrum, beam the variation of the physical parameters with depth in an the half value layer (HVL) has been introduced as practical index absorber is diverse. The photon beam fluence decreases continu- for x-ray energies up to 400 keV. The HVL is the thickness of a ously with depth because of the attenuation while the electron specified material that attenuates the x-ray beam to the half of fluence decreases at the same rate as the photon fluence only its initial intensity. The measurement of HVL, based on beam for depths where transient charged particle equilibrium (TCPE) attenuation, should be performed in a narrow beam geometry to is achieved without a significant variation of the energy spec- avoid the scatter component reaching the detector. The HVL is trum of secondary electrons. The maximum energy and the given in aluminium thickness for low energy x-ray and in copper mean energy of primary electron beam decrease continuously thickness for high x-ray energies. A better characterisation of the from a maximum at the phantom surface to zero at the depth x-ray beam quality could be given by the indication of the first of the maximum range in the material because of the continu- and the second HVL. Another beam quality index of a heteroge- ous slowing down process. The electron beam inside the accel- neous photon beam is the effective energy, which is defined as erator tube just before hitting the accelerator window is named the energy of a monoenergetic photon beam that has the same intrinsic electron beam and exhibits a small energy and angular HVL as the heterogeneous x-ray beam. In accelerators, brems- spread. As the intrinsic electron beam passes through the linac strahlung x-ray beams are produced by accelerated high-energy exit window, scattering foil, transmission monitor chamber and electrons hitting a target. The photon beam quality is specified air reaching the phantom surface, the energy and angular spread in megavolts (MV), i.e. the nominal accelerating potential of the of the beam increase significantly. In Figure B.33a and con- highest energy photon in the continuous bremsstrahlung spec- secutive Figure B.33b, the electron energy spectra are shown, trum produced by the electron beam of average energy (MeV) respectively, before the exit window (a), at the phantom surface accelerated by the linac. As the shape of the high-energy x-ray (0) and at a depth z in the phantom (z). In the spectra, the mean spectrum and consequently the beam quality depends on the energy, the most probable energy and the maximum energy of nature and the thickness of the target used to produce the brems- the beam are also shown. strahlung as well as on the various filters or absorbers through In Figure B.34, an electron depth-dose curve is shown with the which the beam has passed, the depth-dose distribution as well indication of R100, R85, R50 and Rp, respectively as the attenuation of the beam could be different for beams speci- To describe the electron beam quality, several parameters are fied by the same nominal energy. This requires a more appropri- therefore used as the most probable energy Ep,0 at the phantom ate indication of the beam quality, which is also essential for the surface, the mean energy E0 on the phantom surface and R50 the dosimetric requirements to choose the energy-dependent param- depth at which the absorbed dose falls to 50% of the maximum eters utilised into the determination of the absorbed dose from dose. The most probable energy Ep,0 on the phantom surface is ionisation measurement. For a high-energy x-ray beam, it is not empirically related to the practical range Rp by Ep,0 = 0.22 + 1.98 convenient to use HVLs because of the practical difficulties in Rp + 0.0025 R 2 p where Ep,0 is expressed in MeV and Rp in cm. avoiding the scatter radiation reaching the detector, and since the The mean electron energy at the phantom surface is related to attenuation coefficient for high Z materials shows a minimum for the half-value depth R50 by E0 = CR50 where C = 2.33 MeV/cm for energies around 2–3 MeV because of the pair production present water. R50 is calculated from the measured value of I50, the depth in the photon beam interactions. As narrow-beam conditions are at which the ionisation curve falls to 50% of its maximum. R50 is not easy to be fulfilled, it has been suggested that broad-beam determined by ionisation measurements in water by depth doses or ionisations could be used to specify high-energy photon beam quality adopting water as reference material. R50 = 1.029 I50 - 0.06(cm) (for 2 £ I50 £ 10cm) Additionally, the water attenuation coefficient decreases gradu- ally as photon energy increases and reaches a minimum for an energy of about 50 MeV. The quality index is indicated by I10/I20 R50 = 1.059 I50 - 0.37(cm) (for I50 > 10cm) Beam reproducibility 95 Beam restrictor φE/φE(Ep) (φE)z/φE(Ep,z) (φE)0/φE(Ep,0) (φE)a/φE(Ep,a) 1.0 (z) (0) (a) B rz r0 ra 0.5 E (MeV) – – – Ez Ep,z Emax,z E0 Ep,0 Emax,0 Ea Ep,a Emax,a (a) (b) FIGURE B.33 Electron energy spectra. — D D—— × 100 15 max Rp R 100 50 85 10 R85 Ds 5 50 R100 0 10 20 30 Ep,0 (MeV) Dx FIGURE B.35 Variation of range parameters with the modal energy at the phantom surface Ep,0. 0 R100 R85 R50 Rp R (cm) FIGURE B.34 Electron depth–dose curve. Further Reading: ISO. 1993. Guide to the Expression of Uncertainty in Measurement, International Organization for Standardization, Geneva, Switzerland. The mean energy at a depth z in a water phantom is related to the Beam restrictor practical range Rp by Ez = E0 (1 - z / Rp ). (Diagnostic Radiology) There are various beam restrictors used In Figure B.35, R100, R85, R50 and Rp are reported versus the in x-ray imaging system. These are diaphragms, collimators, most probable energy Ep,0. wedges, cones, etc. The purpose of the beam restrictor is to pro- duce an x-ray beam with the size of the detector, thus avoiding Beam reproducibility exposure of the non-imaged parts of the body. Usually, these (Radiotherapy) Reproducibility of results of measurements devices are made of highly absorbent metals like lead. indicates the closeness of the agreement between the results of Diaphragm is usually only a metal piece (absorber) with an the same measurements carried out under changed conditions. opening, mounted beneath the x-ray tube. It simply restricts the Reproducibility may be expressed quantitatively in terms of x-ray beam filed. Such circular diaphragm is used in dental x-ray the dispersion characteristics of the results. Reproducibility is tubes (usually restricting the beam to a circle of 6 cm at 20 cm different from repeatability, which measures the success rate from the focal spot). The simple tool (diaphragm) should not be in successive experiments, possibly conducted by the same mistaken with the light beam diaphragm in front of the x-ray tube experimenters. Beam reproducibility is related to the variation housing, what is in fact a collimator. of the output of photon and electron beams produced by a linear A very simple beam restrictor is the cone. This is simply a accelerator. metal extender (cylinder cone) attached to the diaphragm. Its R (cm water) Beam spectrum 96 Beam steering B FIGURE B.37 Special wedge filter used in the collimator for improved subtracted image of the nose. FIGURE B.36 A set of special beam restrictors (wedge filters) used in DSA. (Courtesy of CGR.) X-ray spectrum KV 1 0 0 length may vary – for example 20–40 cm. It has fixed aperture and is often used for radiographs of the skull, as it minimises the secondary radiation (which is normally prominent in radiography of bones). More complex restrictor is the wedge filter (also called absorp- tion filter or compensating filter), which is used for radiography of anatomical parts with significant absorption variation (chest radiography, skull radiography, etc.). For example, an aluminium wedge filter will produce x-ray beam with decreased intensity at one end (where the wedge is thicker). This beam can be useful in 0 20 40 60 80 120 skull profile radiography, as it will produce a reasonable radio- Photon energy (keV) graph of the nasal bones (exposed with the decreased intensity beam), while the other part of the beam (exposed with the rest FIGURE B.38 Typical x-ray beam spectrum (filtered, with max energy of the non-attenuated beam) will produce reasonable image of 100 kVp). Note: The diagram presents only the stopping radiation (the the other facial skull bones. A set of wedge filters are shown in characteristic radiation is not shown). (Courtesy of Sprawls Foundation, Figure B.36. www .sprawls .org) Complex beam restrictors are collimators with specially shaped lead shutters (jaws) to produce x-ray beam. Most often amount of filtration in the beam and the ‘KV’ or electrical poten- these collimators produce rectangular fields (corresponding to the tial applied to the tube. The KV is the principle adjustable factor x-ray film sizes). When those collimators include light sources that is used to optimise the spectrum for specific applications. mimicking the x-ray beam shape, they are called light beam dia- phragm (LBD). Some collimators are shaped to specific anatomi- cal region (e.g. the nose) – Figure B.37. Collimators used in CT Beam steering scanners restrict the beam width to the size of the slice thickness. (Ultrasound) Beam steering allows the operator to alter the angle Some of these devices have a sensor that detects the size of for the beam without moving the transducer. It is an extension of the detector (film cassette) and automatically restricts the beam to the principles used for dynamic focusing as it is concerned with this size. This is achieved by exact movement of the lead jaws of ensuring pulse transmission and echo receipt accounts for the the diaphragm (most often, the Bucky diaphragm). This system is varying distance the focal point is from all the crystal elements also known as positive beam limitation device (PBL). across the array. Whereas dynamic

focusing usually considers the Related Articles: Attenuation, Step wedge, Filter compensat- focal point to be located in the centre of the array, beam steering ing, Diaphragm, Collimator enables focus points to be at the side or even outside the array window, and as long as the elements diverge sufficiently to allow Beam spectrum transmission and receipt from the desired location the same prin- (Diagnostic Radiology) An x-ray beam consists of photons of ciples apply. different energies. The spectrum is the range and distribution of In the figure the focal distance is at P1 > P2 > P3. Therefore, photon energies for a specific beam as illustrated in Figure B.38. during the transmission phase, a pulse from element P1 will The spectrum of an x-ray beam is determined and controlled be sent first, before P2, giving them the ‘head-start’ over P3, by several factors including the tube anode material, type and and they require respectively to ensure all three pulses arrive Photons Beam symmetry 97 Beamforming simultaneously at the focal point. Similarly in the receiving relative to the other fields. Treatment planning systems usually phase, echoes arriving at element P3 and P2 will receive a ‘time- allow the choice of two weighting methods. For treatments at delay’ to ensure the echo has been received at P1 before summing fixed SSDs, the weight represents a multiplying factor for the dose the received signal across all the elements from the focal point to at a depth at the maximum dose, usually expressed as a percent- produce the screen image. age. For isocentric treatments, the field weight is defined as either the relative contribution of the beam to dose at dmax or the relative B contribution of the beam to the dose at isocentre. In this latter case, the beam weights can be altered to attain 100% at isocentre. This is done by applying the weights as a reciprocal of the number of beams. Weights can also be altered for applied weighted beams to achieve the required normalisation at isocentre. Beam width (Ultrasound) The width of an ultrasound beam is an important parameter affecting lateral resolution and resolution in the ele- vation plane. For circular transducers, it is symmetric about the Beam steering (different focal distances) – see explanation in text. beam axis but for many ultrasound devices including imaging systems using array transducers, it varies in different directions. Beam symmetry The beam width is dependent on the distance from the transducer, (Radiotherapy) Beam symmetry is a measure of the homogeneity focusing, gain settings, etc. of the dose profile within a certain area of the central axis. The Beam width can be determined using a hydrophone. A Institute of Physical Sciences in Medicine (IPSM) defines beam hydrophone measures the pressure pulses and the pulse-pres- symmetry as the greatest value of equidistant points from the sure-squared integral (ppsi) is calculated for each point on a central axis within the flattened area, i.e. A1/A2 or B1/B2 from surface at the specified distance from the transducer or along Figure B.25. The flattened area for a 40 cm field is the portion a line in a specified direction, Figure B.39. The beam width is from the centre out to 3 cm from the edge of the 50% dose level, the distance between two points on this specified surface (line) as shown in Figure B.25. in a specified direction passing through the point of maximum Similarly, beam flatness is defined as the ratio of the maxi- ppsi in that surface. It is defined as a fraction of the maximum mum dose within the beam by the minimum dose within the flat- ppsi value in the surface, normally −6 dB, which is one-fourth tened area, i.e. MAX/MIN. of the maximum. Linear accelerators (linacs) will be designed to give a ‘flat’ Related Article: Pulse-pressure-squared integral beam at one particular depth, often 10 cm. To achieve this, flat- Further Reading: International Electrotechnical Commission, tening filters are placed in the way of the beam, to alter the photon International standard, IEC 61157. distribution that exits the head. The photon beam produced by the interactions of accelerated electrons at the target is sharply Beamforming forward peaked, which would produce significantly inhomoge- (Ultrasound) The principle of B-mode imaging is based on neous dose profiles. The flattening filter is normally circularly sequentially emitting short pulses and detecting the time of arrival symmetric and energy specific, as thicker filters will be needed and the amplitude of the echoes, Figure B.40. With a linear array for higher energy, and substantially reduces the dose rate at the transducer, hundreds of line measurements build up a cross-sec- centre. However, it also acts as a radiation filter and changes the tional image. These lines are updated continuously, allowing real- energy spectrum across the profile, causing beam hardening in time imaging with a frame rate on the order of 10–60 images/s. the centre of the beam as the low-energy photons are preferen- tially absorbed. Hence, the penetrating power of the beam varies across the width, leading to unflatness in profiles at depths other than 10 cm. If the flatness or symmetry of the beam is out of the toler- ance values, then this may suggest the use of an incorrect filter, or 100% an error in positioning. Alternatively, it may be that the monitor chambers are malfunctioning. The monitor chambers are multi- compartment ion chambers situated in the head of the linac to monitor the delivered dose. However, they are also a vital link in the feedback circuits that control the path of the beam using steer- ing and bending magnets, in order to maintain the flatness and symmetry within tolerance. Related Articles: Linear accelerator, Penumbra 25% Further Reading: Mayles, W. P. M. et al. 1988. Commissioning Beam width and quality assurance of linear accelerators. IPSM Report No. 54, Institute of Physical Sciences in Medicine, York, UK. Distance Beam weight FIGURE B.39 Definition of −6 dB beam width. The ppsi profile is mea- (Radiotherapy) In case of a multiple field treatment, the beam sured along a line in a specified direction. (Courtesy of EMIT project, weight is the contribution of a single beam to the treatment dose www .emerald2 .eu) ppsi Beam-on time 98 B eam-on time The beam former is the part of the ultrasound scanner that alongside the image marks at which depth(s) the transmitted beam determines and optimises the shape of the beams in both trans- is focused, Figure B.42. This position is set by the operator to the mission and reception mode. As it is very difficult (impossible) to region of interest of the specific diagnostic examination. In case produce well-defined thin sound beams due to diffraction, etc., of multiple focus zones, the scan line information will contain a number of different tricks are used to enable high-resolution information from several beam pulses, each produced with dif- B imaging. In transmission mode for instance, the beam former gen- ferent focus depth, using different number of transducer elements erates the time-delayed electric signals that drive the transducer and delay pattern, Figure B.43. This process increases the lateral elements, Figure B.41. Multiple focus zones, apodisation, dynamic resolution but reduces the frame rate. focusing, dynamic aperture, delaying, summation, amplification, Related Articles: B-mode, B-scanner, Apodisation, Dynamic TGC and A/D-conversion are all processes that take part in opti- focusing, Dynamic aperture, Time gain compensation (TGC) mised beamforming. The scanner computer steers and controls the main part of these processes. Beam-on time When operating an ultrasound scanner, it’s most often pos- (Radiotherapy) This is the time during which the beam is deliv- sible to choose and change the position and number of trans- ering radiation. Originally, treatment machines were controlled mission focuses. When more than one focus point is used, it is by timing devices, examples of which are Cobalt 60, orthovolt- called choosing multiple focus zones. One or several indicators age and superficial machines where machine radiation output was measured in centigray (cGy) per minute (min), i.e. (cGy/min). Beam profile, focus and output display As beam moves along array, echoes detected depend on beam width-affects spatial and contrast resolution FIGURE B.40 The principle of B-mode scanning and the beam widths affection on lateral resolution. (Courtesy of EMIT project, www . FIGURE B.42 Multiple transmit focus position marks. (Courtesy of emerald2 .eu) EMIT project, www .emerald2 .eu) Timing of elements to transmit beam Wave front Time difference Time difference for focussing for steering + For phased array and for doppler steering FIGURE B.41 A group of elements are used to form the transmit beam. Time-delay patterns are used for focusing and steering. (Courtesy of EMIT project, www .emerald2 .eu) Beam’s eye view 99 Bending magnet support it well (especially at high rotation speeds such as 9000 rpm) and the wobbly rotation will lead to an unstable focal spot and mechanical stress of the glass envelope. Although the anode disc is connected to the rotor through a Zone 1 molybdenum stem (which is not a very good thermal conductor), the rotor and the bearings are heated during the x-ray exposure. If B the ball bearings expand and block the smooth rotation, the target surface will immediately be overheated (even melted on places). Due to this reason, special dry-lubricated bearings are used (using metallic silver, lead, etc.). The preparation and lubrication of these Zone 2 bearings are often kept secret by the producers. The high demands to the bearing of the anode (especially for the heavy high-power anodes) led recently to a new design of the x-ray tube. This has been firstly applied by Philips in their Super- Rotalix-Ceramic tube. The anode in this tube rotates on a stem, Zone 3 supported by bearings at each end. This mechanical support allows much larger anodes (i.e. with increased thermal capacity) to be used. From a mechanical point of view, this construction is only possible if the tube uses a metal envelope (with internal ceramic coating). This tube has three high voltage ceramic insulators (alu- minium oxide) – for the two high voltage cables and the anode stem – see article on Ceramic x-ray tube with double bearings. The latest developments in this field are x-ray tubes with spe- cial spiral groove bearings with liquid metal (Ga-In-Sn alloy) – FIGURE B.43 The principle of multiple focus zones. (Courtesy of see the articles on Liquid metal bearing. EMIT project, www .emerald2 .eu) Related Articles: Anode, Rotating anode, Glass envelope, Metal x-ray tube, Liquid metal bearing, Ceramic x-ray tube with double bearing The beam-on time for linear accelerators (linacs) and modern orthovoltage units is governed by a counter on which a number of Becquerel units are set called monitor units. The monitor counts downwards (Radiation Protection) The Becquerel (Bq) is the SI unit of radio- from the number set to zero at which point the beam is switched active decay. It is the quantity of radioactive material in which one off. Modern record-and-verify systems of linacs count the moni- atom is transformed per second: tor units up until the set limit is reached and the beam is switched 1 Bq = 1 transformation per second off. Typically, machine radiation output is measured in monitor The unit was named after the French physicist, Henri units (MU) per 100 cGy, i.e. (MU/100 cGy or MU/Gy). Becquerel, who is credited as having discovered the phenomenon of radioactivity in 1896. Beam’s eye view Prior to the adoption of Becquerel as the SI unit of radioactiv- (Radiotherapy) The technique of displaying the patient structures, ity, the unit was the Curie (Ci), named after Pierre Curie. This unit which simulates the beam geometry in a plane perpendicular to was equivalent to the rate of disintegration of 1 g of radium-226, the central axis is called beam’s eye view (BEV). BEV displays which was 3.7 × 1010 disintegrations per second. Thus, the equiva- the anatomy of the patient constructed from computed tomogra- lence between Bq and Ci is phy (CT) image data as viewed by the point of view of the source of the radiation. The BEV display permits adjustment of the size and the orientation of the radiation beam to generate optimal 1Ci =3.7´1010 Bq beam arrangement to shield patient organ at risk. Divergence from the radiation source is taken into account so that the rela- Related Article: Curie tionship of the beam collimation

with respect to the patient anat- Further Reading: ICRU. 1998. Fundamental quantities omy is displayed correctly for the divergent beam. BEV could be and units for ionizing radiation. ICRU Report 60, International used in planning conformal therapy where high gradient regions Commission on Radiation Units and Measurements, Bethesda, are customised to the 3D shape of the tumour to spare healthy MD. tissues with small margins. BEV display is used not only in the determination of the best location and shape of the radiation fields BED (biological effective dose) but also in verifying the set up of the treatment isocentre. (Radiotherapy) See Biological effective dose (BED) Related Article: Organ at risk Bearing BEIR (Diagnostic Radiology) Inside rotating anode x-ray tubes, the (Radiation Protection) See Biological effect of ionising radiation anode and the rotor are within the glass envelope, supported by special ball bearings (the stator is fitted outside the glass enve- Bending magnet lope). The whole anode is supported by the bearing, which is (Radiotherapy) In many high-energy linacs, the electron beam of prime importance for the stable rotation of the anode. If the from the waveguide is horizontal, and must be rotated 90° in anode disc is very heavy, the single bearing would not be able to order to direct it at the target. Bernoulli effect 100 Beta decay The simplest system rotates the beam through 90°, as seen in Figure B.44, and was the design for early linacs. This design has the advantage that it requires very little vertical height. Unfortunately, any transverse, angular or energy displacement of the electrons entering this system results in a positional displace- B ment for the output electrons, resulting in a broad focal spot. This FIGURE B.46 Schematic of a slalom magnet. design is no longer used commercially, in favour of the following designs. The 270° magnet bends the beam back on itself, crossing room in the treatment head. This design is favoured by Elekta the incident beam, as shown in Figure B.45. One method for (previously Philips). achieving this is the ‘locally tilted pole gap’, which has adjust- able entrant, and exit pole faces, enabling optimal beam focusing. Bernoulli effect Siemens uses this method. (Ultrasound) Bernoulli’s principle states that for an incompress- An alternative method for the 270° design is the ‘three sec- ible fluid, if no work is performed on the fluid and there are no tor uniform gap’ magnet, which is used by Varian, and consists frictional or other energy losses, then for steady flow, there is con- of three uniform-field dipole magnets with interconnecting drift servation of energy along a streamline as follows: tubes. The slalom magnet or 112.5° magnet rotates as shown in rV 2 Figure B.46, giving a small focal spot and requiring less vertical + rgh + P = constant 2 where ρ is the fluid density V is the velocity at a point in the streamline P is the pressure at the point h is the height of the point above a datum g is the gravity The three components of energy are the dynamic energy, pres- sure energy and gravitational potential energy. In the human circulation, the principle and equation are use- ful in describing the change in energy as velocity increases, for example in a stenosis. If there is negligible change in gravitational energy, for example in a stenosis across a short distance, then the equation simplifies to rV 2 + P = constant 2 By comparing pre-stenotic velocities with those in the stenosis, an estimate of the pressure reduction in the stenosis can be made. FIGURE B.44 Schematic of a 90° bending magnet. BERT (Radiation Protection) See Background equivalent radiation time (BERT) Bessel function (Nuclear Medicine) Bessel functions are solutions to the Bessel differential equation 2 d2y dy x 2 + x + y(x2 - a2 ) = 0 dx dx and where α is the order of the function. Bessel functions are used to describe electromagnetic waves, heat conduction and problems related to lattices. In nuclear medicine, Bessel functions are used in image processing, for example image filtering. Beta decay (Nuclear Medicine) The process in which an atom decays by beta- particle emission is called beta decay. It can occur in two different ways: β− and β+ decay. The kinetic energy of beta particles has a continuous spec- trum, which depends on parent and daughter nuclear states par- FIGURE B.45 Schematic of a 270° magnet. ticipating in the decay. Beta decay 101 Beta radiation In a β− emission, a neutron transforms into a proton, an elec- nucleus, a neutron is transformed into a proton with the emission tron and a neutrino. The β− decay is of no greater interest in clinical of an electron, which, for historical reasons, is called a beta par- tomographic imaging since the range of the electron in tissue is only ticle: n → p + e−. An anti-neutrino (ν*) is also emitted. The trans- on the order of a few millimetres; thus, for a β− particle originat- formed (daughter) nucleus has one less neutron and one additional ing inside the human body, the probability of reaching a detector is proton. Gamma rays may also be emitted such that the daughter low. β−-emitting nuclei are used in preclinical imaging and for some nucleus decays to its ground state. B therapeutic purposes. An example of beta decay with no emission of gamma rays, β+ emission, on the other hand, is one of the basic conditions as it decays directly to the ground state, is the decay of phospho- for PET imaging. The different steps in a β+-decay are schemati- rous-32 into stable sulphur-32: cally represented in Figure B.47. β+ decay is essentially a transfor- mation of a neutron into a proton, a positron (positively charged 32 32 P ® S + b- electron) and a neutrino: + n * 15 16 p+ ® n + e+ + v + energy The decay scheme is shown in Figure B.48. Emitted positrons will travel a few millimetres in tissue before The beta particle (an electron, also sometimes called a nega- they are brought to a near stop. The distance travelled depends on tron) emitted can have a kinetic energy of any value up to a definite the positron kinetic energy. When stopped, the positron combines maximum energy, which is characteristic of the nuclide involved. with an electron and for a short while they form a short-lived par- The shape of the spectrum is generally similar for many nuclides – ticle known as positronium. Both particles are consumed in a pro- see Figure B.48. Where more complex spectra are observed, these cess called annihilation, in which two photons are created. These are found to be composed of several simple spectra superimposed. two photons can be registered using a PET imaging system. The The average energy of the beta particles in the spectrum is photon energy is equivalent to the mass of each particle (511 keV) approximately one-third of the maximum energy. The anti-neu- and the photons are emitted in opposing directions (conservation trino carries off the difference in energy between the beta particle of momentum for a stationary electron–positron pair). However, and the maximum energy in beta minus decay. The neutrino is the particles are seldom brought to a full stop, and the residual often omitted from decay equations. momentum results in a deviation from a 180° emission angle. The Related Articles: Beta particles, Decay schemes, Neutrinos, positron range and deviating emission angle degrade the spatial Nucleus, Radioactive decay, Stable nuclei resolution in PET imaging (separate article). A schematic representation of a β+-decay. A 11C-atom decays, Beta particle which produces a neutrino, a 11B atom and a β+-particle. The β+- (General) The term used for an electron when it is emitted from particle is slowed via interactions with surrounding matter. When an atom as a result of one type of radioactive decay – beta decay. Symbol is β− brought to a near or total halt, the positron forms a particle, a . so-called positronium, with an electron. The positronium is con- A beta particle is emitted from an unstable nucleus (radionu- sumed in an annihilation process that emits two photons with clide) as the result of the transformation of a neutron into a proton 180° opposing directions. and an electron: n → p+ + e−. The electron (beta particle) is ejected Related Articles: Annihilation PET, Spatial resolution PET from the nucleus with kinetic energy of any value up to a definite Further Reading: Cherry, S. R., J. A. Sorenson and M. E. maximum energy, which is characteristic of the decay process Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, (Figures B.48 and B.49). Philadelphia, PA, pp. 25–26. The use of the term beta particle is historical as three differ- ent types of emissions from radioactive decay (alpha, beta and Beta decay gamma) were identified in the early days of studying radioactive (Radiation Protection) Beta decay (sometimes called beta minus decay by Ernest Rutherford (1871–1937) and others, but their decay) is one of the mechanisms of radioactive decay. In the exact nature – type, charge and mass – was not known. Related Articles: Radioactive decay, Radioactivity Beta radiation (Radiation Protection) Beta radiation is a specific form of ion- ising radiation arising from the emission of beta particles from 11 6C ν 32 – 15P 14.3 d e– + β– β+ – – + β+ – 11 + 5B 32 16S 511 keV – + Positronium – + z + 1 180° 511 keV FIGURE B.48 The decay of 32 15 P. Energy of the atom is plotted on the FIGURE B.47 Scheme of different steps in a β+-decay. x-axis and the atomic number on the y-axis. Energy of atom Beta+ radiation 102 Biangular anode disc – Electron central orbit E 0.3 Eβmax Coils Coils Filament Magnet B Injector E Deflector βmax Evacuated doughnut Electron beam E Kinetic energy FIGURE B.50 Betatron. FIGURE B.49 Energy spectrum of beta particles in a typical beta decay process. energies is from about 4 to 20 MeV and linacs can easily produce this electron energy range. Abbreviations: linac = Linear accelerator. radioactive material where the radionuclide involved is decaying Further Readings: Mould, R. F. 1985. Radiotherapy Treatment by beta decay. Planning, 2nd edn., Adam Hilger Ltd, Bristol, UK, pp. 133–134; Beta particles are electrons emitted during beta decay. Podgorsak, E. B. ed. 2005. Radiation Oncology Physics: A Related Articles: Beta particles, Beta decay, Electron, Ionising Handbook for Teachers and Students, International Atomic Energy radiation, Radionuclide Agency, Vienna, Austria, p. 134. Beta+ radiation Bethe–Bloch equation (Radiation Protection) Beta+ radiation (beta plus radiation, (Radiation Protection) The Bethe–Bloch equation describes the β+) is composed of beta+ particles, which are commonly called energy lost by interactions of fast-moving charged particles (e.g. positrons (positive electrons), and is emitted from radionuclides protons, alpha particles, but not electrons/beta particles) travers- undergoing radioactive decay through the decay process normally ing matter. It is named after two theoretical physicists: Hans referred to as ‘positron emission’. Albrecht Bethe and Felix Bloch. When the beta+ particle, or positron, leaves the nucleus, it loses Charged particles moving through matter interact with the its kinetic energy by interacting with surrounding atoms. When it electric fields of the electrons and nuclei of atoms in the mate- has lost its kinetic energy, it is annihilated by combining with an rial. The interaction excites or ionises the atoms. This implies the electron from one of the surrounding atoms to form two photons, transfer of energy from the incident particle. referred to as annihilation radiation: The equation describes the energy loss per unit distance trav- elled (otherwise known as the stopping power of the material tra- b+ + e- ® g + g versed) – it describes the absorption process of charged particles travelling through matter, and implies that the particles will have The mass of the particles is converted into electromagnetic radia- a finite range in matter whereby all incident and secondary par- tion according to Einstein’s equation, E = mc2. The two photons ticles have been totally absorbed within the medium (as opposed each have an energy of 0.51 MeV and they travel, to a first approx- to x- or gamma radiation that is attenuated exponentially with at imation, in opposite directions so that their net momentum is zero.

least some transmission through the material and out the other Related Articles: Beta radiation, Beta decay, Radioactive side). decay The Bethe–Bloch equation, together with the Klein–Nishina differential cross-section, Molière scattering theory and others, Betatron attempts to describe interactions between ionising radiation and (Radiotherapy) The betatron was a device that accelerated elec- matter at an atomic level, and forms the mathematical basis for trons and was developed for research purposes in 1940, but its radiation dosimetry based on Monte Carlo statistical modelling. application for treatment in radiotherapy was soon recognised. Related Articles: Stopping power, Collision mass stopping Electrons are accelerated in an evacuated doughnut-shaped power, Linear stopping power, Klein–Nishina differential cross ring by a magnetic field generated by an alternating current of section, Molière scattering theory between 50 and 200 Hz (see Figure B.50). Electrons are injected from a filament and held in a central Biangular anode disc orbit by the magnetic field. The increasing magnetic field induces (Diagnostic Radiology) X-ray tubes with biangular anode discs an accelerating voltage in the direction of the electron ring simi- are used for dual focus tubes, where there is a need of very small lar to the secondary coil of a transformer. The electrons can then effective focal spot. Such tubes are used for cardio, neurologi- be deflected using electrodes placed in the evacuated doughnut. cal and other x-ray examinations requiring radiographs with high In the 1950s, these machines played an important part in radio- spatial resolution. Biangular anode construction is predominantly therapy, but their role has been taken over by linear accelerators. used for powerful rotational anode x-ray tubes. This is partly due to the greater output possible with linacs (10 Gy/ In these tubes, the line-focus principle allows for reducing min for linacs versus 1 Gy/min for betatrons). Although betatrons the effective focal spot, by using target with smaller bevel. From can produce electrons with energies from a few MeV up to about Figure B.51, it is obvious that the same filament wire will pro- 45 MeV, the clinically advantageous sharp cut-off seen with elec- duce smaller effective focal spot at angle α, compared with those trons is lost after about 20 MeV; so in therapy, the useful range of at angle β. Number of β– particles per unit energy interval Bias 103 Binary counter Bidding process (General) A sequence of actions aimed at generating an offer for Anode a project, a product or a service. It generally starts with a bid from the subject interested in acquiring a product, e.g. the hospital purchasing manager describ- W ing the needs and specifications, together with the estimated cost. B targets A group of vendors analyses the bid and responds with their α offers, usually trying to propose products or services exceeding α, β Cathode the basic needs and specifications as well as reducing the cost. -anode The manager will privilege the bid from the vendor who best angles meets the specifications at the lowest cost. Related Articles: Procurement, Tendering process, Disposal Further Reading: Iadanza, E. 2019. Clinical Engineering β Handbook, 2nd edn., Academic Press, Elsevier, ISBN: 9780128134672; Miniati, R., E. Iadanza and F. Dori. 2016. Clinical Engineering (from Devices to Systems), Academic Press, Elsevier, ISBN 9780128037676; Willson, K., K. Ison and S. Tabakov. 2014. Medical Equipment Management, Taylor & FF BF Francis Group, ISBN: 9780429130373. FIGURE B.51 Biangular anode disc (section) producing two different effective focal spots from each angular surface. The smaller effective Bidirectional focal spot is always from the inner surface. (Nuclear Medicine) Bidirectional refers to a process or a flow moving in two different directions (often opposite directions). For The separation of the effective focal spots creates two actual example, the two photons in positron annihilation have bidirec- focal spots (Figure B.51), which allows better heat distribution tional propagation. when consecutive exposures with different focal spots are made (compared with the situation of overlapping filament – see Focal Big data spot, actual). (General) Big data refers to large, complex sets of information, Such x-ray tubes could have a third smaller angle (closer to the rapidly collected in huge quantities. Big data sets are categorised anode stem), thus producing even smaller focal spots (microfocus by key properties, including volume, velocity and variety. x-ray tubes used in macroradiography). Big data categorisation: Related Articles: Stationary anode, Rotating anode, Target, Line-focus principle, Biangular anode disc, Focal spot actual, Focal spot effective, Focal spot • Unstructured – unorganised information, not fitting a pre-determined format Bias • Structured – consists of information already managed (General) Bias is an electronics term referring to the establishment in databases of a background DC voltage or current at a point in a circuit. A bias is designed into a circuit in order to keep a device or circuit The primary use of big data is in data analysis directed towards element operating around a preferred quiescent voltage or current. exploring the relationships and finding correlations between dif- For example, in an amplifier design, a transistor may be biased to ferent types of data, originating from different sources. Big data keep it operating in a linear region of its characteristic. In the sim- has numerous applications in social science, research, industry, ple amplifier design shown later, resistors R1 and R2 act as a poten- finance and healthcare. tial divider to keep the base of the transistor at a predetermined Healthcare specific applications of big data include but are bias voltage. The input AC signal coupled to the transistor base not limited to the analysis of linked data sets from different sec- via a capacitor rides on this bias voltage – see the electric circuit. tors in search of opportunities to improve healthcare, exploratory biomedical research, data-driven analysis and computer-aided Vcc diagnosis. Further Reading: Bulletin of the World Health Organization. R1 2015. 93:203–208. doi: http://dx .doi .org /10 .2471 /BLT .14 .139022; IEEE Big data, https://bigdata .ieee .org/; International Science Council, https ://co uncil .scie nce/p ublic ation s/ope n-dat a-in- a-big -data -worl d/; United Nations Big Data Global Working Group, Vin https://unstats .un .org /bigdata/. Vout Binary counter R2 (General) Binary counter is a digital electronic circuit that counts clock pulses (cycles). The circuit has a clock input (or uses its own clock) and the counting of pulses can be either related to the pulse 0V raising front (edge) or the falling one. The circuit would usually Electric circuit related to bias voltage. have a reset to zero. Usually, the timers of medical equipment use Electric circuit related to bias voltage. various types of binary counters. Binding affinity 104 Bioeffects Binding affinity τ τ τ (Nuclear Medicine) See Affinity Binding energy Binomial (Radiation Protection) Binding energy is the energy required to RF pulses B separate a part of an entity from the whole entity (or in some cases, π/16 3π/16 3π/16 π/16 the entire entity into its separate parts). The term can be applied to a range of objects, from the nucleus of an atom (nuclear binding ΄On resonance΄ energy) to a celestial object (gravitational binding energy). In this magnetisation article, only electron binding energy will be considered. z΄ The binding energy of an electron orbiting around the nucleus of an atom is the energy required to remove it from the atom to y΄ infinity. In effect, ‘infinity’ means a distance such that the elec- x΄ trostatic force of the nucleus is negligible. The process of removal of an electron from an atom, leaving it in a charged state, is called ionisation. ΄Off resonance΄ magnetisation The electrons in each shell (or orbit) of an atom have both potential and kinetic energy: z΄ y΄ x΄ Energyof electrons = Potentialenergy+Kineticenergy Potential energy is normally taken as zero at infinity and so the potential energy of an electron in a shell is negative. The potential FIGURE B.52 On-resonance spins do not precess in a rotating frame energy is numerically greater than the kinetic energy and there- of reference and the resultant flip is a simple accumulation of the flips fore the total energy will have a negative sign. The binding energy applied. Off-resonance spins with frequencies ±1/2τ Hz different from of an electron will be numerically equal to the total energy but of the centre frequency do precess between RF applications (as indicated by opposite sign. the light grey vectors). For hydrogen, Potential energy (PE) = −26 eV case results in the required total of 90°. For an off-resonant spin Kinetic energy (KE) = +12.5 eV with a frequency 1/2τ different from the resonant frequency, the Total energy = PE + KE = −26 + 12.5 = −13.5 eV transverse component slips through 180° between each RF appli- Binding energy = + 13.5 eV cation. As a result, the accumulation of flips in this case amounts to a net flip of zero. Equally, it can be shown that where every second binomial The higher the atomic number of the atom the greater the bind- coefficient is negative, there is a null at the ‘on resonance’ centre ing energy of the electrons in a particular shell, varying for the frequency of the RF spectrum and peaks at frequencies 1/2τ Hz K-shell as the square of the atomic number. The energy of the either side of the centre frequency. Appropriate application of a electrons in the outer shells is complicated by the interactions binomial excitation can then excite either resonant or off-resonant between electrons. tissue. For example, as fat and water have slightly different preces- Related Article: Ionisation sion frequencies, the fat being off-resonance from water by about 220 Hz in a 1.5 T system, either fat or water can be selectively Binomial excitation excited by this technique through an appropriate choice of τ. (Magnetic Resonance) Binomial excitation is a technique using composite RF excitation pulses to achieve frequency selectivity Binomials for use in, for example fat suppression or in magnetisation trans- (General) In algebra, binomial refers to polynomials with two fer. In binomial excitation, a train of RF pulses is applied, with terms, or the sum of two monomials. each pulse having a relative flip angle determined by the binomial coefficients qn,1 qn,2 … qn,n: Bioeffects (Radiation Protection) Ionising radiation causes biological dam- æ n ö n! n,m = age at a cellular level. Such radiation exposure may lead to direct q ç ÷ = è m ø (n - m)!m! interactions on the DNA within the nucleus of each cell, or by more indirect means – interactions with the water and other When binomial coefficients are used to relatively weight a series biological molecules within the cytoplasm, leading to chemical of RF pulses equally spaced apart by a time τ, it can be shown reactions and thus to biological damage. For more detail on these than the resultant RF excitation spectrum has null points at fre- interactions, see Radiation damage. quencies at odd multiples of 1/2τ Hz from the centre frequency. From an understanding of the effects of ionising radiation at Frequencies ‘off resonance’ from the centre frequency by 1/2τ Hz the cellular level, it is possible to relate the resulting cellular dam- will then experience no net excitation. age to observable biological effects at a macroscopic level – i.e. As an example, consider the binomial series (1, 3, 3, 1). For a the effects to the organism (human or animal) exposed. Firstly, total flip angle of 90°, the binomial excitation will consist of the though, it is necessary to classify such effects: four RF pulses: π/16, 3π/16, 3π/16 and π/16. In Figure B.52, note The biological effects of ionising radiation on an exposed that for an ‘on resonance’ spin, the accumulation of flips in this individual are classified as being either somatic, i.e. effects that Biological dosimeter 105 Biological effective dose (BED) occur within the lifetime of the individual irradiated, or genetic, periodically reviewing the levels of radiation to which human which are effects occurring in progeny due to radiation damage to populations are exposed and improving assessment of the somatic the reproductive cells of the parent. Such genetic (or hereditary) and genetic risks of radiation exposure. effects may only become evident after several generations. The BEIR reports include the following: A further classification describes those effects that manifest in individuals who were exposed to ionising radiation in utero, such • BEIR Committee (The Advisory Committee on the B damage

being called teratogenic effects. These effects may be Biological Effects of Ionising Radiation). 1972. BEIR-I: either somatic to the foetus exposed, manifesting in utero or post- The effects on populations of exposure to low levels partum, or heritable (genetic) in nature, affecting future progeny of ionising radiation. Division of Medical Sciences, as the individual exposed in utero reaches adulthood and has chil- the National Academy of Sciences, National Research dren of his or her own. Council, Washington, DC. 2006. (Generally known as Somatic effects may also be further classified by the probabil- the BEIR Report.) ity of occurrence in irradiated populations: • BEIR Committee. 1979. BEIR-II: The effects on popu- Deterministic effects are those definite effects that will occur lations of exposure to low levels of ionising radiation. in all persons exposed to high doses of radiation – the ‘radiation Draft Report. Division of Medical Sciences, Assembly syndromes’ or ‘radiation sickness’. of Life Sciences, National Research Council, National Stochastic effects are those that occur only to a proportion of Academy of Sciences. (Generally known as the BEIR-II those exposed based purely on chance. Cancer is the main sto- Report.) chastic effect. If a population is exposed to ionising radiation, • BEIR Committee. 1980. BEIR-III: The effects on popu- then a number of people will get cancer. The more the radiation lations of exposure to low levels of ionising radiation. dose received, the larger the number of people expected to get Final Report. Division of Medical Sciences, Assembly cancer. of Life Sciences, National Research Council, National All damage expressed in either the person exposed, or prog- Academy of Sciences. (Generally known as the BIER- eny, may also be referred to as adverse effects or adverse radiation III Report, Final.) effects. • BEIR Committee. 1988. BEIR-IV: Health risks of radon Related Articles: Radiation damage, Repair of radiation dam- and other internally deposited alpha-emitters. National age, Adverse effects, Adverse radiation effects, Stochastic effects, Research Council. Committee on the Biological Effects Deterministic effects of Ionising Radiations. National Academy Press, Washington, DC. Biological dosimeter • BEIR Committee. 1990. BEIR-V: Health effects of (Radiation Protection) Biological dosimetry means the detection, exposures to low levels of ionising radiation. National and if possible, the quantification, of exposure to ionising radia- Research Council. Committee on the Biological Effects tion, using biological indicators. There are, in fact, some meth- of Ionising Radiations. National Academy Press, ods to measure the biological effects of the ionising radiations Washington, DC. in the human body. The bio-dosimetry assessment is based on • BEIR Committee. 1999. Health effects of exposure to the simple assumption that the exposure or contamination will radon. BEIR VI: National Academy Press, Washington, produce effects that are measurable (given a certain variability DC. due to various responses). The most sophisticated example is the • BEIR Committee. 2006. Health risks from exposure measurement of chromosome aberrations. This kind of analysis to low levels of ionising radiation. BEIR VII: Phase 2 has been carried out from 1982 and is reliable and sophisticated; National Academy Press, Washington, DC. due to its complexity, it can be applied only in selected cases. The • BEIR Committee. principle is that ionising radiation causes structural chromosome 2005. Health risks from exposure to low levels of ion- aberrations, a proportion of which give rise to chromosome frag- ising radiation. BEIR VII: National Academy Press, ments. When cells divide, some of these fragments form small Washington, DC. nuclei within the cytoplasm. These micronuclei can be counted, providing an in situ biological dosimeter. In the United States, the Federal Office for Radiation Protection is the most competent Related Article: UNSCEAR body in the field of chromosome aberrations. Competent institu- Hyperlink: National Academy Press: http://www .nap .edu/ tions are present in each European country. Other indicators for the exposure to ionising radiation are the lymphocyte lifetime, the Biological effective dose (BED) presence of erythemas of different levels, etc. (Radiotherapy) The physical radiation dose received by a tissue or tumour does not necessarily reflect the resulting biological Biological Effect of Ionising Radiation (BEIR) effect. This will also depend on the radiosensitivity of the tumour (Radiation Protection) Several major international committees or tissue and the fractionation scheme. The biological effective and several national scientific bodies, relevant to radiation protec- dose (BED), also known as the extrapolated response dose (ERD), tion, came into existence in the mid-1950s. Namely, the United is a convenient tool that allows quantification of the biological Nations Scientific Committee on the Effects of Atomic Radiation effect and is derived from the linear-quadratic model, as shown in (UNSCEAR), the Committee on the Biological Effects of Atomic Figure B.53. It allows different fractionation schedules to be com- Radiation (BEAR). pared in terms of their biological effectiveness and for alternative The BEAR was renamed the Committee on the Biological regimes to be calculated, for example to correct for an unwanted Effects of Ionising Radiation (BEIR) in 1972, set up by the interruption of treatment. US National Academy of Sciences. The UNSCEAR and the In Equation B.6, the BED is equal to the log cell kill by BEIR committees have continued their work up to the present, radiation. However, this ignores the effect of repopulation. For Biological effective dose (BED) 106 Biological effective dose (BED) higher than that prescribed may occur within the target volume Linear-quadratic model: and involve critical normal tissues. These should be considered in E = log cellkill from n fractions of d gray any calculations. For example, the BED in the region of a 110% ‘hot spot’ will require the dose per fraction, d, to be multiplied = n(αd + βd2 ) (B.3) by 1.10 in the calculation. Potential overdosing problems can be B = nd(α + βd) (B.4) prevented by performing any BED matching for the calculation of biological equivalent dose fractionation schedules at the ‘hot spot’ Divide by α to obtain BED in units of dose: region, but it should be noted that this may result in a reduction in tumour control, particularly if inhomogeneities extend to ‘cold E = nd 1+ dβ  (B 5) α  α  . spots’ within the tumour volume. In the future, the use of biologi- cal effective dose-volume histograms as suggested by Wheldon   et al. (1998) or the concept of equivalent uniform dose (EUD) as = nd 1 d  + (B.6)  α/β  suggested by Niemierko (1997) may prove of benefit when assess- ing non-uniform planning volumes. Other dose-volume models for tumour control probability and normal tissue control prob- ability are beginning to be incorporated into some computerised FIGURE B.53 The biological effective dose (BED) is derived from the treatment planning systems where they may be useful for compar- linear quadratic model. ing plans but do not as yet provide a reliable model for absolute calculations. The parameters α/β and K in the BED equations are both tis- irradiated tissue that is concurrently repopulating, we must sub- sue specific; therefore, appropriate values must be selected. Since tract the log cell repopulation from the log cell kill, as given in precise values of α/β are rarely known for individual patients, it is Equation B.7: generally recommended that a range of values are used in multiple calculations. In practice, generic values of 5 and 10 Gy for tumour é d ù BED = nd ê1+ ú - K (T - T (B.7) a b delay ) and 3 Gy for late normal tissue effects are frequently used, the ë / û latter often reduced to 2 or even 1.5 Gy in the case of CNS, which where is known to be particularly sensitive to the fractionation regime. T is the overall treatment time Precise values of K and Tdelay are also unknown. For most latere- Tdelay is the delay time (from beginning of treatment) before the sponding normal tissues, the proliferation rate is usually so small onset of significant repopulation that K can be neglected, i.e. K = 0. Dale et al. (2002) provide some K is the biological dose per day required to compensate for guidance on selection of K and Tdelay for various human tumours, ongoing tumour cell repopulation but it should be noted that these values have large uncertainties associated with them. The biological effective dose received by a uniformly irradi- Values of BED can be difficult to interpret because they are ated tissue that is concurrently repopulating. not physical dose. Most clinical experience of the tolerance of Equation B.7 is valid when treatment is delivered with well- normal tissues has been obtained from treatment regimes that use spaced fractions, but when two or more fractions are delivered per 2 Gy fractions. Therefore, ‘equivalence’ can be simply obtained day, there may be increased biological damage as a result of the and expressed as the equivalent dose in 2 Gy fractions (EQD2), incomplete repair of sub-lethal damage between fractions. Factor Figure B.54. Note that EQD2 is sometimes called the normalised h in Equation B.8 reflects this increased damage. A discussion of total dose in 2 Gy fractions (NTD): its derivation can be found in the appendix of the paper by Dale A worked example of the use of BED to find an alternative et al. (2002): fractionation scheme that produces the same biological effect to the tumour is given in Figure B.55. The dose per fraction required é d to deliver the same biological effect to the tumour in 20 fractions (1+ h) ù BED = nd ê1+ ú - K (T - T (B. ) a b delay ) 8 as the regime that delivers 60 Gy in 30 fractions is calculated. In ëê / ûú The biological effective dose equation requires the addition of a factor, h, when multiple fractions per day are delivered to reflect the increased damage resulting from the incomplete repair of sub- If fractionation regime delivering n1 fractions at d1, gray per frac- lethal damage. n2 fractions at d2 gray per fraction, then: The conventional practice to use the α/β value used in the BED calculation as a subscript to both the BED and its numerical     n d 1+ d1  = n d d2 symbol emphasises the point that the biological effective dose is 1 1 2 2 1 /  + α β  (B.9) not a real physical dose. For example, a stated BEDx of 120 Gy    α /β  x indicates that an α/β of x Gy was used to calculate that particular If d1 is 2 Gy per fraction, then n1d1 = EQD2. Rearranging biological dose. Biological doses expressed in Gyx can be added to other Gyx values to provide a measure of resultant effect, but it is not permissible to add Gyz values to Gyx values.  /  EQD2 = n2d d2 + α β 2   (B.10) Since BED is derived from the linear-quadratic (LQ) model,  2 + α /β  the same limits apply regarding its validity (for more details, see the article on LQ model). It should also be noted that a single BED calculation may not be sufficient: where there are significant FIGURE B.54 BED can be transformed back to physical dose equiva- dose inhomogeneities, the BED will vary. ‘Hot spots’ of doses lence via the calculation of the EQD2. Biological half-life 107 Biological purity linearquadratic transformation of dose-volume histograms in Maintain dose to tumour. Substitute into Equation B.9, n1 fractionated radiotherapy. Radiother. Oncol. 46:285–295. 20, d1 = ?, n2 = 30, d2 = 2 Gy: Biological half-life 20d 1+ d1   1  = 1  +  30 × 2 1 2  (Nuclear Medicine) The disappearance with time of a pharma-  0  10  ceutical in a biological system can be described by a law similar B 2d2 1 + 20d1 − 72 = 0 to that of radioactive decay. The biological half-life is therefore defined as the time required for half of the pharmaceutical to dis- Solve for d1 appear from the biological system. The biochemical elimination of this substance from the body is achieved through processes d1 = 2.81 Gy such as excretion, metabolism and diffusion. Equivalent dose in 2 Gy fractions for normal tissues. Further Readings: Chandra, R. 2004. Nuclear Medicine Substitute into Equation B.10, n2 = 20, d2 =

2.81 Gy: Physics the Basics, 6th edn., Lippincott Williams and Wilkins, Philadelphia, PA; Cherry, S., J. Sorenson and M. Phelps. 2003. EQD2 = 20 × 2.81 2.81+ 3   2 + 3  Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA. = 65.3Gy3 Biological parameter (Radiotherapy) In radiotherapy, a biological parameter is a term in a mathematical model used to describe a radiobiological phe- FIGURE B.55 Worked example. nomenon. An example of such a parameter is the alpha–beta ratio in the linear-quadratic model that describes the response of mam- malian cells to radiation. this example, repopulation effects are ignored and generic values Related Articles: Alpha–beta ratio, Linear-quadratic (LQ) of α/β of 10 Gy for the tumour and 3 Gy for normal tissue are model, Normal tissue complication probability, Radiobiological used. The effect of this alternative regime on the normal tissues models, Tumour control probability can be evaluated by calculating the equivalent dose in 2 Gy frac- tions. The clinician can draw on their clinical experience of toler- Biological purity ance doses to determine if this regime will be acceptable in terms (Nuclear Medicine) Biological purity may be defined as the of the probability of inducing an adverse effect. It is important to absence of undesirable biological components in a radiophar- note that it is not possible to maintain the BED both for tumour maceutical, for example free from microorganisms and toxic and for normal tissues. As expected from a hypofractionated microbial by-products, such as bacterial endotoxins. Almost all regime, the effect of maintaining the same biological effect to the radiopharmaceuticals are intended for parenteral administration, tumour increases the biological effect to normal tissues. i.e. given intravenously, subcutaneously or muscularly, and must More details on the use of BED with many worked examples be prepared by aseptic processing. can be found in the book by Dale and Jones (2007). Measurement of the sterility, microbiological control and bac- In the United Kingdom, the Royal College of Radiologists’ terial endotoxin testing (BET) must be performed regularly and 2008 guidance on dealing with unplanned interruptions of radio- on randomly selected batches of the radiopharmaceutical to check therapy recommend that radiobiological calculations are per- the sufficiency of the aseptic technique, or whenever there might formed by suitably trained physicists or clinicians, preferably be a need for it. who have attended an appropriate training course. It cannot be All sterile radiopharmaceuticals should contain no pathogenic emphasised enough that quantitative radiobiological assessments or nonpathogenic living organisms, meaning that all products that inform but do not replace clinical judgement. are intended for parenteral administration must be sterilised by Abbreviations: BED = Biological effective dose, ERD = autoclaving or by membrane filtration. The most common method Extrapolated response dose, EQD2 = Equivalent dose in 2 Gy frac- is to filter the solution through sterile 0.22 μm Millipore filters. tions and NTD = Normalised total dose in 2 Gy fractions. All products designed for intravenous administration are Related Articles: Adverse effect, Alpha-beta ratio, Dosevolume required to be free from pyrogens. These are metabolic products histogram, Fractions, Fractionation, Interruption of treatment, of microorganisms that cause a pyretic response (fever) in the Linear-quadratic (LQ) model, Normal tissue complication prob- patient between 45 and 90 min post injection. Endotoxin is the ability, Repopulation, Radiosensitivity, Tolerance, Tumour control most significant pyrogen and detected by BET. Endotoxin is not probability removed by membrane filtrations, which is why a solution may Further Readings: Dale, R. and B. Jones. 2007. Radiobiological be sterile but not pyrogen free. Since the most likely sources of Modelling in Radiation Oncology, British Institute of Radiology, pyrogens are impure water and chemicals used in the prepara- London, UK; Dale, R. G. et al. 2002. Practical methods for com- tion, pyrogenic contamination can be prevented by using high- pensating for missed treatment days in radiotherapy, with particu- quality water and chemicals, and glassware dry-heated at 250°C lar reference to head and neck schedules. Clin. Oncol. 14:382–393; for 3 min. Fowler, J. F. 2006. Development of radiobiology for oncology – A Related Articles: GMP, Quality control, Radionuclide purity, personal view. Phys. Med. Biol. 51:R263–R286; Niemierko, A. Radiochemical purity 1997. Reporting and analyzing dose distributions: A concept Further Readings: European pharmacopeia, European of equivalent uniform dose. Med. Phys. 24:103–110; The Royal Directorate for the Quality of Medicines (EDQM), Council College of Radiologists. 2008. The Timely Delivery of Radical of Europe [http: / /www .edqm .eu /e n /eur opean -phar macop oeia- Radiotherapy: Standards and Guidelines for the Management of publi catio ns -14 01 .ht ml]; Kowalsky, R. J. and S. W. Falen. 2004. Unscheduled Treatment Interruptions, 3rd edn., Royal College Radiopharmaceuticals in Nuclear Pharmacy and Nuclear of Radiologists, London, UK; Wheldon, T. E. et al. 1998. The Medicine, 2nd edn., American Pharmacists Association, Biological response models 108 B ipolar pulse in amplifier for radiation detector Washington, DC; Saha, G. B. 2004. Fundamentals of Nuclear Two-directional or biplane imaging specifically overcomes Pharmacy, 5th edn., Springer, New York; Zolle, I. ed. 2007. some of the limitations of a one-directional view through a Technetium-99m Pharmaceuticals – Preparation and Quality body, which does not provide depth information. Control in Nuclear Medicine, Springer, Heidelberg, Germany. A major application of biplane imaging is for the heart. With the two fluoroscopic images, the physician can view the coronary B Biological response models arteries from two directions as they are being filled with contrast (Radiotherapy) Biological response models are mathematical media. Defects such as a stenosis, which are visible in one plane, models that describe radiobiological phenomena. In general, might not be visible in the other plane because of the many differ- biological response models describe how the probability or fre- ent directions and orientations of the arteries. quency of a specific biological response changes with some other Biplane (and sometimes multi-plane) imaging is an advantage parameter, usually dose level. Examples of such models include for fluoroscopic (or radiographic) imaging. Other imaging modal- the linear-quadratic model, tumour control probability and nor- ities such as CT and MRI present the same effect by reconstruct- mal tissue complication probability. ing images in multiple planes. Related Articles: Linear-quadratic (LQ) model, Normal tissue Related Article: Biplane cine system complication probability, Radiobiological models, Tumour con- trol probability Bipolar gradient (Magnetic Resonance) A bipolar gradient is composed of two gra- Biological target volume (BTV) dient pulses that are equal in magnitude and duration but opposite (Radiotherapy) The gross tumour volume (GTV), clinical tar- in sign. This re-phases stationary spins while flowing spins expe- get volume (CTV) and planning target volume (PTV) are rience a phase shift (Figure B.56). The velocity f in the direction well-established volumes used in radiotherapy planning. The of the gradient corresponds to a net shift in the phase image. The biological target volume (BTV) is another planning volume that main application is in flow quantification and in phase-contrast allows for inhomogeneous dose prescriptions. angiography. By putting two bipolar gradients of opposite sign For example, the tumour cell density may not be homogeneous together (back-to-back), the phase of the uniformly flowing spins throughout an organ, or even within a tumour, within an organ. If can be refocused. In this way, the signal loss due to flow-induced there is a region within the PTV that is known to be particularly dephasing is minimised in gradient-echo sequences. aggressive or prolific, then this region can be further delineated Related Article: Phase contrast angiography to become the BTV and receive a higher dose. This information Further Reading: Haacke E. M., R. W. Brown, M. R. Thomson is now becoming more readily available with the complemen- and R. Venkatesan. 1999. Magnetic Resonance Imaging. Physical tary use of modern imaging techniques, including MR spectros- Principles and Sequence Design, John Wiley & Sons, New York. copy, SPECT and PET. This approach is called ‘biological image guided radiotherapy’. Bipolar pulse in amplifier for radiation detector Abbreviation: BTV = Biological target volume. (Radiation Protection) The linear amplifier for radiation detector Related Articles: Gross tumour volume (GTV), Planning tar- is used to amplify and shape a linear tail pulse (voltage or current) get volume (PTV), Clinical target volume (CTV) (Figure B.57). The amplitude of a pulse (signal) may be used to Further Readings: ICRU. 1993. Prescribing, reporting evaluate the energy of radiation recorded by a detector. and recording photon beam therapy. ICRU 50, International If monopolar pulses (Figure B.58a) are used, their amplitude Commission on Radiation Units and Measurements, Washington, is measured relative to a baseline (zero). The shift of the baseline DC; ICRU. 1999. Prescribing, recording and reporting photon causes an error in estimation of a pulse amplitude, thus in radia- beam therapy. ICRU Report 62 (Supplement to ICRU Report 50), tion energy. The bipolar pulse (Figure B.58b) has almost equal International Commission on Radiation Units and Measurements, amounts of positive and negative area. Thus, the net voltage is Washington, DC. zero and a bipolar pulse can pass by a capacitor without alteration Biplane cine system (Diagnostic Radiology) Some fluoroscopic systems need to pres- Bipolar ent simultaneous images of an organ from two angles. This is gradient very important for cardio examinations (invasive radiology). pulse These systems use two C-arms, each with a separate x-ray tube and image intensifier, connected to a separate cine cam- EPI era. Contemporary biplane systems may use fast digital image recording instead of cine camera. The movement of both C-arms RF excitation pulse Signal readout around the patient should be independent to allow various angles of observation, as well as space for the surgical intervention. It is possible that the movement of the C-arms is isocentric, thus always keeping the heart in the middle of the observation field. Naturally, the biplane fluoroscopic system has two TV monitors Net phase – one for each viewing angle. Phase shift Related Articles: Fluoroscopy, Image intensifier, C-arm FIGURE B.56 A ramped trapezoidal bipolar gradient re-phases the sta- Biplane imaging tionary spins (continuous line) while the flowing spins experience a phase (Diagnostic Radiology) There are various x-ray procedures that shift (dotted line) that corresponds to average velocity in a reconstructed benefit from simultaneous imaging in two directions or planes. phase image. Birdcage coil 109 Bismuth germanate (BGO) Radiation source Radiation detector Pre amplifier Amplifier SCA or MCA B Recording device FIGURE B.57 Scheme of basic detector and amplification electronics. V(t) V(t) t t Base line shift (a) (b) FIGURE B.58 Pulse waveforms: (a) monopolar and (b) bipolar. of the baseline. The signal-to-noise ratio (SNR) is poorer after bipolar shaping than after monopolar, but the pulse width is the same. The monopolar shaping is used at low counting rates, whereas bipolar shaping at high counting rates. Further Reading: Knoll, G. F. 2000. Radiation Detection and Measurement, John Wiley & Sons, Inc., New York. Birdcage coil (Magnetic Resonance) A birdcage coil is a particular design for an RF volume coil that can be used for both RF transmission and reception in MR imaging. The basic structure of a birdcage coil consists of a number of parallel conductors distributed at equal spacing around the surface of a cylinder and orientated along the direction of the static magnetic field. Birdcage coils provide the high uniformity required of volume coils, providing good constancy in gain throughout an imaged volume when used in receive mode and by the principle of reciprocity, good uniformity FIGURE B.59 Birdcage coil design. in flip angle in the imaged volume when used in transmit mode (Figure B.59). Theoretical analysis shows that an ideal cylindrical volume also less fragile and hygroscopic than NaI(Tl), allowing a wider coil with a sinusoidal distribution of current with radial angular use in more demanding conditions. One of the disadvantages of position ϕ on the coil surface demonstrates excellent uniformity BGO is the relatively low light yield, which is only 10%–20% of excitation throughout a volume in transmit mode (and by reci- that of NaI(Tl). Another disadvantage is the high refractive index procity, excellent uniformity in receive mode). The distribution (2.15), which makes the transition of light photons between the of current in the conductors of a birdcage coil can be shown to crystal and photomultiplier tube less effective. approximate this ideal. BGO has a relatively short signal decay time compared to NaI(Tl) and some other scintillators. Bismuth germanate (BGO) Further Reading: Knoll, G. F. 2000. Radiation Detection (Nuclear Medicine) Bismuth germanate (BGO) is an alternative

to and Measurement, 3rd edn., John Wiley & Sons, New York, pp. the common NaI(Tl) crystal. It has a high atomic number (83) and 239–240. high density (7.13 g/cm3) and is therefore suited for the detection Related Articles: Inorganic scintillators, Scintillators, NaI(Tl) of high-energy photons such as those from positron emitters. It is detector crystal Bit 110 B loch equations Bit with advanced periodontal involvement because of the increased (General) Bit (binary digit) is the basic unit of information storage coverage of the alveolar bone (Figure B.60). in a computer system. The bit can take two values – 0 or 1. A byte is The name ‘bitewing radiograph’ derives from the original a collection of 8 bits, representing 28 = 256 values (from 0 to 255). technique which required the patient to bite on a small wing (also File sizes are usually expressed in kilobytes (1 kB = 103 bytes), named tab) attached to an intraoral film packet. B megabytes (1 MB = 106 bytes), etc. Modern techniques use holders and digital image detectors instead of wings and films, but the terminology and clinical indi- Bit depth cations have remained the same (Figure B.61). (Diagnostic Radiology) See Matrix size The ideal acquisition technique is described in detail else- where. Specific requirements include: Bite-block (Radiotherapy) The aim of planning radiotherapy is to deliver a • Patient position radiation dose to an accurately localised target to ensure control • Image receptor placement, size and orientation of the tumour with minimal effect on the surrounding tissues. • Number of images best suited for a specific patient In treatment for head and neck cancer, the tongue and mandible • Vertical and horizontal angulation and point of entry of are not static structures. Moreover, during a fractionated course the x-ray beam of treatment, there are day-to-day unavoidable variations in the position and alignment of the patient. To minimise these varia- A mismatch between the ideal and actual acquisition technique tions, a face mask or a bite-block or both can be used. In treat- can lead to sub-optimal or incorrect results. ing maxilla, a bite-block can reduce the volume of the tongue by a compression on the dorsum of the tongue in a reproducible Blackening position. (Diagnostic Radiology) See Film blackening Bitewing radiograph Blank exposure (Diagnostic Radiology) (Diagnostic Radiology) A blank exposure is a radiographic film Bitewing radiograph depicts the crowns of the teeth and the that was not exposed because of an equipment malfunction, oper- maxillary and mandibular alveolar crests, allowing: ator error or other abnormal condition. • Detection of interproximal caries Bloch equations • Monitoring the progression of dental caries (Magnetic Resonance) The Bloch equations are differential equa- • Assessment of existing restorations and the periodontal tions describing the behaviour of a single spin magnetisation vec- status tor in a magnetic field. The basic equation is given by Two types of bitewing radiograph can be acquired. The most æ M / x T2 ö common is the horizontal bitewing, obtained when the image dM ( = M ´ B) ç ÷ g · - ç My / T2 dt ÷ detector is aligned with its long axis horizontally. For a vertical bitewing, the long axis is aligned vertically; it is used in patients ç è(M0 - MZ ) / T ÷ 1 ø FIGURE B.60 Horizontal (A) and vertical (B) bitewing. Block design 111 B lock design B FIGURE B.61 Wing attached to an intraoral film packet (A) and an example of a holder (B). where The vectors M→ and B→ are the magnetisation vectors and applied magnetic field respectively γ is the gyromagnetic ratio M0 is the spin magnetisation equilibrium value Mx, My and Mz are the components of M→ along the directions A B A B i, j, and k The first term reflects the interaction of the M→ and B→ vectors, and represents simple precession where B→ is the static main field B→0. The second term represents the observed decay of the transverse components of magnetisation Mxi and My j according to T2 relaxation, and the recovery of the longi- tudinal component of magnetisation Mzk to M0 according to T FIGURE B.62 A block design consisting of (A) on epochs and (B) off 1 relaxation. epochs and, below, the corresponding BOLD response. The equation can be solved to determine the time course of the magnetisation from some initial magnetisation M→ in response to an applied B field consisting of an RF field B1(t) and the static Further Reading: Amaro, E. and Barker, J. 2006. Study field B→0. design in fMRI: Basic principles. Brain Cogn 60:220–232. Block design Block design (Magnetic Resonance) During a functional MRI (fMRI) study, (Radiotherapy) The main jaws of a linac are able to produce fields the subject undertakes a series of tasks known as a paradigm. of various sizes, but are limited to simple squares or rectangles. These tasks may be arranged either in a block design (also known One of the major limitations of this is that tumours have not such as a boxcar design), or an event-related design. simple shapes, and so by selecting a sufficiently large field to In the simplest type of AB block design, an epoch of stimulus fully encompass the tumour, additional healthy tissue will also (an ‘on-period’, A) is followed by an epoch of rest (an ‘off-period’, be within the field. It is therefore desirable that the dose to this B). Typically, they may each be 30 s long, with a 5 min scan con- healthy tissue is minimised while maintaining the dose level to sisting of five blocks and five intervals. As the stimulus and rest the tumour, and this has traditionally been achieved by designing epochs are distributed evenly throughout the scan, the stimulus blocks to shield the healthy tissue. and rest signals should be equally affected by temporal variations In some cases, these blocks have simple shapes that just cut in scanner sensitivity, subject concentration and subject motion off the corners of the field, but with the increasing use of con- (Figure B.62). formal radiotherapy, the design of blocks became more complex. Statistically, block designs are very powerful. By carefully The blocks were designed according to the tumour shape as seen choosing appropriate block lengths, it can be ensured that the on the BEV image on the treatment planning computer. The avail- subject’s haemodynamic response function maximises during ability of MLCs on most linacs has now significantly reduced the the on-period, and returns completely to baseline during the off- workload involved in making blocks, but there are still certain period. For this reason, the block design often produces strongly situations where a block is still required, i.e. shielding of the cen- significant results. tral part of the field. The basic design can be elaborated by introducing addi- The blocks are typically made from a low-melting-point alloy tional blocks of stimuli (e.g. ABC block design), or randomis- (e.g. Lipowitz’s alloy, cerrobend) and are on the order of 7 cm ing the order of the different blocks (ABBCACB… instead of thick in order to significantly attenuate (to approximately 5%) the ABCABCABC.…). dose relative to the open field value. Related Articles: fMRI (functional magnetic resonance imag- Abbreviations: BEV = Beam’s eye view and MLC = Multileaf ing), Haemodynamic response function, Event-related design collimator. Block transmission factor 112 Blood–brain barrier leakage Related Articles: Block transmission factor, Block tray, Conformal radiotherapy, Custom blocking, Low-melting-point TABLE B.3 alloy, Cerrobend, Beam’s eye view Blocking Tray Attenuation Factors Type of Blocking Tray Attenuation Factor B Block transmission factor Low-density metal (honeycomb) 0.995 (Radiotherapy) Treatment volumes are complex in shape and radiation fields often require additional shaping to treat the Perspex (6 mm) 0.970 tumour and avoid radiosensitive structures near the tumour. This Perspex (10 mm) 0.951 shaping can be achieved by the use of customised blocks placed in the beam, which will attenuate the radiation. These blocks can be fabricated from low-melting-point alloy or lead and must be able to attenuate or block the beam so that underlying structures output from the machine in order to take the attenuation of the are adequately protected. The degree to which they can block the tray into account. See Table B.3 for examples of tray attenuation radiation is measured by the beam blocking attenuation factor and factors (Metcalfe, 1997, p. 233). is the ratio of the dose at a point in the patient or phantom with the Blocks may be placed on the tray if it is flat, i.e. gantry at 0° block present to that with it absent: or 180°. When oblique or lateral beams are blocked, the blocks are attached to the tray by means of screws. The screw holes are drilled through the tray and into the block in the case of Perspex. B = Db t ´100 (B.11) Low-density metal (of the kind used to construct aircraft, such as Dub the alloy Dural) can also be used. These trays are thin and easy to work with and are made with many holes drilled into them. Block attenuation factor, Bt, where Db is the dose measured with Related Article: Block transmission factor the block present and Dub is the dose measured with the block Further Readings: Bentel, G. C. 1996. Radiation Therapy absent. Planning, 2nd edn., McGraw-Hill, New York, pp. 219–241; The block transmission factor Bt (<1.00) can be applied to Metcalfe, P., T. Kron and P. Hoban. 1997. The Physics of the percentage depth dose at the point of interest in tissue for the Radiotherapy X-Rays from Linear Accelerators, Medical Physics unblocked field so that the dose at that point can be calculated. Publishing, Madison, WI, p. 233; Walter, M. 2003. Textbook of Block transmission factors can be measured by placing a radia- Radiotherapy Radiation Physics, Therapy and Oncology, 6th tion detector at the isocentre and recording the response with and edn., Churchill Livingstone, Edinburgh, New York; Williams, without the block in the beam. These measurements are made J. R. and D. Thwaites. 2000. Radiotherapy Physics in Practice, under narrow field conditions to minimise collimator and phan- Oxford Medical Publications, Oxford, UK. tom scatter effects. Measurements may be done with a build-up cap on the detector to account for backscatter, which would be present in water. See Table B.2. Blocking layer It is also possible to modulate the beam intensity by varying (Diagnostic Radiology) Blocking layer is a common term refer- the thickness of the block so as to produce varying degrees of ring to a layer of material included in a radiation detector or image transmission and hence modulation of the dose under the block. receptor to block exposure from an unwanted type of radiation. Further Reading: Metcalfe, P., T. Kron and P. Hoban. 1997. The Physics of Radiotherapy X-Rays from Linear Accelerators, Blood–brain barrier Medical Physics Publishing, Madison, WI, pp. 232–233, 285. (Magnetic Resonance) The blood–brain barrier (BBB) is a cellu- lar structure that acts primarily to protect the brain from certain Block tray chemicals in the blood, while still allowing essential metabolic (Radiotherapy) This is a tray on which attenuation blocks can be function. The barrier is composed of very tightly packed endothe- fixed to allow parts of the radiation beam to be shielded. The tray lial cells in brain capillaries. This higher density alters the perme- is attached to the treatment head at a suitable distance from the ability of brain capillaries, so that some substances are prohibited patient, which does not lead to scattered radiation from the block- from entering the brain tissue, while other substances are permit- ing material reaching the patient. The tray is made from Perspex ted to enter freely. or some other material that does not attenuate the beam appre- ciably. Nevertheless, a factor needs to be applied to the radiation Blood–brain barrier leakage (Magnetic Resonance) An intact BBB is a prerequisite for using a common gadolinium-chelate MRI contrast agent as an intravas- TABLE B.2 cular tracer for the assessment of cerebral perfusion in dynamic Examples of Block Transmission Factors susceptibility contrast MRI (DSC-MRI). When the BBB is dam- aged, the contrast agent will leak into the extravascular extracel- Thickness and Type of Block Transmission (%) lular space. This may lead to reduced T2* as well as T1 relaxation 10 cm lead 0.7 times of the extravascular spin population. These two effects will

counteract the signal loss normally caused by the contrast-agent- 10 cm low-melting-point alloy 3.8 induced susceptibility effects in DSC-MRI, leading to errors in, Source: Metcalfe, P. et al., The physics of radiotherapy x-rays from linear primarily, the quantification of cerebral blood volume (CBV). accelerators, medical physics publishing, madison, wi, 285, pp. Two ways of minimising the effects of BBB leakage in DSC- 232–233, 1997. MRI have been presented. One approach is to correct for the effects using algorithms in the post-processing step, as proposed Blood oxygenation level–dependent (BOLD) contrast 113 Blue light hazard by, for example Weisskoff et al. (1994), Haselhorst et al. (2000) brain region is marked by increased MR signal on T2* and Vonken et al. (2000). The second approach is to give a small weighted images (typically, the signal intensity rises by pre-dose of contrast agent before the actual DSC-MRI experi- a few percent). ment (Sorensen and Reimer, 2000). This pre-dose will lead to a shortened T1, and when the main bolus arrives, the T1 relaxation is In fMRI, BOLD responses are detected by statistical analysis more or less saturated and no further shortening of T1 will occur. of a time series of images. The intensity of each voxel within B Related Articles: Blood–brain barrier, Dynamic susceptibil- the image is typically statistically compared to a model of the ity contrast MRI, Cerebral blood volume expected BOLD response to determine whether activation of a Further Readings: Haselhorst, R., L. Kappos, D. Bilecen, particular brain region has occurred. K. Scheffler, D. Mori, E. W. Radu and J. Seelig. 2000. Dynamic As the magnitude of the BOLD response varies with field susceptibility contrast MR imaging of plaque development in strength, fMRI sensitivity can often be improved by moving to multiple sclerosis: Application of an extended blood-brain bar- higher field strengths. rier leakage correction. J. Magn. Reson. Imaging 11:495–505; Related Articles: fMRI (functional magnetic resonance imag- Sorensen, A. G. and R. P. Cerebral. 2000. MR Perfusion Imaging; ing), Oxyhaemoglobin, Haemodynamic response function Principles and Current Applications, Georg Thieme Verlag, Further Reading: Arthurs, O. J. and S. Boniface. January Stuttgart, Germany; Vonken, E. P., M. J. van Osch, C. J. Bakker 2002. How well do we understand the neural origins of the fMRI and M. A. Viergever. 2000. Simultaneous quantitative cerebral BOLD signal? Trends Neurosci. 25(1):27–31, Review. Erratum in: perfusion and Gd-DTPA extravasation measurement with dual- Trends Neurosci. March 2002, 25(3:169); Kwong, K. et al. 1992. echo dynamic susceptibility contrast MRI. Magn. Reson. Med. Dynamic magnetic resonance imaging of human brain activity 43:820–827; Weisskoff, R. M., J. L. Boxerman, A. G. Sorensen, S. during primary sensory stimulation. Proc. Natl. Acad. Sci. USA M. Kulke, T. A. Campbell and B. R. Rosen. 1994. Simultaneous 89:5675–5679; Logothetis, N. K. et al. 2001. Neurophysiological blood volume and permeability mapping using a single Gd-based investigation of the basis of the fMRI signal. Nature 412:150–157; contrast injection. Proceedings of the Second International Ogawa, S. et al. 1990. Brain magnetic resonance imaging with Society for Magnetic Resonance in Medicine, San Francisco, CA, contrast dependent on blood oxygenation. Proc. Natl. Acad. Sci. p. 279. USA 87:9868–9872. Blood oxygenation level–dependent (BOLD) contrast Blue light hazard (Magnetic Resonance) The brain requires a steady flow of oxygen (Non-Ionising Radiation) This term is used to indicate the pho- to function, which is transported by haemoglobin in the blood. tochemical hazard that visible light, in the 300–780nm region, In functional MRI, blood oxygenation level–dependent (BOLD) but primarily the blue region (400–450 nm) can cause to the eye. contrast is the most commonly used mechanism for detecting The effect is not cumulative and depends on how the opti- brain activity. cal energy enters the eye, and in particular, both the energy and Whilst it has been shown that a BOLD response directly dimension of the light beam. The latter depends on the actual size reflects an increase in neural activity (see Logothetis et al., 2001), and distance of the artificial optical source and is described in the understanding of its neural origins and evolution remains terms of the solid angle of exposure. unclear. At a basic level, the process can be traced as shown in The hazard is due to the particularly high sensitivity of the Figure B.63: retina’s photoreceptors (cones and rods) at those wavelengths. To measure this hazard, ICNIRP refers to a blue light action spec- 1. An fMRI task or stimulus causes an increase in neuro- trum (see Action spectra). nal activity in brain areas related to the task. Related Articles: AORD, Action spectra, Cornea, Eye, 2. The increase in neuronal activity induces complex neu- ICNIRP, Lens, Skin, Skin cancer rovascular coupling. Further Readings: Coleman, A., F. Fedele, M. Khazova, 3. The neurovascular coupling activates a haemody- P. Freeman and R. Sarkany. 2010. A survey of the optical haz- namic response (see related article: Haemodynamic ards associated with hospital light sources with reference to the response function) such that blood flow to the active Control of Artificial Optical Radiation at Work Regulations 2010. brain region is increased and the local level of oxy- J. Radiol. Prot. 30(3):469; ICNIRP. A closer look at the thresh- haemoglobin becomes elevated relative to the level of olds of thermal damage: Workshop report by an ICNIRP task deoxyhaemoglobin. group. Health Phys. 111(3):300–306; ICNIRP. 2003. Guidelines 4. As oxyhaemoglobin is diamagnetic, and deoxyhaemo- on limits of exposure to incoherent visible and infrared radia- globin is paramagnetic, the increase in oxyhaemoglo- tion. Health Phys. 105(1):74–91; ICNIRP. 2013. Guidelines on bin prompts a local increase in T2*. Thus, the active limits of exposure to laser radiation of wavelengths between (1) Neuronal (3) Haemodynamic fMRI BOLD activity (2) Neurovascular response response (4) Detection by Stimulus coupling MRI scanner Or modulation in background activity FIGURE B.63 The basis of fMRI BOLD contrast. (Image courtesy of Arthurs and Boniface.) Blurring 114 Body coil 180 nm and 1,000 µm. Health Phys. 105(3): 271–295; ICNIRP. 2004. Guidelines on limits of exposure to ultraviolet radia- tion of wavelengths between 180 nm and 400 nm (incoherent optical radiation). Health Phys. 87(2):171–186; ICNIRP. 2000. Revision of the guidelines on limits of exposure to laser radia- B tion of wavelengths between 400 nm and 1.4 µm. Health Phys. 79(4):431–440; Sihota, Ramanjit and Radhika Tandon. 2011. Parsons' Diseases of the Eye, Elsevier India. Blurring (Diagnostic Radiology) See Detail resolution B-mode (Ultrasound) Brightness-mode (B-mode) imaging is the most common way to display diagnostic ultrasound measurements. Emitting short pulses and detecting the time of arrival and the amplitude of the echoes build the image. In B-mode imaging, the echo amplitudes are converted to gray levels where white dots correspond to high amplitudes. With a linear array transducer, hundreds of line measurements build up a cross-sectional image. These lines are updated continuously, allowing real-time imaging FIGURE B.65 B-mode imaging. with a frame rate on the order of 10–60 images/s (depending on penetration depth, number of lines, number of focus zones, etc.) (Figures B.64 and B.65). Body coil Related Articles: A-mode, M-mode, B-scanner (Magnetic Resonance) The body coil is an RF coil built into the MRI system. It surrounds the bore of the magnet and is hidden BMUS from view by the protective covers lining the bore. The body coil (Ultrasound) The British Medical Ultrasound Society is a may be used both as a transmit coil and a receive coil (Figures multidisciplinary society organisation with a membership of B.66 and B.67). Body coils are volume coil designs and display sonographers, physicians, technologists, scientists, physicists good uniformity over a large field of view. Where a large, uniform and engineers. The society’s stated aims are to advance the sci- field of view is required, the body coil will frequently be utilised ence and application of ultrasound in medicine and to provide as RF transmitter and receiver. For example, the body coil is par- education, guidance and information for its members and the ticularly suitable for general imaging of the thorax and abdomen. public. It achieves this through its annual meetings, journal and However, where a higher SNR is required and a smaller field website. of view is adequate, it will be generally more appropriate to take Hyperlink: BMUS: www .bmus .org advantage of the better SNR achievable using a surface coil or a smaller volume coil (e.g. a head coil) as receiver. In addition, the use of separate receiver coils can enable the implementation of parallel imaging techniques. The body coil is generally used as the transmit coil even with other coils being used as the receiver. A-mode B-mode M-mode FIGURE B.66 MRI system with the covers removed, showing the body FIGURE B.64 The principle of A-, B- and M-mode display. coil (innermost elements) and the surrounding gradient coils. Body contour 115 Body protein monitor B FIGURE B.67 Close-up of lower surface of the body coil shown in FIGURE B.68 Body contouring orbit. Figure B.66. Body habitus As a transmitter, the body coil has the advantage of delivering a (General) A term used in medicine to describe the physical char- uniform distribution of flip angles. The only exception to utilisa- acteristics of an individual in relation to body size. For example, tion of the body coil as a transmitter is where the imaging coil ‘poor body habitus’ in an ultrasound report would describe an used also has transmit capability (e.g. some head and extremity individual who was obese. coil designs). Body mass index (BMI) Body contour (General) This is a statistical measure of the weight of a person (Radiotherapy) In general, the body contour represents the outline scaled according to height. Body mass index is defined as the of the patient as indicated on an image to be used for treatment individual’s body weight divided by the square of their height. planning, for example, CT or MRI scan slices. Other, simpler, tra- ditional methods include the use of lead wire or mechanical rods Body protein monitor to form the shape of the patient surface. These are then traced on (Radiotherapy) paper before being digitised onto the planning system. Background: The body comprises distinct and measurable The body contour also indicates the shape of the skin surface body compartments. The status and rate of change of these com- onto which the treatment field is incident. Typically, this will pro- partments reflects the health of a person and the response of treat- duce an oblique angle of incidence and so some distortion of the ment for a specific disease. Whereas measurement of body weight dose distribution will result. is a basic and useful parameter, it can be a misleading measure of Abbreviations: CT = Computed tomography and MRI = response to treatment, which may increase oedema and fat while Magnetic resonance imaging. depleting protein. An important compartment is the total body protein (TBP), which is a measure of muscle and visceral mass. It is determined Body contouring orbit directly by measurement of the total body nitrogen (TBN). The (Nuclear Medicine) Body contouring orbit describes non-circular relationship between nitrogen and protein is detector motion during a SPECT acquisition. In a non-circular detector orbit, also known as ‘body con- TBP = 6.25* TBN touring’, the detector will keep the camera heads in close prox- imity to the surface of the body during each view (Figure B.68). TBN is regarded as the superior measure of protein status in dis- This provides superior spatial resolution when compared with eased subjects. a circular orbit for imaging irregular structures such as the Methods: In vivo methods are used to obtain TBN values abdomen. before and after therapy, so as to determine the effect of therapy. On some systems, the technologist must specify the non-cir- The measurement of normal values is required so as to com- cular orbit before starting the acquisition by placing the detectors pare patient values, according to a defined nitrogen index, NI = as close to the patient as possible at several angles. The camera’s TBNp/TBNn, where p designates the patient and n the normal val- computer will then determine the optimal orbit. ues for healthy subjects. Alternatively, modern systems are able to perform automatic The importance of this index relates to the impact on dis- body contouring whereby collision sensors on the detectors are ease prognosis, and has clinical significance in cancer and renal used to bring the heads in as close to the body surface as possible

disease. for each view. In vivo interrogation techniques are used. These can give a Further Reading: Bushberg, J. T., J. A. Siebert, E. M. measure of the whole compartment, or in some cases, the spatial Leidholdt Jr and J. M. Boone. The Essential Physics of Medical distribution of the compartment. Imaging, 2nd edn. Nitrogen is measured in a body protein monitor (BPM) using Related Articles: Circular orbit, Single photon emission com- a Cf252 or PuBe neutron source. The patient is moved over a puted tomography (SPECT) collimated neutron beam. The fast neutrons emitted by these BOLD contrast 116 Bolus radioactive sources are moderated in tissue and captured by nitro- If particle creation/destruction processes (e.g. photoelectric gen and hydrogen nuclei in the patient. The 11.4 MeV ground- absorption, or pair production) are absent, then the derivative of f state gamma ray emitted after nitrogen capture can be measured with time must be zero (reflecting that the sum probability for par- by NaI detectors, being of higher energy than all background ticle number within the phase space is conserved). Considering radiations. The hydrogen gamma ray (2.2 MeV) is very intense the multivariable chain rule and then re-including creation/ B and easily detected. The ratio of nitrogen to hydrogen yields elim- destruction processes under the umbrella term of collision events, inates the dependence of gamma ray attenuation on body habitus. the generalised form of the BTE is arrived at: Abbreviations: BPM = Body protein monitor, NI = Nitrogen index, TBP = Total body protein and TBN = Total body nitrogen. Related Articles: In vivo body composition, Total body nitro- ¶f dr dp ¶f + Ñ + Ñ = é ù r f p f ¶t dt dt ëê ¶ ûú (B.13) t gen, Total body potassium, Total body water, Total body fat collision Under the assumption that no self-interactions of the transported BOLD contrast particle fluences occur, i.e. interactions are with the medium that (Magnetic Resonance) See Blood oxygenation level–dependent they are transported through only, the Boltzmann transport equa- (BOLD) contrast tion is made linear. It is beyond the scope of a short article to show how Equation B.13 can be redefined under the terms and specific Boltzmann distribution interactions more commonly utilised for radiation transport in (Magnetic Resonance) Classically, when a strong external mag- radiotherapy (see the reference material for this), but the more netic field B0 is applied to nuclei with non-zero nuclear spin, applicable form for photon becomes these nuclei are forced to precess about an axis either parallel or antiparallel to the magnetic field. Spins precessing parallel to W ×ÑFg + sgFg = qgg + qg (B.14) B0 are in a state of lower energy than those precessing antipar- allel to it, such that the parallel spin state (lower energy state) Ω is the unit normal in the direction of interest, Φγ is the angular claims the greatest share of the spin population. The population photon fluence, σγ is the total photon cross-section, qγγ is the dif- difference varies with temperature according to the Boltzmann ferential cross-section for photon to photon (e.g. Compton) scat- distribution: tering and qγ is a source term for new photons entering the volume element from an external source. NAP æ dE ö = exp Conceptually, the first term on the left of Equation B.14 thus ç ÷ NP è k BT ø represents streaming of particles out of a volume, and the second attenuation, which also removes particles from the volume. The where first term on the right, on the other hand, represents scattering NAP is the number of spins in the antiparallel state into the volume, and the second is a new term allowing for addi- NP is the number of spins in the parallel state tional in-flux from an external photon source. The equation is thus T is the temperature δ one which balances the number of photons, thus linking back to E is the energy difference between the two states Equation B.12. kB is the Boltzmann constant With the cross sections in Equation B.14 explicitly written out, it is seen to be a complex integro-differential equation, solvable When a patient (body temperature ∼310 K) is placed in a 1.5 only with approximated numerical methods. T scanner, the population difference between the two levels is Related Articles: Linear Boltzmann transport equation approximately 1 in 106. Such a population difference means that (LBTE) solver the vector sum of spins will be non-zero and will point parallel to Further Readings: Akpochafor, M. O. et al. 2014. Simulation the magnetic field. In other words, in the presence of a magnetic of the linear Boltzmann transport equation in modelling of pho- field B0, tissue becomes magnetised with net magnetisation M0. ton beam data. IOSR J. Appl. Phys. 5(6):72–86; Boman, E. 2007. This induced magnetisation M0 is the source of signal in all MR Radiotherapy Forward and Inverse Problem Applying Boltzmann experiments. Transport Equation, Department of Physics, University of Kuopio, Kuopio. Boltzmann transport equation (Radiotherapy) The Boltzmann transport equation (BTE) describes the statistical evolution of a system of particles with Bolus time, through consideration of an appropriate probability dis- (Radiotherapy) Bolus is a tissue equivalent material placed tribution for their positions and momenta. Although it was ini- directly on the skin surface to even out the irregular patient con- tially devised by Ludwig Boltzmann to describe the diffusion of tour and thereby provide a flat surface for normal beam incidence. gases within thermodynamic theory, it has been adapted to apply In principle, the use of bolus is straightforward and practical; equally well to radiation transfer problems. however, it suffers a serious drawback: for megavoltage photon Consider a particle to have position r and momentum p within beams, it results in the loss of the skin sparing effect in the skin a (six-dimensional) phase space encompassing all possible posi- under the bolus layer (i.e. skin sparing occurs in the bolus). tions and momenta of that particle. At a given time, the number of Bolus is placed in contact with the skin to achieve one or both particles occupying a certain finite element of the phase space can of the following: be found by integrating over volume elements as 1. Increase the surface dose: To increase the surface dose, N = ò d3r ò d3 p f (r, p,t ) (B.12) a layer of uniform thickness bolus is often used (0.5–1.5 cm), since it does not significantly change the shape of where f is a probability density function. the isodose curves at depth. Several materials have been Bolus injection 117 Bone developed commercially for this purpose, which are agents (for computerised tomography or magnetic resonance floppy and malleable and easily shaped to the patient’s imaging). surface. However, wet towels or gauze wrapped in cel- Related Articles: Dynamic susceptibility contrast MRI, Bolus lophane offer a low-cost substitute. tracking 2. Compensate for missing tissue: To compensate for missing tissue or a sloping surface, a custom-made Bolus tracking B bolus arrangement can be made that conforms to the (Magnetic Resonance) When performing contrast-enhanced patient’s skin on one side and yields a flat perpendicular MRA (see Contrast-enhanced angiography), a contrast media incidence to the beam (see Figure B.69). The result is an are injected intravenously, and images are acquired as the bolus isodose distribution that is identical to that produced on reaches the area of interest. The timing of the MRA acquisition a flat phantom; however, skin sparing is not maintained. is essential to obtain optimal image quality for the vascular type This can be overcome by retracting the bolus (taking (arterial or venous phase) and/or the vascular territory (large or divergence into account) as in Figure B.69b. small vessels). Therefore, the determination of the time delay between the injection of contrast agent and the arrival of this A common material used for this kind of bolus is wax, which agent bolus to the region of interest is important. There are two is essentially tissue equivalent and when heated is malleable and main methods to determine accurate timing of the data acquisi- can be fitted precisely to the patient’s contour. Bolus can also be tion from ‘tracking the contrast agent bolus’. used to compensate for lack of scatter, such as near the extremities The first method utilises a test bolus, i.e. a small amount of or the head during total body irradiation (TBI). Saline or rice bags contrast agent. The test bolus is injected, and a special bolus can be used as bolus in these treatments. imaging scan is started. This is a dynamic scan of usually one Related Articles: Compensating wedge, Tissue compensation, slice with good time resolution to track the bolus. As dynamic Compensating filters, Compensation, Compensator scans are acquired during the passage of the test bolus, reviewing the image set gives the optimal time point after contrast injec- Bolus injection tion to start the data acquisition in the subsequent CE-MRA scan. (Nuclear Medicine) When discussing a bolus injection in nuclear This method requires two injections, and a revision of images in medicine, one is referring to a prompt intravenous injection of between scans. Furthermore, it assumes that the timing of the a small volume of the radiopharmaceutical. An example of an bolus does not change, for example due to changes in the heart examination where it is very important that a bolus is injected is rate. first-pass radionuclide ventriculography. The second method uses a real-time view of the area of inter- est. A simplified version of the CE-MRA sequence is run dynam- Bolus of tracer ically, and the operator can interactively start the subsequent (Magnetic Resonance) A bolus of a tracer is the single admin- CE-MRA scan as the contrast is seen to reach the vessels to be istration into a blood vessel over relatively short time to achieve studied. This method requires knowledge of the profile order of a blood concentration at a relatively high level. In the theoreti- the subsequent CE-MRA scan – if the order is centric, the scan cal modelling of tracer kinetics, the concept of an ‘ideal’ bolus should be started as the contrast reaches the area of interest, while is introduced. The injection or input of such an ‘ideal bolus’ is it, in the case of linear order, should be started before the contrast assumed to be instantaneous, i.e. the dose distribution over time arrival, to obtain the k = 0 samples with correct timing. is described by the Kronecker delta function. Tracers employed Related Articles: Contrast-enhanced angiography, Contrast in medical imaging are often radiolabelled substances or contrast media Bone (General) Compensator Density 1000–2000 kg/m3 CT number 400–3000 HU Wax bolue Bone is a rigid organ that forms the endoskeleton of vertebrates. There are 206 bones in the adult human body. Bone has multiple functions, including to provide shape, movement and protection of the organs of the body, to allow sound transduction, to produce Patient Patient blood cells, to store minerals, growth factors and fat, to detoxify and to achieve acid–base balance. Bone consists predominantly of mineralised osseous tissue containing calcium hydroxyapa- tite, which provides bone with its rigidity and strength as well as (a) (b) relatively low density due to its pseudo-honeycomb internal struc- ture. There are two types of bone, including cortical (or compact) bone and trabecular (or cancellous) bone. Cortical bone makes FIGURE B.69 In (a) a wax bolus is placed on the skin, producing a flat radiation distribution. Skin sparing is lost with bolus. In (b) a compen- up 80% of bone by mass and provides a hard smooth outer layer. sator achieving the same dose distribution as in (a) is constructed and Trabecular bone forms a porous interior network of bone, provid- attached to the treatment unit. Due to the large air gap, skin sparing is ing its lightweight structure and a high surface area, containing maintained. bone marrow and blood vessels. Bone consists of osteoblast cells, Bone densitometry 118 Bonn Call for Action which build bone and produce hormones; osteocyte cells, which be processed to subtract out soft tissue attenuation leaving an form and maintain bone and function in calcium homeostasis; and image of the bones. Data from the bone images are used to cal- osteoclast cells, which resorb and remodel bone. culate bone density values (BMD), which are compared with the Bone is assessed in

medicine by various techniques, most expected normal values for the age of the patient (Figure B.71). commonly conventional radiography or computed tomography B (CT) to assess anatomical structure. Functional imaging can be Bone–soft tissue interface carried out, such as a nuclear medicine bone scan by administra- (Radiotherapy) There is an enhanced dose pattern at the interface tion of a radioisotope (technetium-99m) labelled phosphate ana- of bone and soft tissue for photon beams due to logue tracer, methylene diphosphonate (MDP). Osteoporosis is usually assessed by a DEXA (dual-energy x-ray absorptiometry) 1. Backscattered photons scan to calculate bone mineral density (BMD), but can also be 2. Backscattering of secondary electrons set in motion in determined by other techniques such as ultrasonic densitometry. the soft-tissue medium upstream of the bone Metastatic bone disease can be treated by some combination of 3. Backscattering of secondary electrons set in motion surgery, chemotherapy, radiotherapy and radionuclide therapy. within the bone Related Articles: Bone densitometry, Bone–soft tissue inter- face, Tc-99m-labelled bone imaging agents It has been shown (Das and Khan, 1989) that the main component of this enhanced dose is due to the secondary electron backscat- Bone densitometry tering at the interface. All dose enhancement is upstream of the (Diagnostic Radiology) Bone densitometry is the measurement of bone. For a number of photon energies studied, from cobalt 60 to bone density or bone mass associated with both the thickness and 24 MV, the range of the backscattered electrons upstream (beam structure of bones. A primary application is to diagnose, evaluate entry side) is limited to a few millimetres and the magnitude of and monitor osteoporosis, a common disease that increases the the dose enhancement effect (approximately 8%) is similar. On risk of bone fracture. the transmission side of the bone, the forward scattered electrons Bone density can be measured with several techniques includ- from bone make the situation more complicated and energy depen- ing quantitative ultrasound (QUS), quantitative computed tomog- dent. Khan (2003) has shown that for energies less than 10 MV, raphy (QCT) and single-energy x-ray absorptiometry (SEXA or there is an initial dose reduction at the interface relative to dose SXA). However, the most common method for clinical bone den- in a homogeneous soft-tissue volume. The dose then builds up to sitometry is dual-energy x-ray absorptiometry (DEXA or DXA) – a value slightly higher than that for the homogeneous case. For Figure B.70. higher energies, there is an enhanced dose at the interface due to DEXA scans a selected region of the patient's body with a electron fluence increase in bone resulting from pair production. small beam of x-rays (pencil shaped or cone shaped). The method See Figure B.72 for an illustration of the dose pattern. is most often used to estimate bone mineral density (BMD – g/ Further Readings: Das, I. and F. Khan. 1989. Backscatter dose cm2) of the lumbar spine and femoral neck. The beam applies perturbation at high atomic number interfaces in megavoltage pho- alternatively two different x-ray photon energy spectra (by using ton beams. Med. Phys. 16(3): 367–375; Khan, F. M. 2003. The two different kVp). The method uses two scanning x-ray beams Physics of Radiation Therapy, 3rd edn., Lippincott Williams & (e.g. 70kV and 140 kV), produced either by an x-ray tube with Wilkins, Baltimore, MD. a special x-ray generator (medium frequency generator with two alternative frequencies) or through moving special metal filtration Bonn Call for Action in front of the x-ray tube (e.g. Cerium filter). The accuracy of the (Radiation Protection) The Bonn Call for Action seeks to method is approximately 3% and the patient dose is of the order foster coordinated work to address issues arising in radiation of 1–3 μSv. protection in medicine. It was issued at an IAEA organised Because of the different x-ray attenuation characteristics international conference held in Bonn, Germany, in 2012, and for the two spectra in soft tissue and bone, the scan data can strengthened at the follow-up conference in Vienna, Austria, in 2017. The 2012 conference details were as follows: • Title: International Conference on Radiation Protection in Medicine: Setting the Scene for Next Decade • Date and venue: 3–7 December 2012, Bonn, Germany • Organised by the IAEA with 16 participating organisations • Hosted by the Government of Germany and co-spon- sored by the WHO • Over 500 participants from 77 countries Objective: To improve radiation protection in medicine in the next decade All stakeholders were encouraged to take part in: FIGURE B.70 Typical DEXA system (Hologic QDR4500) in a posi- • Identifying and highlighting 10 main actions essential tion for densitometry of the lumbar spine. (Image courtesy of Hologic, to strengthen radiation protection in medicine Bedford, MA.) • Specifying related sub-actions and activities Bonn Call for Action 119 Bonn Call for Action B FIGURE B.71 Part of a typical software interface used to calculate bone densitometry (sample from measurements and calculations in the region of the lumbar spine). Sub-sections and activities: • Enhance the implementation of the principle of justification • Enhance the implementation of the principle of optimi- sation of protection and safety Tissue • Strengthen manufacturers’ role in contributing to the overall safety regime Photons • Strengthen radiation protection education and training of health professionals 2 Electrons produced • Shape and promote a strategic research agenda for radi- in tissue 3 Bone ation protection in medicine Electrons produced • Increase availability of improved global information on 1 in bone medical and occupational exposures in medicine 1 See type in text • Improve prevention of medical radiation incidents and Tissue accidents • Strengthen radiation safety culture in health care • Foster an improved radiation benefit-risk-dialogue • Strengthen the implementation of safety requirements globally FIGURE B.72 An illustration of the scattering at a bone soft tissue Created as a result of ICRPM 2017, the Bonn Call for Action interface. Implementation Toolkit. Boost (brachytherapy) 120 Boron This is an online live platform offering existing and new Boost dose implementation resources for improving radiation protection in (Radiotherapy) Boost dose is defined as a supplemental dose of medicine with links to different initiatives radiation delivered to the tumour bed following a conservative Further Reading: Bonn Call For Action Platform, IAEA, surgery in the treatment of the breast. The boost dose can be Vienna, www .i aea .o rg /re sourc es /rp op /re sourc es /bo nn -ca ll -fo r - delivered either with an external radiation source and interstitial B act ion -p latfo rm; Bonn Call For Action, Supporting the radioactive implant. In case of boost with an external irradiation, Implementation of the Bonn Call for Action, IAEA, Vienna, www . a smaller treatment field is used. w ho .in t /ion izing _radi ation /medi cal _r adiat ion _e xposu re /ca ll -fo r - act ion /e n/ Born approximation Hyperlinks: International Atomic Energy Agency (IAEA) (Ultrasound) To evaluate the pressure scattered from a region official website, www .iaea .org of inhomogeneities, an integral equation needs to be solved. In this expression, the total pressure (i.e. both the incident and the Boost (brachytherapy) scattered pressure) is needed, but this is unknown, since one is (Radiotherapy, Brachytherapy) Brachytherapy is often given as a trying to calculate the scattered pressure. A common approxi- boost together with external beam radiotherapy. Two examples of mation is then to use the Born approximation, which is valid if radiotherapy regimes with brachytherapy boosts are given. the scattered pressure is much smaller than the incident pressure. Another way to put this is that the incident pressure is virtually unchanged as it passes through the volume in question, i.e. ptot ≈ EXAMPLE 1: LOCALISED HIGH- pi. The scattered pressure will be small if the compressibility and RISK PROSTATE CANCER density variations of the scattering objects are small compared to the surrounding medium. A consequence of this is also that mul- External beam radiotherapy is used to give 50 Gy in 25 tiple scattering (scattering of the scattered wave) can be ignored. fractions to the prostate with a ‘larger’ margin (confor- By employing the Born approximation, the scattered pressure mal beam technique or IMRT technique). HDR interstitial can be evaluated directly, and not by using, for instance, succes- brachytherapy is used to give a boost to the prostate with sive approximation. narrow margins, consisting of 2 fractions of 10 Gy each. The dose to organs at risk has to be kept below accepted Boron biological effect limits; urethra, rectum, etc. (General) Symbol B Element category Metalloid (semimetal) EXAMPLE 2: CANCER OF THE CERVIX, Mass number A 10, 11 (stable isotopes) RADICAL RADIOTHERAPY Atomic number Z 5 Atomic weight 10.811 g/mol External beam radiotherapy with concomitant chemother- Electronic configuration 1s2 2s2 2p1 apy (chemoradiation) is used to give 46.8 Gy in 26 fractions Melting point 2349 K to a larger target volume – pelvic lymph nodes including Boiling point 4200 K ‘uterus and tumour’, with an external beam boost of 3.6 Density near room temperature 2.34 g/cm3 Gy in 2 fractions to a smaller target volume – ‘uterus and tumour with margins’; in total, 50.4 Gy in 28 fractions is given to this volume. HDR intracavitary brachytherapy History: Boron was discovered jointly by Sir Humphry Davy, (ring applicator) is used to give a boost to the tumour with Joseph Louis Gay-Lussac and Louis Jacques Thénard in 1808. appropriate margins (see the GEC-ESTRO ref for a discus- Isotopes of Boron: Boron is found naturally only in compound sion of the brachytherapy target volume), consisting of 5 form, notably as sodium borate (borax) and hydrated sodium cal- fractions of 5 Gy each. The dose to organs at risk, rectum, cium borate hydroxide (ulexite). Two stable isotopes exist: 10B and bladder, sigmoid colon, must be considered. 11B (19.9% and 80.1% relative abundance, respectively). Medical Applications: Antiseptic – Diluted boric acid is sometimes used as a topical application for minor cuts, burns and Abbreviations: HDR = High dose rate and IMRT = Intensity- wounds due to its anti-bacterial and antiseptic action. modulated radiotherapy. Boron neutron capture therapy (BNCT) – An application cur- Related Articles: Intracavitary brachytherapy, Temporary rently under investigation for the treatment of aggressive brain implant, Interstitial brachytherapy tumours. The therapy is delivered in two parts and utilises the Further Readings: Kovács, G. et al. 2005. GEC/ESTRO- high affinity of 10B for capturing thermal neutrons. An intrave- EAU recommendations on temporary brachytherapy using step- nous injection is administered containing a 10B-labelled agent, ping sources for localised prostate cancer. Radiother. Oncol. which is selectively taken up at the tumour site. The region is then 74:137–148; Pötter, R. et al. 2006. Recommendations from gyn- irradiated with a beam of low-energy neutrons, which lose energy aecological (GYN) GEC ESTRO working group (II): Concepts in tissue to become thermal neutrons and are absorbed by 10B in and terms in 3D image-based treatment planning in cervix cancer the tumour to form lithium ions and alpha particles: brachytherapy – 3D dose-volume parameters and aspects of 3D image-based anatomy, radiation physics, radiobiology. Radiother. Oncol. 78:67–77. 10 B+1n ®11 B* ®7 Li+4He Boron neutron capture 121 Boundary layer Both 7Li and 4He cause ionisations in tissue with a path length which is more than two times that of 10B(n, α)7Li reaction. In the of 5–12 μm, approximately a single-cell diameter. This theoreti- BNCT, the heavy charged particles release 3.3 MeV only within cally provides a significant dose to the tumour, with substantial their total trajectory. Such limited energy transfer within short reduction in collateral damage to healthy tissue compared with trajectory in tissue can save serious radiation injury to the normal conventional radiotherapy techniques. However, the feasibility of brain surrounding the tumour. However, dose distribution in the this therapy has not currently been proven and is the subject of tumour is sharply dependent on the microdistribution of 10B in B ongoing trials. tumour that is heterogeneous and has non-uniform dose distribu- Neutron Shielding: Boron has a high thermal neutron absorp- tion, aside from a variety of proliferation conditions of the tumour tion cross section, favouring its use as a shielding material in cells. Dose distribution of Gd-based NCT is more uniform than high-energy radiotherapy linear accelerator units. Boron is added that of boron-based NCT and might be suitable for pathological to a liquid, usually paraffin, which is incorporated into the door heterogeneity

of malignant tumours. of the linear accelerator maze to prevent the escape of potentially Related Article: Linear energy transfer (LET) harmful neutrons. Related Article: Neutron capture therapy Boundary layer (Ultrasound) Boundary layer is a term used in fluid dynamics. Boron neutron capture When fluid flows over a surface, the fluid close to the surface (Radiotherapy) Neutron capture is a nuclear reaction in which a moves at a velocity slower than that in the free stream. The thick- neutron collides with an atomic nucleus and they merge to form a ness of the boundary layer is usually described as the distance heavier nucleus. The incident neutron is swallowed up, giving rise from the surface to a point where velocity is 99% of free stream to more or less stable nuclear excited states of the resulting iso- velocity (Figure B.73). tope that may lead to an immediate de-excitation by the emission In Figure B.73, the shear stress at the wall causes the fluid of a gamma photon as well as a delayed emission process when at the wall to be stationary. This in turn creates a drag on adja- one or more nuclear particle may be emitted. The usual notation cent fluid, slowing it. Further from the wall, the fluid flows at free for the reaction is (n,p), (n,n), (n,2n), (n,a) or (n,fission). The heavy stream velocity. The region of slower moving fluid is the bound- charged particles produced in the nuclear reaction, i.e. proton and ary layer. alpha particle, have a higher LET than gamma rays. The energy Boundary layers can be laminar or, at higher Reynolds num- of the heavy particle is therefore distributed throughout a lim- bers, turbulent. In turbulent boundary layers, flow is characterised ited volume of approximately a sphere of 14 micron diameter, i.e. by unsteady flow with vortices. In tubes, laminar flows are present slightly larger than the diameter of a red blood cell. Boron neutron at Reynolds numbers <2000 and turbulent >4000. Between these capture therapy (BNCT) is a modality of radiotherapeutic treat- values, transition occurs. ment of brain tumours that utilises a neutron beam that interacts The presence and extent of boundary layers in arteries and with boron-10 (10B) injected to a patient. 10B, which is a stable veins lead to complex velocity distributions, which are evident in isotope, absorbs a thermal neutron, producing the nuclear reac- the Doppler sonograms. One practical consideration in Doppler tion 10B(n, a) 7Li. Alpha particles and lithium ions have a com- ultrasound quality assurance is that the flow profiles in a tube bined path length of approximately 12 μm and therefore deposit change as flow develops along it. This is shown diagrammatically most of their energy within the cell containing the original 10B in Figure B.74 where steady flow in a constant diameter straight atom. If a high concentration of 10B is obtained in tumour cells, a tube only develops a parabolic flow profile when the boundary high dose could be delivered to tumour cells, leaving normal cells layer extends across the whole diameter. This is important if flow unaffected. The interval between administration of the capture phantoms are to be used to produce a known parabolic flow pro- agent and neutron irradiation could be optimised to have the high- file where peak velocity = 2 × mean velocity. The entry length for est differential 10B concentrations between normal tissues and the flow to become established is approximately 120 × diameter for tumour. Alpha particles and lithium ions will produce a signifi- laminar flow. cant radiobiological effect because of their high linear energy As flow moves down the pipe, the boundary layer (grey) transfer (LET) as the particles experience closely spaced events. increases until it extends across the pipe. At this point, flow is Alpha particles have another biological advantage as they do not require oxygen to enhance their biological effectiveness. This makes them very effective in cases where the tumour has limited oxygen supply. The neutrons are created either in a nuclear reac- tor or in particle accelerators, making a proton beam collide into targets made of lithium or beryllium. The neutrons pass through V a moderator, which makes the neutron energy spectrum suitable max for BNCT treatment, and a collimation device, which shapes the beam before entering the patient. Passing through the patient tis- sue, the neutrons are slowed by elastic scattering and become low- energy thermal neutrons. The thermal neutrons undergo reaction Boundary layer with the boron-10 nuclei, forming an excited compound nucleus boron-11 (11B), which then promptly disintegrates to an alpha particle and 7Li. The investigation of carriers more efficacious than boron has recently addressed the use of 157Gd. 157Gd has a large thermal neutron cross section of 255,000 barn, which is 65 Wall times that of 10B and releases Auger electrons, internal conversion electrons, gamma and x-rays by a single thermal neutron capture reaction sharing among them the total kinetic energy of 7.7 MeV, FIGURE B.73 Flow profile of fluid flowing over a surface (wall). Bow-tie filter 122 Brachytherapy Boxcar function (Magnetic Resonance) Boxcar function or boxcar design is also known as block design. For further details, see Block design (MRI). B Brachytherapy (Radiotherapy, Brachytherapy) Depending on the distance Boundary layer between the radiation source and the target volume, i.e. the tis- sues to be treated, radiotherapy is divided into two categories: FIGURE B.74 If flow starts in a pipe from a container, it will initially teletherapy and brachytherapy: have blunt flow profile (left). • In teletherapy, the source is far from the target (Greek word ‘tele’, meaning distant, far away). said to be fully developed and for constant laminar flow, has a • In brachytherapy, the source is placed close to or inside parabolic flow profile. the target (Greek word ‘brachys’, meaning short). Related Articles: Reynolds number, Laminar flow, Turbulent flow Brachytherapy conventionally uses sealed radioactive sources. Vocabulary in Brachytherapy: There are many different Bow-tie filter ways of characterising brachytherapy (also called curietherapy), (Diagnostic Radiology) The bow-tie filter, also referred to as using words both from Latin and from Greek: ‘beam shaping’ filter, or sometimes ‘wedge’, is a physical filter used on CT scanners to modify the x-ray beam profile in the scan • In intracavitary brachytherapy, applicators/sources are (x–y) plane. It is usually employed in addition to the flat filter used placed in existing cavities (Latin words intra, meaning to remove low-energy x-rays from the beam (Figure B.75). within, inside, and cavitus, meaning cavity, hole). The bow-tie filter compensates for the shape of the patient • In interstitial brachytherapy, applicators/sources are cross section by having a reduced thickness on the central axis placed within the tumour using needles or catheters, of the beam where patient attenuation is highest. It provides addi- ‘you make the cavities yourself’ (Latin words inter, tional beam hardening where the path of the beam through the meaning between, and interstitium, meaning gap, ‘a patient is short, and so results in a more uniform beam energy at space between’). the detectors. The uniformity of dose and image quality across • In endobronchial brachytherapy, applicators/sources the patient cross section is also improved. are placed in the bronchus, a type of intracavitary tech- Some models of CT scanners will have two or more differ- nique (Greek word endon, meaning within). ent bow-tie filters, designed for head and body applications or for • In endoluminal brachytherapy, the applicators/sources patients of different sizes. More recently, some CT scanner mod- are placed within (Greek, ‘endon’) the lumen (Latin, els incorporate a special bow-tie filter for cardiac CT scans. These meaning passage within a tubular organ); also a type of are shaped so as to reduce the dose, and thereby the image quality, intracavitary technique. Using Latin (intra) instead of in the area outside the central region of the beam as this is not of Greek (endo) for the word within, the term is intralumi- primary interest when imaging the heart. nary or intraluminal brachytherapy. Related Article: Beam hardening • Then there is endocurietherapy, plesiocurietherapy (Greek words ‘endon’, meaning within, and ‘plesios’, meaning near), intravascular brachytherapy. Brachytherapy Characteristics in General: Bow tie filter • Brachytherapy conventionally uses radioactive sources; thus, there is always a hazard of radiation exposure. More intense • Today, remotely controlled afterloading equipment is beam at used for high-dose-rate (and pulsed-dose-rate) brachy- centre therapy, and also for low-dose-rate brachytherapy, in principle eliminating the radiation hazard for the staff More attenuation at involved. centre by patient • Brachytherapy gives a high dose to a small target vol- ume, with a rapid fall off outside the target, an advan- tage of brachytherapy. The steep dose gradient could on the other hand also lead to target under dosage at the periphery. More uniform signal • Brachytherapy cannot be used to treat very large tar- at detectors get volumes. Further, brachytherapy can only be used where the target volume is accessible for application/ insertion of the appropriate applicators. • The dose in the target volume is inherently very inho- FIGURE B.75 Diagrammatic representation of a bow-tie filter on a CT mogeneous in brachytherapy, with regions with doses scanner. (Courtesy of ImPACT, UK, www .impactscan .org) larger than 50% and more of the peripheral dose. Thus, Brachytherapy sources 123 Brachytherapy sources large parts of the target receive doses markedly higher Brachytherapy, ESTRO, Brussels available at the ESTRO web than the prescribed dose. This is in contrast to the case site: www .estro .be (accessed 31 July 2012); Nag, S. ed. 1994. High in external beam therapy, where the dose to the target Dose Rate Brachytherapy, Futura Publishing Company, New is almost constant, equal to the prescribed dose, with York; Nag, S. ed. 1997. Principles and Practice of Brachytherapy, a variation about ±5%. For the same specified dose, Futura Publishing Company, New York; Thomadsen, B. R., M. J. brachytherapy gives large parts of the target volume a Rivard, and W. M. Butler. eds. 2005. Brachytherapy Physics, 2nd B much higher dose than external beam therapy. edn., Medical Physics Monograph No 31, American Association • Another noticeable difference between external beam of Physicists in Medicine, Madison, WI. radiotherapy and brachytherapy is the impact of patient motion, both external motions of the patient and motion Brachytherapy sources of internal organs. This is a big problem in external (Radiotherapy, Brachytherapy) beam therapy, and elaborate techniques are developed Introduction: When brachytherapy was introduced, the to handle different types of patient motions. In brachy- naturally occurring radioactive isotopes radium-226 and radon- therapy, the problem of these movements is minimised 222 were used as photon-emitting brachytherapy sources. Later, when the applicators/sources are placed directly in particle accelerators and nuclear reactors made it possible to pro- the target volume! However, permanently implanted duce new radionuclides with a wide range of physical properties. sources can migrate within the target volume. Today, there are a number of radionuclides used in brachytherapy, • In modern high-dose-rate and pulsed-dose-rate brachy- emitting photon, beta and neutron radiations therapy, it is possible to manipulate the dose distribu- The historical radium and radon sources are in principle not tion to match the target volume. (‘Brachytherapy is the used any longer, and should not be used, primarily because of ultimate form of conformal radiotherapy’!) safety concerns. However, there is a large amount of clinical • Brachytherapy uses image-guided techniques today, experience gained using radium sources and specific treatment image-guided brachytherapy – IGBT, using ultrasound, techniques, which must not be forgotten. CT, MR, fluoroscopy, etc. Some photon-emitting nuclides are denoted ‘radium substi- • Brachytherapy is a radio-surgical procedure with spe- tutes’, for example caesium-137, iridium-192 and cobalt-60, as the cial requirements for the radiation physics part of the dose rate distributions for these sources and for radium sources treatment. are very similar up to about 5 cm distance from the source. • Brachytherapy requires proper education and training Originally, source strengths were specified in terms of ‘mg Ra eq’ for the whole brachytherapy team. (mg radium equivalent). Source Characteristics: Brachytherapy sources are in general Different Ways to Characterise Brachytherapy: Brachytherapy encapsulated; the radioactive material is contained in a capsule can be divided into several different types as follows: often made of stainless steel or titanium (non-toxic materials). The capsule further serves as an absorber for ‘unwanted decay prod- 1. According to the placement of the sources ucts’, such as alpha particles

and for photon sources also beta radia- a. Intracavitary techniques tion. Modern sources are small, and generally cylindrical in shape. b. Interstitial techniques Photon-emitting sources are by far the most commonly used c. Surface applications sources for all types of brachytherapy. Beta-emitting sources are 2. According to dose rate used for the treatment of superficial tumours/lesions, for example a. Low dose rate – LDR in ophthalmic applicators (and in intracoronary brachytherapy). b. Medium dose rate – MDR Properties of some brachytherapy sources are given as fol- c. High dose rate – HDR lows; the energy values – nominal energy values, principal or d. Pulsed dose rate – PDR mean values – are given for a typical encapsulated source, half 3. According to duration of treatment value layers are given for lead and the mass values in the last col- a. Temporary implants umn are theoretical calculated maximum values. b. Permanent implants 4. According to the handling of the sources a. Manual handling HVL Mass for b. Manual afterloading Energy (mm 100 MBq c. Remotely controlled afterloading Radionuclide Type Half-Life (MeV) Pb) (μg) 5. According to radiation quality a. Sources emitting mainly photons Ra-226, Photon 1600 years 0.83 16 2732 b. Sources emitting beta radiation historical c. Sources emitting neutron radiation Rn-222, Photon 3.82 days 0.83 16 0.02 historical Co-60 Photon 5.27 years 1.25 11 2.39 These types are presented in some more detail under related articles. Cs-137 Photon 30.07 years 0.662 6.5 31.1 Related Articles: Intracavitary brachytherapy, Intraluminary Ir-192 Photon 73.83 days 0.380 3.0 0.29 brachytherapy, Intravascular brachytherapy, Interstitial brachyther- I-125 Photon 59.4 days 0.028 0.025 0.15 apy, Dose rates in brachytherapy, Temporary implant, Permanent Pd-103 Photon 17.0 days 0.021 0.013 0.04 implant, Source loading in brachytherapy, Brachytherapy sources Sr-90/Y-90 Beta 28.8 years 0.55–2.28 0.14 19.6 Further Readings: Gerbaulet, A., R. Pötter, J.-J. Mazeron (Cf-252) Neutron 2.65 years 2.15 — and E. van Limbergen. eds. The GEC ESTRO Handbook of Bragg peak 124 Bragg peak spreading Some characteristics to consider in the choice of a photon-emit- where ting brachytherapy source for a clinical application are k0 is the 8.99 × 109 N m2/C2 z is the atomic number of the heavy particle 1. Energy e is the magnitude of the electron charge a. ‘Low’; minimises radiation protection requirements n is the number of electrons per unit volume in the medium B b. ‘Reasonably high’ (radium substitute sources) m is the electron rest mass 2. Half-life c is the speed of light in vacuum a. Longer; temporary implant sources – long working β = v/c is the speed of the particle relative to c life I is the mean excitation energy of the medium b. Shorter; sources for permanent implants 3. Specific source strength At high energies when β ≅ 1, the logarithmic term makes the a. High; small source dimensions and small applica- stopping power increase. At low energies, the factor in front of the tor dimensions bracket increases as β ≅ 0 and the logarithm term then decreases, 4. Toxicity (problems with radium) causing the so-called Bragg peak. The linear rate of energy loss 5. Decay products of the charged particle is a maximum there. The shape of depth– a. No gaseous products (problem with radium–radon) dose distribution of the heavy charged particle makes possible the b. No electrons that are not absorbed in the irradiation of a strictly localised region within a biological object encapsulation in correspondence of the Bragg peak depth. In Figure B.76, the Bragg peak of a 62 MeV proton beam in water is shown. Brachytherapy using radionuclide sources like iridium-192 for The exact location of the Bragg peak can be measured with temporary implants and iodine-125 for permanent implants is a use of the Bragg peak chamber. The diameter of the chamber is well-established clinical procedure. large enough to measure the proton beam diameter. Developments: The brachytherapy sources mentioned in Related Articles: Stopping power, Deposition of dose the earlier table are not the only sources that have been used for brachytherapy; other sources are, for example tantalum-182, Bragg peak spreading ruthenium-106, americium-241, samarium-145 and ytterbium-169. (Radiotherapy) Since the energy loss per unit path length for pro- There are a large number of radionuclides, both naturally tons is inversely proportional to the square of their velocity, protons occurring and artificially produced, that could be used for differ- deliver their maximum radiation dose at the point in the medium at ent types of brachytherapy, and developments of new sources and which they stop (Khan and Gibbons, 2014). This gives rise to the applications are continuously ongoing. Bragg peak. The Bragg peak of a mono-energetic proton beam or Other interesting developments are the brachytherapy sources pristine Bragg peak is too narrow to cover the dimensions of most that do not use radionuclides for the production of photons, con- clinical tumours. It is, therefore, necessary to modulate pristine sisting of miniature x-ray sources. Bragg peaks in energy and range to create spread-out Bragg peak Related Articles: Brachytherapy, Radium, Radium substi- (SOBP) beams to cover the dimensions of the tumour. There are tute isotope, Iodine-125, Iridium-192, Intravascular irradiation, three main techniques to broaden a narrow mono-energetic proton Equivalent mass of radium beam and achieve the desirable dose coverage of the tumour: (1) passive scattering; (2) uniform scanning; and (3) active scanning (McGowan et al., 2013). Bragg peak In passive scattering, a range modulator wheel is used to (Radiotherapy) Heavy charged particles such as protons or modulate the range or depth of penetration and spread the beam heavier particles interacting with a medium show a characteris- energy. Two scattering foils are used to widen the lateral dimen- tic-shaped depth-dose distribution, if their nuclear interactions in sions of the beam. Patient-specific compensators and collimators the medium are negligible. If some of the particles will be subject to inelastic nuclear reaction before they have come to rest, lighter fragments with longer ranges than that of the primary ion are produced, depositing the dose observed beyond the Bragg peak. As the charged particle proceeds through a medium, its rate of energy loss or ionisation per unit path length increases with 5 decreasing particle velocity until it reaches a maximum in cor- respondence of the Bragg peak near the end of its path. The shape 4 of the depth-dose curve is determined primarily by the type of ion and its initial energy spectrum. The shape of the distribu- 3 tion is due to the fact that the heavy charged particle spends the first half of its initial kinetic energy along a pathlength x and the remaining half of the energy will be spent in a distance roughly 2 equal to x/3, increasing in this way the rate of energy deposition at the end of its track. Using relativistic quantum mechanics, the 1 stopping power of a uniform medium for a heavy charged par- ticle is given by 0 0 10 20 30 dE 4pk2z2e4 0 n é 2mc2b2 ù mm in water - = n 2 dx 2 êl - mc b2 b (1 b2) ú ë I - û FIGURE B.76 The Bragg peak (62 MeV protons). Ionisation Bragg–Gray cavity theory 125 Breakdown voltage can be used to shape the beam in the depth and in the lateral The theoretical approach to Bragg–Gray dosimetry requires directions, respectively. A high-energy proton beam is used to the cavity and therefore the radiation detector so small that, cover the distal end of the tumour and a superposition of decreas- when inserted into a medium, it does not disturb the fluence ing energy proton beams is used to cover the proximal end of the of charged particles existing in the medium. This means that tumour (Khan and Gibbons, 2014). the ideal Bragg–Gray cavity is one of infinitesimal dimensions Uniform scanning is similar to passive scattering. However, and therefore the detector must be a point detector to fulfil B instead of using scattering foils to spread the beam laterally, the hypothesis of the theory. In practice, such detectors do not it uses magnets to scan the beam laterally. As in passive scat- exist but many real detectors may, in a first approximation, be tering, patient-specific collimators and compensators are used treated as Bragg–Gray detectors to a high degree of accuracy. to shape the beam laterally and in the depth direction, respec- However, perturbation corrections are needed to account for tively (McGowan et al., 2013). the deviation of the signal from a practical detector from that In active scanning, the target is divided into a series of layers. of an ideal. Each layer requires a specific proton beam energy with its associ- Related Article: Stopping power ated pristine Bragg peak to reach it. Magnets are used to scan the layer in the x and y directions. Once the layer has been painted, Braking radiation the energy of the beam can be adjusted by either using the energy (Radiation Protection) Braking radiation is more usually called selection system or by placing material in the beam with a range bremsstrahlung. For more information, see eponymous article. shifter. The next layer is then painted and so on until the whole Related Articles: Bremsstrahlung, Characteristic x-rays target has been covered. Active scanning allows for the delivery of individual Bragg peaks within the target, generating an effec- Breakdown voltage tive spread-out Bragg peak without the need of patient-specific (General) Breakdown voltage of an insulator is the minimum compensators and collimators (McGowan et al., 2013). voltage that causes a portion of an insulator to become electri- The reader is referred to IPEM Report 75 for diagrams of pas- cally conductive. sive scattering, uniform scanning and active scanning delivery Breakdown voltage of a diode is the minimum reverse voltage systems. to make the diode conduct in reverse, as described later. Some Related Articles: Bragg peak, Beam modulation, Modulation devices (such as TRIACs) also have a forward breakdown voltage. wheel, Range modulation, Spread-out Bragg peak (SOBP) Figure B.77 shows the dependence of the current flow on the Further Readings: Horton, P. and D. Eaton. 2017. Design and voltage for a diode (p–n junction). This p–n junction characteris- Shielding of Radiotherapy Treatment Facilities, IPEM Report 75, tic is nonlinear. 2nd edn, IOP Publishing; Khan, F. M. and J. P. Gibbons. 2014. The current flowing across the p–n junction depends on the Khan’s the Physics of Radiation Therapy, 5th edn., Wolters direction of the potential difference (bias). The current increases Kluwer Health; McGowan, S. E., N. G. Burnet and A. J. Lomax. with the voltage increasing in a forward direction (OF). If the nor- 2013. Treatment planning optimisation in proton therapy. Br. J. mal reverse bias is applied to the diode, the low reverse current Radiol. 86(1021):20120288. doi:10.1259/bjr.20120288. (OR) flows due to minority carriers. The diode can be used as rectifier. The further increase of the reverse voltage makes the Bragg–Gray cavity theory diode conduct in reverse (RB). The diode stops to act as a rectifier. (Radiotherapy) The Bragg–Gray theory considers a medium uni- The minimum reverse voltage causing that the reverse current formly irradiated by photons and whose dimensions are adequate increases strongly is called breakdown voltage. At this voltage, to establish a condition of electronic equilibrium in which a small the minority carriers are accelerated to get enough energy to ion- gas-filled cavity is inserted. The theory relates Dm, the absorbed ise molecules. As a result of this ionisation, the number of charge dose in the medium, to J, the charge per unit of mass resulting from ionisation produced by electrons in the gas. If the cavity gas is air, Dm is given by Current W S ö m = æ öæ D Jair ç è e ÷ç r ÷ F øè øm,air where Jair is the ionisation charge per unit mass of air in the cavity (S/ρ)m,air is the ratio of the mean unrestricted collision mass stopping power of material m to that of air W/e is the quotient of the average energy expended to produce an ion pair by the electronic charge O The theory requires that R Voltage • Charged particle equilibrium exists in the absence of the cavity • The cavity does not disturb the charged particle fluence or its distribution in energy and direction B • The mass stopping power ratio does not vary with energy • Secondary charged particles

lose energy by a process of continuous slowing down FIGURE B.77 Current as a function of voltage for the p–n junction. Breast 126 Bremsstrahlung carriers (electrons) increases and the reverse current grows. It is the so-called solid-state multiplication. Abbreviation: TRIAC = Triode for alternating current. Related Article: Semiconductor detector Further Reading: Graham, D. T. and P. Cloke. 2003. B Principles of Radiological Physics, 4th edn., Elsevier Science Limited, Edinburgh, London, UK, p. 168. Breast (General) Density ∼1000 kg/m3 CT number −100 to 0 HU The breast is the upper ventral region of the torso containing mod- ified sweat (mammary) glands in females, which lactate (secrete milk) for breast-feeding. Estrogens in females promote breast development. Each breast has a nipple surrounded by the areola, which have sebaceous glands, and the mammary glands drain to the nipple via a lactiferous duct network. The main structure of the breast consists of connective tissue, adipose tissue and Cooper’s ligaments, and the pectoralis major muscle lies underneath the FIGURE B.78 Breast coil in position on the table of an MRI. breast. The lymphatic drainage of breasts predominantly travels through the ipsilateral axillary lymph nodes, which is pertinent to oncology since breast cancer is common and metastatic spread occurs via the lymphatic system. Breast tissue is assessed in medicine usually for oncological purposes, predominantly by mammography, as used in the breast screening programme in the United Kingdom. The breast may also be assessed by ultrasound, MRI and nuclear medicine sen- tinel node scans to locate lymph nodes for biopsy. Breast cancer is treated by some combination of surgery, chemotherapy and radiotherapy. Related Articles: Digital mammography, Mammography, Radiotherapy, Ultrasound Breast coil (Magnetic Resonance) A breast coil is used in MRI as the RF coil for breast imaging. During imaging, the patient lies prone on the MRI table. Recesses in the breast coil housing accept each breast and adjustable paddles may be used to hold the breasts in position (Figures B.78 and B.79). FIGURE B.79 Close-up of a breast coil showing recesses and paddles used to hold the breasts. Breast density (Diagnostic Radiology) Breast density is defined as the amount of fibroglandular tissue in the breast compared with the total amount and consistent density measurements. It is a strong predictor of of breast tissue. The white areas of a mammogram result from the the risk of developing breast cancer, and women with the dens- high attenuation of x-rays in denser tissue (fibroglandular tissue est breasts are estimated to be 4–6 times more likely to develop consisting of connective tissue and epithelium). The darker areas cancer compared with those with the lowest densities. Some of the mammogram correspond to the less attenuating adipose software tools to assess volumetric breast density are Volpara (fatty) tissue. There are three ways of describing breast density as (Volpara Health Technologies), Quantra (Hologic) and Insight BD seen on a mammogram: pattern-based, area-based and volumet- (Siemens Healthcare) ric-based estimated respectively by visual, semi-automated and Further Reading: Ng, K. H. and S. Lau. 2015. Vision 20/20: fully automated. In addition to x-ray mammography, magnetic Mammographic breast density and its clinical applications. Med. resonance imaging (MRI) and ultrasonography (US) are alterna- Phys. 42(12):7059–7077. tive imaging modalities to assess breast density. It is generally accepted that a woman’s breast density is a Breast phantom strong predictor of the failure of mammography to detect breast (Diagnostic Radiology) See Mammographic phantoms cancer and thus alternate modalities such as ultrasound might be considered. Furthermore, it is proven that breast density is becom- Bremsstrahlung ing an increasingly important clinical tool for screening and pre- (Radiation Protection) Bremsstrahlung is a German word that diction of risk; hence, there is an increasing need for accurate means ‘braking radiation’. Bremsstrahlung contamination 127 Bridge circuit 100 90 Bremsstrahlung 80 x-rays 70 Bremsstrahlung 60 contamination 50 B 40 30 20 + 10 – 0 Incident 0 10 20 30 40 50 electron Depth (mm) Path of electron FIGURE B.81 A depth–dose profile illustrating the contribution from the bremsstrahlung contamination. FIGURE B.80 Illustration of braking radiation (bremsstrahlung). Bridge circuit When the electrons are incident upon the anode target, they are (General) A bridge circuit is a term commonly used to describe a rapidly decelerated by interactions with the electric fields within specific architecture of four components. the atoms of the target. There are two mechanisms by which the The bridge rectifier is a commonly available unit formed electrons are decelerated (Figure B.80). from four diodes, and used to convert AC signals or power to DC Firstly, incident electrons may interact with the electric field (Figure B.82). of the nuclei. The path of the incident electron is changed by The impedance bridge is designed to provide an output sig- this interaction, implying that the electron is decelerated – i.e. nal, which is related to the ratio of impedances of the conductors braked, giving off energy in the form of x-ray photons. Because while being relatively insensitive to changes in the supply voltage there are any number of different paths by which the electrons or ambient temperature (Figure B.83). may traverse through the atoms of the material, the x-rays emit- When used with a sensitive current meter across the out- ted will have a spectrum of energies. This continuous spectrum put arms, it becomes a traditional method for determining the of x-rays is therefore known as bremsstrahlung, or braking unknown impedance of a component: one half is made of two radiation. identical impedances Z1, Z2, the third is calibrated variable Secondly, they may interact with the electric field of an inner bound atomic electron, knocking that electron out of its atomic orbit, and losing some or all of its kinetic energy in the process. If not all its energy is lost, the incident electron is deflected by AC in the collision. The vacancy in the inner orbit is filled by an elec- tron dropping down from an outer orbit. The electron must emit energy in the form of an x-ray photon. The energy of the photon is + equal to the energy gap between the orbits, and as such is charac- teristic of the atom of the material – hence these x-rays are known as characteristic x-rays. DC Related Articles: Braking radiation, Characteristic x-rays out – Bremsstrahlung contamination (Radiotherapy) The electron beams from a linac always have some x-ray photons present, and these are known as brems- FIGURE B.82 The bridge rectifier. strahlung contamination. They are produced in the head of the linac and result in a bremsstrahlung tail being present in electron percentage depth-dose plots. This contamination is typically of the order of 3% (Figure B.81). Knowledge of the bremsstrahlung contamination of an electron beam is also of importance for a specialist treatment known as total skin elec- tron therapy. Vin Abbreviation: PDD = Percentage depth dose. Related Article: Electron ranges Further Reading: Klevenhagen, S. C. 1985. Physics of Vout Electron Beam Therapy Medical Physics Handbooks 13, Adam Hilger Ltd, Bristol, UK. Brick (Radiation Protection) See Radiation shielding FIGURE B.83 The impedance bridge. PDD (%) Brightness 128 BTV impedance Z3 and the fourth is the unknown Z4. The variable Brightness induction effect impedance Z3 is merely adjusted to nullify any current flowing in the current metre, at which point the variable impedance MUST equal the impedance of the unknown component. It has the two advantages of being a highly sensitive circuit whilst being insen- B sitive to the amplitude of the signal applied to power the circuit or any parameter that affects the components similarly (e.g. tem- perature change). Related Articles: Diode, Rectifier, Four rectifier circuit Further Reading: Horowitz, P. and W. Hill. 2006. The Art of Electronics, Cambridge University Press, Cambridge, UK. Brightness Which circle appears to be brighter? (Diagnostic Radiology) Brightness is the characteristic of a light source or reflecting surface describing the visual response or per- FIGURE B.84 This is an illustration of the brightness induction effect. ception related to the amount of light emitted. Both circles have the same physical brightness but might appear to be It is related to the measurable physical quantity luminance, different because of the backgrounds. (Courtesy of Sprawls Foundation, which is the density of luminous intensity emitted in a specific www .sprawls .org) direction. The SI unit for luminance is the candela per square metre (cd/m2). B-scanner In diagnostic radiology, the brightness (or luminance) of (Ultrasound) Basic diagnostic ultrasound imaging (greyscale images (film on illuminators and digital displays) has an effect images) is performed with a B-scanner. Other common words on visual perception and must be adjusted to an appropriate level. for ultrasound imaging instrumentation are linear scanner, ultra- When the perceived power of light is considered, then lumi- sound imager and ultrasound diagnostic scanner. A simplified nous flux is used – measured in Lumen (lm) – an SI unit. block diagram is shown in Figure B.85. The pulser generates high-voltage transmission pulses of – two 1 lm = 1 lx × m2 = 1 cd × sr (candela. steradian) to three cycles and peak-to-peak amplitudes of 200–300 V. The transducer, which converts the electrical pulse to an acoustical pulse, contains an array of elements, in some transducers several Here lux (lx) is the SI unit of illuminance and luminous emittance. hundred. Groups of elements are excited to generate each acoustic transmission pulse. The electrical pulses applied to the different 1 lx = 1 lm/m2 elements are time delayed to perform transmission focusing and beam steering. Lux is used for light incident on a surface, while lumen is used for The received echo signals are time delayed in a similar way light emitted from a surface. (dynamic focusing) before summation to build up the line data Hyperlink: http://en .wikipedia .org /wiki /Lumen_(unit) stored in the image memory in the scan converter. The beam former, controlled by the computer, distributes appropriate time- delay patterns for both transmission and reception. Brightness control Amplification and processing of the received echo signals (Diagnostic Radiology) See Automatic brightness control (ABC) are performed in several steps. The TGC compensates for sound beam attenuation so that the echo amplitude from a target close to Brightness induction effect the transducer is of the same size as one from a longer distance. (Diagnostic Radiology) Brightness induction is the apparent dif- Demodulation is used to convert the pulse (RF signal) to a ‘spike’ ference in the brightness or intensity of an object when the back- and compression is used to modify the spike amplitudes from a ground is changed from light to dark as illustrated in Figure B.84. large dynamic range to a lower, which is more suitable for images and the display. The scan converter transforms the line data to Brightness stabilisation pixel values in the image memory. (Diagnostic Radiology) See Automatic brightness control (ABC) In modern scanners, almost all adjustments and processing are performed within the computer. The software can be updated with new options and improvements. Broad-beam geometry A B-scanner is often provided with other modes and options (Nuclear Medicine) Broad-beam geometry refers to situations such as M- and Doppler-modes. in which the radiation beam profile is larger than a few cm2. Related Articles: B-mode, Transmitter, Linear array trans- The opposite situation is referred to as narrow-beam geometry. ducer, Dynamic focusing, Scan converter, Beam former, TGC, Consider two different beam geometries striking a perpendicular Demodulation, Compression, M-mode, Doppler imaging modes surface. The dose along the central line in the two cases will differ since in the broad-beam situation, photons outside the central line can scatter inwards and contribute to the dose. The dose along a BSS (basic safety standards) central line is therefore higher in the broad-beam geometry com- (Radiation Protection) See Basic safety standards (BSS) pared to a narrow beam with equal intensity per unit area. Related Articles: Narrow beam, Absorber, Broad-beam BTV geometry (Radiotherapy) See Biological target volume (BTV) Bucky diaphragm 129 B ucky wall stand B Computer Monitor FIGURE B.85 Block diagram of B-scanner. (Courtesy of EMIT project, www .emerald2 .eu) Bucky diaphragm X-ray tube (Diagnostic Radiology) Bucky diaphragm is the classic name Grid for what is more commonly known as a grid used to selectively focal point absorb scattered radiation in x-ray imaging. It is named after Dr. Gustav Bucky from Germany who developed the first grid or ‘dia- Primary source phragm’ in 1913. A major advancement was made by Dr.

Hollis Potter in 1920, Scatter source who developed a method for moving the grid during the exposure to blur out the undesirable images of the grid lines. This was known as the Potter–Bucky diaphragm. The use of the names has been changed over the years. Now the ‘Bucky diaphragm’ is most commonly known as a grid and the ‘Potter–Bucky diaphragm’ is known as a ‘Bucky’ or a Bucky mechanism. The application of a Bucky diaphragm or grid (anti-scatter grid) is illustrated in Figure B.86. Primary penetration Scatter penetration The grid is composed of parallel metal strips, such as lead, that attenuate the radiation. They are separated by spaces or FIGURE B.86 Bucky diaphragm principle. (Courtesy of Sprawls gaps filled with a low-attenuating material, such as aluminium Foundation, www .sprawls .org) or fibre. In many grids, the strips (spaces) are aligned with or focused on a point in space, the grid focal point. The grid is positioned so that the primary x-ray beam is aligned with the from 70 to 180 cm), but in practice it is usable in a focal range (e.g. strips (and interspaces) and can pass through or penetrate the grid with focal distance of 100 cm is usable in a range from 90 to grid with relatively little attenuation. This occurs when the grid 110 cm) without causing noticeable image non-uniformity. focal point is located at the x-ray tube focal spot. Because the Related Articles: Grid, Grid ratio, Focused grid patient’s body, which is the source of the scattered radiation, is much closer than the x-ray tube to the grid, the scattered radiation is not aligned with the interspaces and much of it is Bucky table attenuated. (Diagnostic Radiology) A Bucky table is a radiographic table with The desirable characteristic of a grid is to attenuate as much a built-in ‘Bucky’ or mechanism for moving the grid during the of the scattered radiation as possible and let the primary beam exposure. pass through. Related Articles: Bucky diaphragm, Grid An important performance variable of a grid is the grid ratio (usually with values from 5:1 to 16:1), i.e. ratio between height Bucky wall stand (thickness) of the strips and width of the interspace. A focused (Diagnostic Radiology) A Bucky wall stand is a stand for the grid is also characterised by the focal distance (usually with values cassette holder or for the flat panel detector with a moving grid Transducer elements Amplifiers TGC Transmitter A/D-converters Time delay Timer Time delay Summation Scan converter D/A-converters Build-up 130 Build-up plates 100 15 MV photons 80 10 MV photons Farmer chamber B 60 6 MV photons 40 20 HVT 3 mm Al HVT 2 mm Cu 0 5 10 15 20 25 30 Build up cap Depth (cm) dmax for 15 MV photons ~ 3 cm FIGURE B.87 Central axis dose build-up and fall-off for different kV FIGURE B.88 Farmer chamber and Perspex build-up cap. and photon beam energies. mounted usually on the wall of a radiography room. It is most often used in chest radiography. Related Articles: Bucky diaphragm, Grid Build-up (Radiotherapy) At megavoltage energies, the scattered radiation is more in the forward direction and gives rise to less scattered radi- ation outside the edges of the beam. This is clear from a compari- son of the isodose charts for kilovoltage and megavoltage beams. The Compton scattering process causes the recoil electrons to be ejected more in the forward direction and with increasing kinetic energy. The range of the recoil electron at megavoltage energies is on the order of a few millimetres, and an electron dose build-up to dmax manifests itself below the surface of the irradiated tissue, see Figure B.87. Related Articles: Percentage depth dose, Build-up dose Further Reading: Walter, M. 2003. Textbook of Radiotherapy Radiation Physics, Therapy and Oncology, 6th edn., Churchill FIGURE B.89 Farmer chamber with build-up cap fitted. Livingstone, Edinburgh, UK. Build-up cap their range. The electron dose build-up thus reaches a maximum (Radiotherapy) If a dose chamber is to be used for exposure mea- at a depth determined by the range of the electrons and there- surements under electron equilibrium conditions in a beam of fore by the energy of the photon beam. In practice, the kerma is energy >0.5 MV, then an appropriate build-up cap must be used. not quite constant but falls, as the primary radiation undergoes Figures B.88 and B.89 show this. absorption and attenuation, with the result that the depth of the Related Articles: Build-up, Ionisation chamber dose maximum is less than the maximum range of the secondary Further Reading: Metcalfe, P., T. Kron and P. Hoban. 1997. electrons; dmax is further reduced because the recoil electrons do The Physics of Radiotherapy X-Rays from Linear Accelerators, not all travel in the direction of the primary radiation. Medical Physics Publishing, Madison, WI. Related Articles: Build-up, Kerma Further Reading: Walter, M. 2003. Textbook of Radiotherapy Build-up dose Radiation Physics, Therapy and Oncology, 6th edn., Churchill (Radiotherapy) This may be explained in simple terms as follows. Livingstone, Edinburgh, UK. Each successive thin layer of tissue produces its quota of recoil electrons, which in turn deposit their kinetic energy through Build-up plates several successive layers of tissue beyond their point of origin. (Radiotherapy) In total body irradiation (TBI) treatments, bet- Although the kinetic energy released in each layer (the kerma) ter dose homogeneity throughout the patient is usually achieved may be constant, the energy deposited in each layer (the absorbed using high-energy photon beams. However, the build-up effect of dose) will be determined by the total number of electrons passing the percentage depth dose for these beams means that superficial through the layer, and that number increases as each layer adds or shallow structures close to the skin may not receive the pre- its quota of recoil electrons to the electron flux from the preced- scribed dose. To get higher dose to these regions, build-up plates ing layers. This number only continues to increase until the new made from Perspex or similar material can be placed between the electrons released only replace those that have come to the end of patient and the radiation source, close to the patient. These plates, Percentage depth dose Build-up region 131 Bullseye image which are usually 1–2 cm thick, provide a source of forward-scat- of examples, Figure B.90 for an illustration of typical build-up tered electrons when irradiated with high-energy photon beams. region data for x-ray beams and Figure B.91 for electron beams. These electrons then increase the dose to shallow structures and While the absorbed dose build-up region is similar in appear- improve dose uniformity in the patient. ance for both megavoltage electron and photon beams, the physi- Further Reading: Van Dyk, J. 1999. The Modern Technology cal reason and size are different: of Radiation Oncology: A Compendium for Medical Physicists B and Radiation Oncologists, Medical Physics Publishing, 1. In the case of x-ray beams, the initial electron fluence is Madison, WI. low, but as an increasing number of secondary electrons are produced, with depth the dose increases. There is Build-up region then a point (depth of maximum dose, dmax) where the (Radiotherapy) One of the benefits of using high-energy mega- fluence of these electrons is matched (in equilibrium) voltage beams instead of kilovoltage beams is that there is a skin with the exponential decrease in x-ray photons, and the sparing effect due to the fact that the dose at the skin surface is dose delivered then starts to decrease with depth. much lower than the dose to tissue at shallow depths. The region 2. For electron beams, the deposition of energy (dose) on the depth-dose curve where the dose increases from the skin begins immediately at the surface, and therefore the surface to the depth of maximum dose is called the build-up skin-sparing effect is reduced compared to x-ray photon region. The build-up region results from the range of secondary beams. As the electrons penetrate further in the mate- electrons produced in the patient/phantom by photon interactions rial, they are scattered at increasingly oblique angles, – that is they deposit their kinetic energy beyond the position at therefore increasing the fluence and thus dose delivered. which they are released. The amount of material contained within the build-up region The significance of the build-up region depends on the energy as well as the significance of the build-up effect will be dictated of the beam since it is much easier to scatter a low-energy elec- by the energy and nature of the beam – see Table B.4 for a list tron, and so a much more oblique pathway will result. Whereas higher energy electrons are less easily scattered and so tend to keep closer to a straight path, resulting in a much reduced build- up effect. TABLE B.4 Related Articles: Build-up, Build-up dose, Skin sparing Typical Surface Dose Values and Build-Up Bullseye image Thicknesses for Different Treatment Beams (Nuclear Medicine) A polar map combines all areas of the myo- Treatment Beam Surface Dose (%) Build-Up Thickness (cm) cardium into a 2D image and is standard in the most commonly used quantification systems. The image is created by wrapping 6 MV x-ray 50 1.5 each short axis image around the previous, starting from the 15 MV x-ray 30 3.0 apex. This type of image usually displays the raw counts, areas 4 MeV electrons 78 0.8 of abnormality and reversibility by combining and comparing the 12 MeV electrons 89 2.4 circumferential profile of each short axis slice into a colour-coded 16 MeV electrons 95 3.2 image and a gender-specific normal database. The points of each circumferential profile are assigned a colour based on normalised FIGURE B.90 Typical build-up characteristics for x-ray beams. Grey – cobalt (1.25 MV), light blue – 4 MV, pink – 6 MV, dark blue – 10 MV, red – 15 MV. Bus 132 b-Value B FIGURE B.91 Typical build-up characteristics for electron beams. Pink – 4 MeV, green – 6 MeV, dark blue – 9 MeV, grey – 12 MeV, light blue – 16 MeV, red – 20 MeV. 13 Bus (General) The data bus in a computer system is the pathway 7 between the central processing unit (CPU) and a peripheral 1 14 18 device. Usually, contemporary buses have parallel data lines 8 2 19 6 12 (most often 32 and 64 bits) between the CPU and any peripheral 3 20 controller (card). The buses have also external ports for cables, 5 15 9 4 11 connecting the computer with a specific device. Each data bus has 17 its own speed of data transfer (MB/s), for example 10 PCI – Peripheral component interconnect bus – 32 bits; 130 16 MB/s AGP – Accelerated graphic port – 32 bits; at least 500 MB/s Distal Mid Basal Anterior 1 7 13 USB – Universal serial bus – 480 MB/s Anteroseptal 2 8 14 Inferoseptal 3 9 15 Hyperlink: www .pcmag .com /encyclopedia _term/ Inferior 4 10 16 Inferolateral 5 11 17 Anterolateral 6 12 18 b-Value Anteroapical 19 (Magnetic Resonance) The b-value is a factor describing the Inferoapical 20 amount of weighting in diffusion-weighted MRI. In particular, it describes the strength of the diffusion encoding, and is the sum- FIGURE B.92 Diagrammatic representation of segmental division of marised influence of all gradients on the diffusion encoding. It the SPECT slices and the assignment of individual segments to individual describes the relationship between the diffusion coefficient, D, coronary arteries. and the signal attenuation. The higher the b-value is the stronger the diffusion encoding will be. For the Stejskal–Tanner experiment (Stejskal and Tanner, count values, and the colour profiles are shaped into concentric 1965), the b-value is defined by rings as shown in Figure B.92. The slice-based polar map is either distance-weighted to adequately localise defects or volume-weighted to equally map b = g2G2d2 (D - d/3) éës/mm2 ùû areas of the myocardium to determine as accurately as possible the size of defect, though this defect size is distorted. Although where this approach to perfusion quantification was a preferred option γ in [Hz/T] is the gyromagnetic ratio when nuclear medicine computers were slow, more realistic 4D G in [T/m] is the amplitude of the diffusion encoding gradient gated SPECT displays are growing in popularity. Despite this, δ in [ms] is the duration of the diffusion encoding gradient polar

map displays are still frequently used. Δ in [ms] is the time interval between the onset of the two Related Articles: Extent, Severity, Gated SPECT gradient pulses Bystander effects 133 Bystander effects 90° 180° (more radiosensitive) or stimulatory (cells tend to be less radio- sensitive) (Figure B.94). G The first report of this phenomenon involved the exposure of a t monolayer of cells to very low fluences of α-particles such that no δ more than 1% of cells were actually traversed by α-particles. The Δ observation that changes were seen in 30%–50% of cells showed B that the target for genetic damage by α-particles is much larger than the nucleus or indeed the cell itself. FIGURE B.93 A basic diffusion-sensitive pulse sequence, known as the Stejskal–Tanner pulse. There are also reports that chromosomal activity may be inherited by the clonal descendants of bystander cells. The bystander effect is becoming better understood, and See Figure B.93. there is evidence that it can be attributed to inter-cell interac- The b-value is often determined only from the applied diffu- tions and the release of factors into the media supporting the sion encoding gradients, but in order to accurately estimate the cell cultures. diffusion sensitivity, all gradients within a pulse sequence (imag- As can be seen in the previous diagram, three cells are irradi- ing and slice selection gradient) should be included. To correctly ated. The middle cell exhibits the classical direct damage leading determine the b-value of a pulse sequence, the time integral given to a mutation. The cell on the left only expresses the damage in later, including all gradient pulses, should be solved the form of genomic instability observed several generations later. The cell on the right is irradiated, but passes on the damage to another cell by some form of chemical messenger. It is then an un- TE æ t 2 ö irradiated cell that expresses the damage. Various theories have b = g2 ò ç òG¢(t )dt ÷ dt been postulated why the irradiated cell might do this; the most ç ÷ 0 è 0 ø obvious is as a survival mechanism. However, there are two key implications of the bystander where G′ effect: Firstly, the target volume for damage in tissue will be is the amplitude of the gradient τ greater than the irradiated volume. Secondly, the response of the is the duration of the gradient (denoted δ for the diffusion body to small doses of radiation may be higher than the linear no- encoding gradient) threshold (LNT) model might suggest. This can be represented in the graph in Figure B.95. Related Articles: ADC, Diffusion encoding, Diffusion time The graph demonstrates that lower doses lead to a dispropor- Further Reading: Stejskal, E. O. and J. E. Tanner. 1965. Spin tionately higher rate of cancer induction than might be assumed diffusion measurements: Spin echoes in the presence of time- using the classical LNT model for radiation protection. dependent field gradient. J. Chem. Phys. 42:288–292. Related Articles: Radiobiological models, Linear no-thresh- old model, Adaptive responses and hormesis Bystander effects Further Readings: Lorimore, S. et al. 1998. Chromosomal (Radiation Protection) Radiation-induced bystander effects are instability in the descendants of unirradiated surviving cells after biological responses in non-irradiated cells, i.e. no energy depo- α-particle irradiation. Proc. Natl. Acad. Sci. USA 95:5730–5733; sition by the radiation in the cell. These cells are often adjacent Nagasawa, H. and J. B. Little. 1992. Induction of sister chroma- to irradiated cells, hence the name bystander effects. Bystander tid exchanges by extremely low doses of alpha particles. Cancer effects are believed to be communicated via cell-to-cell junc- Res. 52:6394–6396; Sgouros, G., S. J. Knox, M. C. Joiner, W. F. tions or by secretion or shedding of soluble factors. The bystander Morgan and A. I. Kassis. 2007. MIRD Continuing education: effect is proportional to the absorbed dose delivered to the tar- Bystander and low-dose-rate effects: Are these relevant to radio- geted cells. The effect in a bystander cell can either be inhibitory nuclide therapy? J. Nucl. Med. 48(10):1683–1691. Bystander effects Radiation-induced bystander effect Linear no-threshold Radiation-induced Abscopal and model genomic instabilities– clastogenic effects gene mutations, chromosomal aberrations, aneuploidy Dose FIGURE B.94 Examples of bystander effects. FIGURE B.95 Theoretical dose response assuming bystander effects. Probability of effect C CAD (computer-aided diagnosis) to create an image whose intensity is proportional to the transmis- (General) See Computer-aided diagnosis (CAD) sion of gamma rays through the tissue. If this is performed for each view acquired in emission, the transmission information can C Cadmium tungstate be incorporated into reconstruction calculations to correct for the (Diagnostic Radiology) Cadmium tungstate (CdWO4) is a phos- attenuation of gamma photons with increasing depth in the tissue. phor material used as a scintillator in some radiation detectors Scintillation Detectors: Caesium fluoride was used as a with photomultiplier tubes (especially in CT scanner detectors). scintillation detector material in early time-of-flight (TOF) PET Cadmium tungstate has conversion efficiency on the order of 40 imaging, as its high temporal resolution allowed improved locali- (while CsI has ∼45 and CaWO4 has ∼15). It emits light with peak sation of radioactivity within the body. It has since been replaced wavelength of 520 nm. with lutetium oxyorthosilicate (LSO) and lutetium–yttrium oxy- Related Articles: CaWO4, Caesium iodide orthosilicate (LYSO), both of which offer increased detection effi- Further Reading: Bushberg, J. T., J. A. Seibert, E. M. ciency and spatial resolution as well as high temporal resolution. Leidholdt and J. M. Boone. 2002. The Essential Physics of Caesium iodide doped with thallium (Cs(TI)) is a common Medical Imaging, 2nd edn., Lippincott Williams & Wilkins, phosphor material used in digital radiography detectors, fluoros- Philadelphia, PA. copy image intensifiers and gamma cameras. The crystal pro- duces flashes of visible light in response to incident gamma and Caesium x-ray photons to enable detection of the radiation. (General) Teletherapy: Radioactive 137CsCl was used in the past as a radiation source in caesium teletherapy units. The use of cae- sium for teletherapy was discontinued because, in comparison Symbol Cs to cobalt-60 (Co-60), caesium-137 (Cs-137) has a relatively low Element category Alkali metal specific activity (86.7 Ci/g) and relatively low energy (662 keV). Mass number A 55 Moreover, CsCl is a salt, highly dispersible in the atmosphere Atomic number Z 133 and soluble in water, making it very dangerous from the point of Atomic weight 10.811 g/mol view of source security and disposal, in comparison to cobalt-60 Electronic configuration 1s2 2s2 2p6 3s2 3p6 3d10 4s2 4p6 5s2 which is a metal. Because of these drawbacks, Cs-137 is no longer 4d10 5p6 6s1 used for radiotherapy; however, it is still used in blood irradiators, Melting point 301.59 K despite security concerns, because its long half-life of 30 years and relatively low gamma ray energy make it more practical than Boiling point 944 K Co-60 for this purpose. Density near room temperature 1.93 g/cm3 Related Articles: SPECT (single photon emission computed tomography), PET (positron emission tomography), Caesium fluo- History: Caesium was discovered by Robert Bunsen and ride, Caesium iodide, Caesium unit Gustav Kirchhoff in 1860, through spectroscopic investigation of mineral water, in which it was seen as characteristic blue lines. Caesium fluoride (CsF) Carl Setterberg first produced the metal through electrolysis of (Diagnostic Radiology) Caesium fluoride (CsF) belongs to the caesium chloride in 1882. chemical family of inorganic fluorides. The compound is usually Although the internationally recognised spelling is Caesium, found as a hygroscopic solid. The caesium ions form a cubic crys- the spelling Cesium is also commonly seen. tal system. Isotopes of Caesium: Caesium is most commonly found in At the beginning of the 1980s, CsF has been investigated as a compound as a chloride or nitrate. It occurs as stable 133Cs with possible scintillation detector for time-of-flight positron tomog- 100% natural abundance. There are 38 other known isotopes with raphy, but its temporal resolution was found to be unsuitable for atomic numbers between 112 and 151, all of which are unstable that purpose. and decay through radioactive processes. The main isotope of interest in medicine is 137Cs, which decays via pure beta emission Caesium iodide to 137mBa with a half-life of 30.23 years. The 137mBa created in this (Diagnostic Radiology) Caesium iodide (CsI) is used in medical decay is often in an excited state and emits a gamma photon to x-ray imaging in structured or columnar phosphor screens, for reach its ground state with a half-life of 2.55 min, so that a sample both flat panel detectors (indirect detection) and digital fluoros- of 137Cs emits both beta and gamma radiation. copy. Within x-ray imaging, the phosphor screen is used to absorb Medical Applications: Nuclear Medicine Transmission the incident radiation (x-ray wavelength) and emit visible wave- Source and Marker – 137Cs is commonly used as a transmission length light photons, which are then detected through a number source in nuclear medicine scanning, to provide information for of optical imaging techniques. The wavelengths of the radiation the attenuation correction of SPECT and PET scans. The 137Cs absorbed and emitted by the phosphor is dependent on the mate- source is positioned behind the object being scanned, and the rial used in the phosphor screen, and the activator impurity intro- gamma emissions are detected from the opposite side of the object duced into it. 135 Caesium unit 136 Calcium A structured phosphor has superior spatial resolution to that Although CsI has an improved spatial resolution compared to of the traditional powdered phosphor of the same thickness, that of a powdered screen, it has several practical drawbacks as it as the columnar structure acts like fibre optic light guide. The is hydroscopic, toxic and very mechanically delicate. Due to these design of traditional powdered phosphor screens is restricted properties, it was first introduced as a phosphor for fluoroscopy by a compromise between x-ray detection efficiency and spa- since it could be protected within the image intensifier; however, tial resolution (Figure C.1). As the thickness of the phosphor recent advances in flat panel detectors have allowed its use in gen- screen is increased, the probability of an incident x-ray inter- eral radiography and mammography. acting with it increases. Related Article: Image intensifier At the same time, by increasing the thickness, the emitted Further Reading: Beutel, J., H. L. Kundel and R. L. Van C light photon must generally travel further from the point of Metter. 2000. Handbook of Medical Imaging: Volume 1, Physics emission before it exits the screen, increasing the spread of the and Psychophysics, SPIE, Bellingham, WA. produced light photons and thus reducing the spatial resolu- tion. In contrast, the CsI columnar structure is produced by an Caesium unit evaporation process, which grows the phosphor to form colum- (Radiotherapy) A Caesium unit is an obsolete machine in tele- nar, crystalline structures. These crystals form an array of nee- therapy that was used mainly in 1970s for palliative and non- dle-like elements, which collimate the light and improve the tumourous treatments with short SSD. spatial resolution. The spatial resolution is improved further Abbreviation: SSD = Source surface distance. by exposing the phosphor to a thermal shock which fractures Related Article: Caesium the crystals allowing air to enter within the cracks, reducing the density by 15%–25%. The difference between the refrac- Calcium tive indices of CsI and air (1.78 and 1, respectively) allows (General) the crystals to act like fibre optics, which guide the light pho- tons towards the horizontal edges of the phosphor with little lateral spread (assuming the photons are emitted at an angle Symbol Ca that allows total internal reflection). The spatial resolution of Element category Alkaline earth metal a columnar screen is therefore less dependent on the screen Mass number A 40 thickness than a powder screen. Atomic number Z 20 The k edges for CsI are at 33 and 36 keV, which gives a high Atomic weight 40.078 g/mol absorption probability for the x-ray energies ordinarily used in Electronic configuration 1s2 2s2 2p6 3s2 3p6 4s2 radiography. In planar radiography, the CsI is generally activated Melting point 1115 K by thallium (CsI:Tl) that emits a green wavelength photon (∼550 Boiling point 1757 K nm), which is effectively absorbed in amorphous silicon photodi- Density near room temperature 1.378 g/cm3 odes used in imaging detectors. In fluoroscopy, the CsI tends to be activated by sodium (CsI:Na) that emits blue wavelength light (∼450

nm), which is best matched to the response of photodiodes History: Calcium is known to have been used in oxide form used in x-ray image intensifiers (XRII). (lime) by the Ancient Romans. It was first isolated in its pure form Incoming x-ray Incoming x-ray Reflective layer Reflective layer Exiting light intensity (a) (b) FIGURE C.1 Lateral diffusion of light for (a) a traditional powder phosphor screen and (b) a columnar Caesium Iodide phosphor screen. Calculation of absorbed dose 137 Calibration depth by Sir Humphry Davy in 1808 using electrolysis of lime in mer- Related Articles: Treatment planning system, Convolution curic oxide. method, Hogstrom algorithm Isotopes of Calcium: There are four stable isotopes of cal- cium: 40Ca, 42Ca, 43Ca and 44Ca, and two metastable isotopes with Calibration mass numbers 46 and 48, which have half-lives sufficiently long (Radiation Protection) Calibration of a measuring instrument to be considered stable. 40Ca is the most abundant natural isotope, (device) consists of establishing the relationship between its with a relative abundance of 97%. There also exists a radioactive readings and the readings of another instrument known as the cosmogenic isotope, 41Ca, which is produced in the upper layers standard. of soil by reaction of 40Ca with neutrons and has a half-life of 103,000 years. Calibration curve (Medical Applications) X-ray Image Intensifiers: Calcium C (Radiation Protection) Calibration curve (Figure C.2) represents tungstate (CaWO4) is a compound of calcium, which exhibits the relationship between a measured signal (dependent variable, fluorescence under short-wave ultraviolet light. Large crystals of e.g. number of counts) and the corresponding value of an indepen- calcium tungstate were routinely used until the mid-1970s as fluo- dent physical variable, for example absorbed dose. rescent screens in real-time x-ray fluoroscopy, with a conversion The linear regression method is used for fitting straight lines. efficiency of 5% and emission wavelength of 420 nm. However, Related Articles: Calibration, Calibration depth, Calibration these have been superseded by rare earth metal screens, which factor have been found to be more suitable for the application. Further Readings: Knoll, G. F. 2000. Radiation Detection Wound Dressings: Calcium alginate in powder form is used in and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, modern wound dressings to absorb fluids secreted from wounds. pp. 629–630; Saha, G. P. 2001. Physics and Radiobiology of On contact with fluid, it becomes a thick, non-adherent gel that Nuclear Medicine, 2nd edn., Springer-Verlag, New York. provides a moist environment, which promotes tissue healing. Calcium Scoring in CT: Calcium is present in hydroxyapatite, Calibration depth which makes up 70% of bone. It is also important in the nervous (Radiation Protection) The personal dose equivalent HP(d) is the system and muscle contraction. Calcium is present in calcified dose equivalent in soft tissue, at an appropriate depth, d, below a plaques, which build up in the wall of the coronary artery and specified point on the body (unit: J/kg). It is used in monitoring of pose a major health concern. The disease can be diagnosed by individuals exposed to external irradiation. The value of a depth d measuring the amount of calcium present in the artery using a is equal to 10 mm for the skin for penetrating radiation like high- computed tomography (CT) scan, a process known as calcium energy x- and gamma rays or neutrons and to 0.07 mm for weakly scoring. penetrating radiation, for example x-rays and electrons. The depth Related Articles: Image intensifier, CT (computed of 0.07 mm corresponds to the thickness of an epidermis. A depth tomography) of 3 mm is used for the eye. The monitoring of the environmental doses is performed with Calculation of absorbed dose so-called spatial dose equivalent H*(10) defined as a dose equiva- (Radiotherapy) The dose can be measured either directly by a lent at the depth d = 10 mm of a sphere made from tissue-like calorimeter observing the heating effect produced by the inter- material and of 30 cm diameter, i.e. ICRU sphere (Figure C.3). actions of the radiation with matter or indirectly by ionisation Abbreviation: ICRU = International Commission on Radiation chamber measurements under so-called Bragg–Gray conditions Units and Measurements. or by chemical dosimetry (e.g. Fricke dosimeter). These detec- Related Articles: Calibration, Calibration curve, Calibration tors are not suitable for quantifying the dose and its distribution factor in patients and therefore algorithms have been developed for cal- Further Readings: Graham, D. T. and P. Cloke. 2003. culating the dose distribution throughout the three-dimensional Principles of Radiological Physics, 4th edn., Elsevier Science irradiated volume. The core of the calculation process is a treatment planning system (TPS) that schematically comprises a computer, input and output devices and appropriate software. The TPS permits the input of the patient anatomical information and produces a representation of the dose distribution in the patient with as much accuracy as possible. B There are many types of dose calculation algorithms, but basically they can be classified into correction-based and model- based algorithms. In correction-based approaches, the dose is first measured in a water phantom using broad irradiation field or reconstructed from a representative sample of these measure- ments. Then the dose distribution is corrected for the presence of beam modifiers, tissue heterogeneities and patient surface A contours. Model-based methods directly compute the dose to a patient, taking into account the beam energy and geometry, and Noise the anatomical description of the patient. The most common methods for dose calculation are, respectively, the convolution/ Dose superposition and the Monte Carlo methods for computing photon beam dose and the Hogstrom pencil beam and the Voxel Monte FIGURE C.2 Calibration curve: (A) limit of detection and (B) limit of Carlo methods for computing electron beam dose. linearity. Measured signal Calibration factor 138 Calorimeter The stability of any hospital ionisation chamber must be veri- fied as part of the quality control of the chambers (recalibrations of the hospital standard chambers are recommended every 2 years). For the long-term stability/constancy checks of a well-type chamber, a source with long half-life, for example Cs-137 and an d insert with stable geometry can be used. Alternatively, the stabil- ity checks of the well-type chamber can be made by irradiation in 15 cm a Co-60 beam under stable geometric conditions. Calibration Source – Quality Control of Source Strength for C ‘Permanent Seeds’: The whole quality control procedure for an interactive permanent seed implant must be performed before the start of the actual implantation of the seeds themselves, includ- FIGURE C.3 Example of measurement of an environmental dose equiv- ing the verification of source strength. Source strength is verified alent H*(10) on the depth d = 10 mm in the ICRU sphere (tissue-like mate- using a well-type chamber calibrated for the type of source used rial) of 30 cm diameter. and using the ‘dedicated insert’, specific for that type of source. In practice, a measurement of source strength is often made for a sample of at least 10% of the number of sources to be used for the Limited, Edinburgh, UK, pp. 347–348; ICRU (International implant (AAPM recommendation: 3% tolerance between hospital Commission on Radiation Units and Measurements). 1992. calibration and manufacturer source certificate for the mean of Measurement of dose equivalent from external photon and electron a batch of sources, and a maximum deviation of 5% for an indi- radiation. ICRU Report 42, Bethesda, MD; ICRU (International vidual source). Commission on Radiation Units and Measurements). 1993. Verification of stability of the well-type chamber for the spe- Quantities and units in radiation protection dosimetry. ICRU cific source type used can also be made using a single source, Report 51, Bethesda, MD. preferably with a smaller calibration uncertainty than the stan- dard sources, ‘a calibration source’. Calibration factor Abbreviations: AAPM = American Association of Physicists (Radiation Protection) The detector used for absorbed dose in Medicine and IAEA = International Atomic Energy Agency. measurement in electron and photon beams in external beam Related Article: Well-type ion chamber radiotherapy should be calibrated using primary standards for Further Readings: IAEA. March 2002. Calibration of pho- absorbed dose to water. The calibration factors are measured for ton and beta ray sources used in brachytherapy. Guidelines on each ionisation chamber and beam quality in terms of exposure or standardized procedures at Secondary Standards Dosimetry air kerma according to the dosimetry protocol recommended by, Laboratories (SSDLs) and hospitals, IAEA TECDOC-1274. for example IAEA (2000). International Atomic Energy Agency, Vienna, Austria; Nath, R., The measured dose D expressed as a product of a calibration L. L. Anderson, G. Luxton, K. E. Weaver, J. F. Williamsson and factor k and the registered signal M should be proportional to M: A. S. Meigoni. 1997. Code of practice for brachytherapy phys- ics. Report of the AAPM radiation therapy committee task group D = k ´ M No. 56. Med. Phys. 24:1557–1598; Venselaar, J. and J. Pérez- Calatayud, eds. 2004. A Practical Guide to Quality Control of In reality, calibration factor k depends on a number of factors, Brachytherapy Equipment, ESTRO Booklet No 8. European including the radiation (photons, electrons etc.), the energy (lin- Society for Therapeutic Radiology and Oncology, Brussels, ear energy transfer (LET)) and intensity of the radiation; envi- Belgium. ronmental parameters, for example temperature, pressure etc.; and the characteristics of the detector, for example wall correc- Calorimeter tion factor. (Radiation Protection) Calorimeter is a device measuring the Abbreviations: IAEA = International Atomic Energy Agency absorbed dose by assessing the heating effect of the radiation (in and LET = Linear energy transfer. Latin ‘calor’ means ‘heat’). Calorimeters are often made from Related Articles: Absorbed dose distribution, Calibration, graphite. The heat measurement is based on the fact that heat ∆Q1 Calibration curve, Calibration depth, Ionisation chamber, lost by B1 is equal to the amount ∆Q2 gained by B2: Primary standard Further Reading: IAEA (International Atomic Energy Agency). 2000. Absorbed dose determination in external beam radiotherapy -DQ1 = DQ2 ® DQ1 + DQ2 = 0 technical report series no. 398. IAEA, Vienna, Austria. The conservation law of heat enables to define the specific heat cs Calibration source as the amount of heat ∆Q lost or gained by a body per unit mass, (Radiotherapy, Brachytherapy) per unit temperature changes ∆T: Calibration Source – Quality Control of Well-Type Chamber Stability: The well-type ionisation chamber is commonly used for DQ = cs ´ m ´ DT brachytherapy source calibrations, both at Secondary Standards Dosimetry Laboratories (SSDLs) and at hospitals. At the hospital where m is body mass. level, it is important to have a chamber that is stable, reliable and According to the first law of thermodynamics (conservation of easy to use, and the well-type chamber is the chamber recom- energy, including heat), the change in the internal energy ∆U of mended by the IAEA for photon-emitting sources (and for some a system is equal to the heat ∆Q added to the system minus the beta sources). work W done by the system: Calorimetry 139 C apacitive reactance DU = DQ - W When W = 0 → ∆U = ∆Q. This way the heat can be expressed in units of energy: Conducting electrode 1cal = 4.186J ®1J = 0.239cal Insulating film The amount of heat produced in the medium of known spe- Contoured cific heat cs and mass, placed in an insulated container, can be conducting silicon measured calorimetrically as change of temperature (∆T). This V(t) backplate C measurement enables us to evaluate the absolute amount medium of energy absorbed by this medium. Related Article: Calorimetry FIGURE C.4 Basic principle of a capacitive transducer. (Courtesy of Further Reading: Graham, D. T. and Cloke, P. 2003. EMIT project, www .emerald2 .eu) Principles of Radiological Physics, Elsevier, Edinburgh, London. Calorimetry The more common term in daily use is a capacitor, which is (Radiation Protection) Calorimetry is a method of heat measure- an electrical (electronic) component having primarily capacitive ment in the medium, which attenuates x-rays or gamma radiation. properties within the field of its intended use. The radiation x or gamma passing through the medium interacts Related Article: Capacitor with its atoms by producing the electrons (photoelectrons, recoil electrons of Compton scattering, pair production) and causes Capacitive micromachined ultrasound transducer (cMUT) many ionisations. The kinetic energy of electrons is absorbed by (Ultrasound) The basic principle for a capacitive transducer is a the atoms of the medium, resulting in an increase in their internal conducting backplate

with a matrix of small cavities, Figure C.4. energy ∆U, which can be measured quantitatively (according to The cavities are sealed with a thin insulating membrane with a the first law of thermodynamics ∆U = ∆Q) as a temperature rise conducting top layer. The electrodes are biased with a DC voltage ∆T. The produced heat ∆Q is equal: and an AC signal can either be applied to move the membranes or be generated by vibrating the membranes. This is due to the fact DQ = cs ´ m ´ DT that the distance between the two conducting planes acts like a capacitor with altering capacitance (Farads). One transducer ele- ment is built up of a matrix of these cavities. Advanced IC fab- where rication processes are used to fabricate the structure of the small cs (J/kg/K) is the specific heat capacity of the medium cavities, thin membranes and electrodes with submicron precision. m (kg) is its mass In recent years, the cMUT technology has emerged as an alter- ∆T (K) is measured temperature rise native to piezoelectric ultrasound transducers offering advantages such as wide bandwidth, ease of fabricating large arrays and Using calorimetry, one can perform an absolute measurement potential for integration with supporting electronic circuits (see of absorbed dose D as a ratio of absorbed energy per unit mass of Khuri-Yakub et al., 2000). the body (medium): Further Reading: Khuri-Yakub, B. T., C. H. Cheng, F. L. Degertekin, S. Ergun, S. Hansen, X. C. Jin and Ö. Oralkan. 2000. DQ D = (J/kg = Gy) ® D = cs ´ DT Silicon micromachined ultrasonic transducers. Jpn. J. Appl. Phys. m 39:2883–2887. Example: for D = 1 Gy → ∆T about 10−4 K Capacitive reactance This method is used for very large absorbed dose of radiation. (General) The imaginary part of the complex impedance Z due to Related Article: Calorimeter presence of capacitance in an electrical circuit is given by Further Reading: Graham, D. T. and P. Cloke. 2003. Principles of Radiological Physics, Elsevier Science Limited, Z = R + iXC Edinburgh, UK. where CAP (computer-aided perception) R is the resistance (General) See Computer-aided perception (CAP) XC is the capacitive reactance i is equal to √(−1) Capacitance (General) Capacitance is the property of body to store electri- The capacitive reactance is given by cal charge. The symbol for capacitance is C, and its value is expressed in farads (F). It is defined as the ratio of the charge Q 1 (C) stored on the body and the voltage difference U (V) across XC = wC the conductive parts of the body. The capacitance of a body is determined by the size and the shape of its conductive parts and where the dielectric properties of the material between the conductive ω is the angular frequency parts. C is the capacitance Capacitor 140 Capacitor v, i Current Voltage 90° t C FIGURE C.5 Graph showing current leading the voltage in AC circuits by phase angle of 90°. Capacitive reactance decreases with frequency. Capacitive reactance causes voltage and current to become out of phase in AC circuits. In purely capacitive circuits, the current FIGURE C.6 Single-layer ceramic capacitors. leads the voltage by the phase angle 90°, Figure C.5. Reactance is expressed in ohms (Ω). Related Articles: Capacitance, AC current, AC voltage Capacitor (General) A capacitor is an electrical or electronic component that has dominantly capacitive properties. It consists of at least two conductors or semiconductors separated by an insulator. The conductive parts of a capacitor are described as electrodes or plates. The value of the capacitance of a capacitor depends on the area of the electrodes, the distance between the electrodes and the dielectric properties of the insulating material between the elec- trodes. For a parallel-plate capacitor, consisting of two parallel conductive plates, both having an area S and separated by a dis- tance d filled with a material having a relative dielectric constant ɛr, the capacitance is equal to S C = e0er d where ɛ0 is the absolute dielectric constant (permittivity of free space) and it equals 8.854 × 10−12 F/m. The dielectric material used in capacitors may be solid, liquid or gaseous. Capacitors FIGURE C.7 Multilayer ceramic capacitors. used in electronic circuits usually have a solid dielectric material separating the conductive plates. Accordingly, they are differenti- ated as ceramic, foil, mica, electrolytic and other types of capaci- insulator used – mica (a silicate mineral). Mica capacitors are tors (Figure C.6). used in high voltage and high frequency applications. They also Ceramic capacitors have ceramic as the dielectric material. have excellent temperature characteristics. Ceramic has high permittivity and therefore enables building large- Electrolytic capacitors have the dielectric layer formed by capacity capacitors of small volume. Nominal values of ceramic electrolytic process. One electrode (anode) is made out of a metal capacitors capacity span from sub-picofarad to a few microfarad conductor, aluminium or tantalum, while the other is formed values. Ceramic capacitors with nominal values larger than a few using conducting electrolyte. Electrolytic capacitors have high nanofarads are often built as multilayer capacitors, Figure C.7. capacitance per unit volume and also large leakage currents Foil capacitors use metallised plastic foils to form a capacitor. (Figure C.9). The foils are wound in order to get large areas of electrodes, i.e. Capacitors using liquid or gaseous dielectric materials are large capacitances. The electrical characteristics of foil capaci- often used in transducer circuits. Variable capacitors (trimmer tors are determined by the insulator foil properties. Commonly capacitors) use air as insulator and the nominal value of their used plastic materials are polypropylene, polystyrene, polyester, capacitance is usually up to a few tenths of picofarads. The sym- polycarbonate and others. Paper may also be used as insulating bols for capacitors are shown in Figures C.10 and C.11. The unit material for capacitors (Figure C.8). for expressing the capacitance of a capacitor is farad (F). Since Mica capacitors are high quality and reliability capacitors the capacitance of most capacitors built in electrical and elec- due to high dielectric strength and high chemical stability of the tronic circuits is several orders of magnitude smaller than a farad, Capacitor discharge generator 141 Capacitor discharge generator C FIGURE C.8 Foil capacitors. FIGURE C.9 Electrolytic capacitors. HV iron transformer Rectifier HV capacitor X-ray tube AC FIGURE C.12 Block diagram of capacitor discharge–high voltage generator. FIGURE C.10 Symbol for capacitor. subunits of farad are in use: millifarads (mF), microfarads (μF), nanofarads (nF) and picofarads (pF). Most capacitors have a fixed nominal value of their capacitance. Related Articles: Capacitance, Circuit(s) electrical + Capacitor discharge generator (Diagnostic Radiology) This high-voltage generator (HVG) is used mainly for low-power mobile x-ray equipment. Its design includes a power high-voltage capacitor (or set of capacitors) – ∼1 μF, connected after the HV rectifier (Figure C.12). During the charging period, the capacitors in the set are normally con- nected in parallel in order to assure equal charge for each one. During the exposure (discharge), the capacitors are connected in FIGURE C.11 Symbol for polarised capacitor. series, so their summary voltage can ensure the necessary high Capacity 142 C arbon voltage (kV). Often a grid-controlled x-ray tube is used with Furthermore, the central volume theorem f = v/t implies that dt/df these HVGs. The kVp waveform follows the discharge – i.e. the = −v/f2 = −t/f. Substituting dt/df = −t/f in Equation C.1 gives the kV decreases during the exposure (see article on Voltage wave- following: form). The kVp drop is an important parameter for this HVG. Normally, it is 1 kV/ mAs (i.e. if an 80 kVp/30 mAs exposure is w( f ) æ t ö æ t dR t made, then the exposure begins at 80 kVp, but ends at 50 kVp). = -ç ÷h(t ) ö ( ) = ç ÷ è f ø f dt (C.2) è ø Before each exposure, the capacitors need approximately 10 s to charge. This HVG is lighter than the battery-powered HVG and Finally, the capillary flow distribution w( f) is typically nor- has no special requirements to the mains electrical supply, but it malised to have unit mean flow and area. C cannot be used for powerful radiographs (e.g. of thick body parts Statistical comparisons of flow heterogeneity, represented or large patients). by w( f), observed in normal tissue and in ischaemic regions in Related Articles: High-voltage generator, Monoblock genera- patients with stroke have indicated that the ischaemic regions tend tor, High-voltage circuit, Voltage waveform to show a more homogeneous flow distribution. It has also been suggested that the region that shows abnormal flow heterogene- ity in the acute phase after stroke correlates well with the final Capacity infarct volume as determined by follow-up T2-weighted magnetic (General) As a general term, this defines the maximum quantity resonance imaging. of a substance that an object can accommodate. The capacity can Related Articles: Dynamic susceptibility contrast MRI, Mean usually be defined in terms of the maximum volume of fluid that transit time can be accommodated. Further Readings: King, R. B., G. M. Raymond and J. B. In electronics, the capacity refers to the amount of electrical Bassingthwaighte. 1996. Modeling blood flow heterogeneity. Ann. charge that can be stored within the device when a specific poten- Biomed. Eng. 24:352–372; Østergaard, L., D. A. Chesler, R. M. tial difference is applied across its electrical connections. Specific Weisskoff, A. G. Sorensen and B. R. Rosen. 1999. Modeling cerebral components, capacitors, are available with specified capacities in blood flow and flow heterogeneity from magnetic resonance residue the range from 10−9 to 10−2 F (C/V). data. J. Cereb. Blood Flow Metab. 19:690–699; Simonsen, C. Z., L. Related Article: Capacitor Røhl, P. Vestergaard-Poulsen, C. Gyldensted, G. Andersen and L. Østergaard. 2002. Final infarct size after acute stroke: Prediction Capillary blockade imaging with flow heterogeneity. Radiology 225:269–275. (Nuclear Medicine) Capillary blockade refers to the process of using radioactive particles to investigate patients with a suspected Carbon embolism, for example 99mTc-MAA (macro aggregated albumin) (General) is used for regional lung perfusion studies. When injected intravenously, the size of the radioactive par- Symbol C ticle used in these studies must be bigger than red blood cells to Element category Non-metal allow them to gather in the capillaries. By studying the accumula- Mass number A of stable isotopes 12 (98.93%), 13 (1.07%) tion of the radioactive particles in the capillaries, it is possible to Atomic number Z 6 draw conclusions regarding the relative blood flow in a specific organ or a specific part of an organ. The particles that gather in Atomic weight 12.0107 kg/kg atom Electronic configuration 1s2 2s2 2p2 the capillaries cause a number of micro-embolisms but the num- ber of particles typically injected is not enough to block more than Melting point 3925 K a small fraction of the capillaries. Boiling point 5100 K Density near room temperature 1900–2300 kg/m3 (1.9–2.3 g/cm3) Capillary flow heterogeneity (Magnetic Resonance) The heterogeneity of microvascular flows History: Since prehistory carbon has been known in the forms is known to be an important determinant of the efficacy of oxy- of soot and charcoal and indeed the element name comes from the gen delivery to tissue. For this, the function h(t) is introduced to Latin ‘carbo’, meaning coal. There are three naturally occurring describe the distribution of times required by the different tracer allotropes: amorphous carbon, as in soot; graphite, a soft mate- molecules to pass through the capillary system (following an rial mainly used as a lubricant; and diamond, one of the hardest instantaneous tracer input). In particular, h(t) is the distribution known materials. of transit times. The impulse tissue residue function R(t) can be In recent years, a new generation of carbon-based material obtained experimentally, for example by dynamic susceptibility called carbon nanotube (CNT) has been developed in nanotech- contrast MRI, and h(t) can then be calculated as the derivative nology laboratories. Carbon nanotubes are ordered molecular of R(t) with respect to time t, i.e. h(t) = −dR(t)/dt. By applying structures formed by carbon, yet different from graphite and dia- the central volume theorem (i.e. flow equals volume divided by mond. They are molecular scale tubes with typical diameter of transit time) to each transit time of the distribution given by h(t), a few nanometres and a height of up to a few millimetres. The and assuming that all vascular paths have

the same volume v, the tubes have remarkable electronic properties and special physical corresponding distribution w( f) of flow components f can be cal- characteristics that make them of great academic as well as com- culated. The application of vascular modelling presented by King mercial interest, especially as a cold cathode source of electrons et al. and Østergaard et al. provides the following relationship: in x-ray tubes in contrast to hot cathodes in Coolidge-type x-ray tubes. In a cold cathode, the electrons are emitted through the ( ) ( ) ( ) ( ) dt = Û = process of field emission; in a hot cathode, through thermionic w f df h t dt w f h t (C.1) df emission. Carbon-11 143 Carbon ion therapy (Medical Applications) Charged Particle Radiotherapy: Skin dose rate from 1 MBq: 0 µSv/h at 30 cm (point source); 0 Due to their Bragg peak and high RBE, carbon ion beams can µSv/h at 1 m (10 mL glass vial) offer therapeutic advantage over photon beams in certain clini- Total absorption (electron 0.320 mm in tissue (Lucite), 25 cm in cal situations (most notably in the treatment of anatomically range): air awkward radioresistant tumours). Carbon ion radiotherapy Biological half-life: bone 12–40 days requires a large initial investment (for a cyclotron or simi- Critical organ: soft tissue/fat lar accelerator), but the technique is rapidly gaining ground ALImin (50 mSv): 90 MBq worldwide. Related Articles: Charged particle therapy, Bragg peak, Absorbed dose: 0.095 mGy/MBq muscle Relative biological effectiveness (RBE), Cyclotron Effective dose: 0.024 mSv/MBq (ingestion) 0.002 mSv/MBq (inhalation) C Carbon-11 (Nuclear Medicine) A radionuclide used for in-vivo PET imaging. 6 3P0 0+ 5730 y Maximum C 14 6C β– 1+ Positron Positron Photon Common Carbon 14 Half-life Fraction Energy Emission Application 12.0107 7N 1s22s22p2 20.4 0.99 960 keV 511 keV Tumour and Q 11.260 β–156.475 3 minutes metabolic imaging Clinical Applications: 14C can be used for in vitro nuclear medicine. It is used to a small extent for, for example glycerol Carbon-11 can formulated into several different radiopharma- tri[1–14C]oleate test for fat malabsorption in the small intestine. ceuticals used to detect/image various types of cancers with PET Other examples are 14C-inulin and 14C-triolin for measurement of imaging. the glomerular filtration rate (GFR) and 14C-urea for CO2 breath tests (Helicobacter pylori infection). 1. 11C choline – this radiopharmaceutical is particularly In chemical and biological research, a wide range of useful in detecting recurrent prostate cancer before it is 14C-labelled radiochemicals are used, examples are 14C-lipids, detected by more-conventional imaging tests. nucleic acids, steroids, amino acids, proteins, carbon monoxide, 2. 11C acetate – this radiopharmaceutical is also used carbon dioxide, dopamine receptor ligands and various other dif- for detecting prostate cancer as well as lymph node ferent drugs in pharmaceutical research. metastases. Further Readings: Annals of the ICRP. 1987. Radiation Dose 3. 11C methionine – this radiopharmaceutical is one of the to Patients from Radiopharmaceuticals, Biokinetic Models and most commonly used PET tracers for evaluating brain Data, ICRP Publication 53, Vol. 18, Pergamon Press, Oxford, tumours. 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 Related Articles: Positron emission tomography, Radionuclide .lu. se /nu clear data/ toi/]; Firestone, R. B. 1999. Table of Isotopes, imaging 8th edn., Update with CD-ROM. [http://ie .lbl .gov /toi .html]; Further Readings: Bushberg, Seibert, Leidholdt and Boone. Kowalsky, R. J. and S. W. Falen, 2004. Radiopharmaceuticals in 2012. The Essential Physics of Medical Imaging, 3rd edn., Nuclear Pharmacy and Nuclear Medicine, 2nd edn., American Lippincott Williams & Wilkins; Cherry, Sorenson and Phelps. Pharmacists Association, Washington, DC; Taylor, D. M. 2004. 2012. Physics in Nuclear Medicine, 4th edn., Elsevier; Mettler Biokinetic models for the behaviour of carbon-14 from labelled and Guiberteau. 2012. Essentials of Nuclear Medicine Imaging, compounds in the human body: Can a single generic model be 6th edn., Elsevier; Zeissman, O’ Malley, Thrall and Fahey. 2014. justified? Rad. Prot. Dosimetry 108:187–202. Nuclear Medicine, 4th edn. Carbon dioxide as contrast agent Carbon-14 [14C] (Diagnostic Radiology) Carbon dioxide (CO2) gas can be injected (Nuclear Medicine) into a blood vessel to produce contrast. Compared to iodine con- trast media, CO2 produces negative contrast because of its low Element: carbon density and atomic number. Some advantages of CO2 as a contrast Isotopes: 2 < N < 16 agent are its low toxicity and allergic effects and rapid absorption and dissipation after an injection. Atomic number (Z): 6 Neutron number (N): 8 Symbol: 14C Carbon ion therapy Production: Reactor (Radiotherapy) Carbon ion therapy (CIRT) is a particle therapy which uses high energy carbon ions to treat tumours within the Daughter: 14N body. Like other forms of particle therapy, such as proton beam Half-life: 5730 years therapy (PBT), it uses the Bragg peak to obtain an advantageous Decay mode: β− – decay dose distribution. Radiation: β− 156.5 keV (max) 52 keV (mean) It was developed to treat patients in the 1970s at Lawrence Gamma energy: none Berkley Laboratory. The success of this work led Japan to create Carcinogenesis 144 C ardiac gating R RR interval R RR interval P T td P T td Q S Sequence Q S Sequence C tw tw FIGURE C.13 The collection of data are synchronised with the cardiac rhythm. When performing gating, data are only collected during a predefined time in the cardiac cycle. its first CIRT centre in the 1980s. Since then additional CIRT cen- to faithfully depict the motion. For both these cases, a synchro- tres have been built in Japan, China, Germany, Italy and Austria. nisation of the imaging sequence to the heart rhythm is neces- The source of carbon ions is usually a synchrotron accelera- sary (Figure C.13). tor as centres may wish to allow treatment with other heavy ions. For synchronisation, a device is used to detect the R-peak of Treatments were originally delivered using passively scattered the ECG. For cardiac imaging, VCG (vector ECG) is most com- carbon ion beams, but following on from work carried out at GSI monly used. If this is not applicable, a PPU (peripheral pulse unit) in Germany in 1997, active scanning treatments are now available can be used instead. The devices produce a signal at the detection at several centres. of an R-peak, and this signal is fed into the MR-scanner as a trig- There are several key differences to PBT, the most com- ger to the acquisition. mon form of particle therapy. Firstly, the range of LET values There are different strategies in how to collect the data in found in carbon treatments is much greater than that in PBT. respect to the R-trigger. The most basic strategy is cardiac gat- Consequently, variable radiobiological effect must be considered ing. Trigger delay (td) and gate width (tw) are defined, and data is in CIRT. Centres typically use a version of one of two models to acquired in the time period where the gate is open. This method is do so, either the microdosimetric kinetic model (MKM) or the especially well-suited for morphological series, for example coro- local effect model (LEM). nary artery and infarct (delayed enhancement) imaging. The gate Further differences are found in the lateral penumbra and the is then open during late diastole, where the heart is not moving. distal edge. Due to the heavier mass of the carbon ion, it scatters For functional imaging, one needs to collect data over as less than the proton and thus the lateral edge of the beam is sig- large a part of the cardiac cycle as possible. The two approaches nificantly sharper. However, due to nuclear fragmentation at the are prospective and retrospective triggering. The prospective end of range, a dose tail extends beyond the Bragg peak. This is triggering starts data acquisition immediately after the R-peak not seen in PBT. is detected, and collects data for a predefined time. The acqui- Further Readings: Ebner, D. K. and T. Kamada. 2016. The sition then waits for the next trigger signal. Retrospective trig- emerging role of carbon-ion radiotherapy. Front. Oncol. [Online] gering acquires data continuously, and uses the trigger signal to 6(June):6–11; Mohamad, O., et al. 2017. Carbon ion radiotherapy: keep track of when in the cardiac cycle a specific set of data was A review of clinical experiences and preclinical research, with an measured. Retrospective triggering is the most commonly used emphasis on DNA damage/repair. Cancers. [Online] 9(6):1–30. method as it covers the complete RR-interval, whereas when using prospective triggering, information is not acquired in late diastole Carcinogenesis (Figure C.14). (Radiotherapy) See Radiation-induced secondary malignancies Cardiac gating Cardiac blood-pool imaging (Nuclear Medicine) Cardiac gating is an imaging technique in (Nuclear Medicine) Cardiac blood-pool imaging is also called which images are synchronised to electrocardiogram (ECG) sig- gated blood-pool imaging or multi-gated acquisition scan nals, thus permitting images of the heart to be formed in different (MUGA). For more information, see Multi-gated acquisition scan (MUGA). R RR interval R Cardiac cineangiography (Diagnostic Radiology) See Cineangiography P T P T Cardiac gating (Magnetic Resonance) When performing cardiac MR, one must Q S Q S deal with the fact that the heart is beating. During a heart beat, Sequence the heart is translated, rotated and twisted. Blood is pumped between atria and ventricles. If the aim is to produce morpho- FIGURE C.14 In retrospective triggering, data are acquired continu- logical images of the heart, it is desired to freeze this motion. If, ously. It is sent to the reconstruction with a time stamp, identifying from on the other hand, functional information is desired, the goal is which phase in the cardiac cycle it belongs to. Cardiolite 145 Carrier-free sample phases of the cardiac cycle. Cardiac images acquired without gat- Carr–Purcell (CP) ing techniques are often blurred due to the contraction and relax- (Magnetic Resonance) The Carr–Purcell pulse sequence consists ation of the cardiac muscle. The technique is used whenever the of a 90° excitation RF pulse followed by a train of 180° RF refo- data acquisition is too slow to occur under a fraction of the car- cusing pulses in order to recall several echoes after one excitation diac cycle. The trigger position in the cardiac cycle is determined pulse. The time separation between the 90° pulse and the first 180° using the ECG, the signal from which is fed live to the acquisition pulse is τ, and the separation between following 180° pulses is 2τ. electronics. Images are thus acquired at different times in the car- In the Carr–Purcell sequence, the phase angle of all RF pulses are diac cycle and arranged in a movie loop. the same, i.e. 90x –τ– 180x –2τ– 180x –2τ– 180x –2τ– 180x …, which The typical count rates in nuclear medicine are too low to pro- leads to an accumulation of phase errors and an additional reduc- duce a sufficient image from a single cardiac cycle. Therefore, tion of the echo amplitude as the echo train is prolonged, occurring an image is typically comprised of an average of 50–100 images. due to, for example imperfections of the 180° pulse. This limiting C A common diagnostic examination in nuclear medicine is gated factor was solved in the CPMG pulse sequence, where a 90° phase blood-pool scanning (also known as multi-gated acquisition or shift was introduced between the excitation pulse and the succes- MUGA). This type of scan can be used to determine the size of sive refocusing pulses. Alternatively, a 180° phase cycling between the ventricles, look for abnormalities in the heart wall (like aneu- successive 180° RF pulses also compensates for this problem. rysms) and studying abnormal blood flow between the ventricles. The Carr–Purcell pulse sequence was originally created to It is used to assess the function of the heart by the calculation of measure the complete echo decay by creation of multiple recalled the left ventricle ejection fraction. echoes needed, for example for T2 measurements. This is also the Related Article: Multi-gated acquisition (MUGA) first multi-echo pulse sequence. Related Article: Carr–Purcell–Meiboom–Gill (CPMG) Cardiolite sequence (Nuclear Medicine) Cardiolite is a trade name for Technetium- 99m Sestamibi. For further information see Tc-99m SestaMIBI Carr–Purcell–Meiboom–Gill (CPMG) sequence Related Article: Tc-99m SestaMIBI (Magnetic Resonance) The Carr–Purcell–Meiboom–Gill pulse sequence is a modification of the Carr–Purcell pulse sequence. It C-arm in fluoroscopy consists of

a 90° excitation RF pulse followed by a train of 180° (Diagnostic Radiology) Special type of C-shaped support holding RF refocusing pulses, which in contrast to the CP pulse sequence the x-ray tube and image intensifier (II). The C-arm can be rotated are phase shifted by 180°. The timing between the RF pulses is around the patient around three rotational axes, thus allowing the same as in the CP pulse sequence, i.e. τ between the 90° and fluoroscopy from various angles. C-arm mobile fluoroscopic sys- the 180° RF pulses and 2τ between successive 180° RF pulses. tems (Figure C.15) are very useful for intra-operative imaging, The CPMG pulse sequence hence consists of the following pulses; as these allow easy operational access of the personnel in the 90y –τ– 180x –2τ– 180−x –2τ– 180x –2τ– 180−x …; the phase shift theatre. C-arm systems are also used for angiographic systems between successive 180° was introduced to inhibit the accumula- (sometimes using two separate C-arms – biplane system). The tion of phase errors. distance between the x-ray tube and II is usually between 90 and Related Article: Carr–Purcell (CP) 120 cm (achieved by moving the II close or away from the tube). Related Articles: Mobile unit, Fluoroscopy, Biplane cine Carrier-added radioisotope system (Nuclear Medicine) When a known amount of the correspond- ing stable isotope has been added to the radioactive sample, it is referred to as carrier-added. The stable isotope is called the carrier. Carrier-free sample (Nuclear Medicine) A sample with a radioisotope can also contain other isotopes of the same element. For instance, a 131I sample can also contain the stable isotope 127I. When an isotope of the same element as the radionuclide of interest is present in a sample, the sample is said to be with carrier. A sample without a stable isotope is said to be carrier free. Such pure samples are pretty much impossible to manufacture without a small fraction of car- rier atoms. Therefore, the term without carriers is also used to describe carrier-free solutions. If the carrier atoms are also radio- active, they are unlikely to share the same decay characteristics, i.e. one isotope is suitable for imaging while the other is not. The second isotope will only contribute to an increase in the total patient radiation dose without giving any diagnostic information. One always strives to attain carrier-free samples, and the choice of production method is often decided by which of the methods produces the sample with the smallest amount of carriers. The sample specific activity is the ratio between radioisotope activity and the total mass of the element present. When there are no car- FIGURE C.15 Typical mobile fluoroscopic x-ray equipment with C-arm. rier isotopes of the same element as the radioisotope of interest, In this system, the II (on top) is fixed. the specific activity is called carrier-free specific activity. Carrier-free specific activity (CFSA) 146 Cataracts (Eye) Further Reading: Cherry, S. R., J. A. Sorenson and M. E. and noise propagation in cascaded imaging systems. Med. Phys. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, 21(3):417–427. Philadelphia, PA, pp. 38–40. Cassette carriage Carrier-free specific activity (CFSA) (Diagnostic Radiology) A tray holding the cassette with x-ray film (Nuclear Medicine) Carrier-free specific activity (CFSA) is the in radiography systems. There are two main types of carriage – highest possible specific activity of a radionuclide. A sample with used in patient table (see Undertable cassette carriage) or used a radioisotope can also contain stable isotopes of the same ele- in vertical x-ray stands for chest imaging. These are most often ment. For instance, a 131I sample can also contain the stable isotope mounted in one assembly with the anti-scatter grid and the detec- 127 C I. When a stable isotope of the same element as the radionuclide tors (dominants, cells) of the automatic exposure control (AEC) of interest is present in a sample, the sample is said to be with car- system. The cassette carriage has metal clips holding different rier. A sample without a stable isotope is said to be carrier free. cassette sizes. The centre of the cassette carriage is aligned with Such pure samples are pretty much impossible to manufacture the centre of the x-ray beam. without a small fraction of impurity atoms. Therefore, the term Related Articles: Radiography, Bucky table, Undertable cas- without carriers is also used to describe carrier-free solutions. sette carriage The sample specific activity is the ratio between radioisotope activity and the total mass of the element present. When there are Cassette changer no polluting isotopes of the same element as the radioisotope of (Diagnostic Radiology) Equipment used to quickly change the interest, the specific activity is called carrier-free specific activity. cassettes with x-ray films in x-ray radiography systems per- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. forming sequences of exposures. This equipment usually has Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, a magazine loaded with cassettes with unexposed x-ray films; Philadelphia, PA, pp. 38–40. a mechanism quickly moving the cassette in front of the x-ray beam; and a magazine for the cassettes with exposed films. Carrier-mediated diffusion of tracers Cassette changers used in fluoroscopic systems can perform (Nuclear Medicine) The term ‘carrier-mediated diffusion of up to – two to three exposures per second, while fast cassette tracers’ refers to a passive transport of tracers in the same direc- changers used in angiography can have around 10 exposures per tion as a concentration gradient within the body. An example of second. These systems are not used anymore, as contemporary carrier-mediated transport is the transport of glucose and amino digital x-ray equipment can acquire images with speed above acids across the blood–brain barrier. The transport uses a carrier 25 fps. molecule to transport the substrate across a barrier. They form a substrate/carrier complex that can physically move across the Cassette, filmless barrier. At the other side of the barrier, the complex decays into (Diagnostic Radiology) An old name for computed radiography the original substrate and carrier molecule. Since there are only (CR) using storage phosphor plate (the cassette does not contain a a finite number of carrier molecules, this type of transport can film, but a storage phosphor plate). be saturated. The carrier molecule is typically an enzyme that is Related Article: Computed radiography neither created nor destroyed in the process but rather acts as a catalyst. Further Reading: Cassette size Cherry, S. R., J. A. Sorenson and M. (Diagnostic Radiology) The sizes of x-ray films and their hold- E. Phelps. 2012. Physics in Nuclear Medicine, Saunders, ers (x-ray cassettes) are standardised according to the anatomi- Philadelphia, PA. cal regions to be radiographed. Usually, the sizes are two types – American/English using inches (e.g. 8 × 10; 10 × 12 in., etc.), Cartesian coordinates and European using centimetres (e.g. 18 × 24, 24 × 30 cm, etc.). (General) The Cartesian coordinate system is a 2D reference system where every point can be determined by two coordinates. Cassette, Wisconsin The coordinates are defined by two perpendicular lines with a (Diagnostic Radiology) See Wisconsin test cassette unit length. The lines are generally called the x- and y-axis. The Cartesian coordinates can be used to model geometric shapes Catapult bucky described by algebraic functions. (Diagnostic Radiology) A mechanism that moves the anti-scatter The 3D case of the Cartesian coordinate system provides the grid during the x-ray exposure to blur and reduce the visibility three physical dimensions of space – length, width and height. of the grid lines in the image. It is also known as a reciprocating The location of a specific point in 3D Cartesian coordinate system Bucky (or in a specific company device – Lysholm raster). is given by the x-, y- and z- coordinates. Cataracts (Eye) Cascaded imaging systems (Non-Ionising Radiation) The term is used to indicate the whiten- (Diagnostic Radiology) The analysis of complex imaging systems ing and clouding of the lens of the eye, this causes the eyesight includes their description by a series of cascaded stages, each one to gradually get worse. Caractogenesis might occur naturally having either a physical and/or mathematical meaning. For exam- with ageing or in the presence of a co-morbidity, such as diabe- ple, cascaded-systems analysis can be used for the assessment of tes. Exposure to solar, and in particular UVA, radiation has been an imaging parameter (e.g. DQE) of a complex detector. associated with developing cataracts. Further Reading: Cunningham, I., M. Westmore and A. Related Articles: AORD, Eye, Lens, UV light hazard Fenster. 2018. A spatial-frequency dependent quantum account- Further Readings: Coleman, A., F. Fedele, M. Khazova, P. ing diagram and detective quantum efficiency model of signal Freeman and R. Sarkany. 2010. A survey of the optical hazards Cathode (of an x-ray tube) 147 Cathode (of an x-ray tube) associated with hospital light sources with reference to the Control The very high temperature of the filament leads to some evap- of Artificial Optical Radiation at Work Regulations 2010. J. oration of the tungsten wire. This evaporation leads to shortening Radiol. Prot. 30(3):469; ICNIRP. A closer look at the thresholds of the life of the cathode (thinning it). Normally, the cathode life of thermal damage: Workshop report by an ICNIRP task group. at this temperature is not more than 1000 working hours. Due Health Phys. 111(3):300–306; ICNIRP. 2013. Guidelines on limits to this reason, the cathode is heated to this high temperature for of exposure to incoherent visible and infrared radiation. Health limited time only (during the x-ray exposure). But to heat the Phys. 105(1):74–91; ICNIRP. 2013. Guidelines on limits of expo- cathode from room temperature to 2700 K takes time. In order to sure to laser radiation of wavelengths between 180 nm and 1000 keep the heating time short, the cathode always stays preheated µm. Health Phys. 105(3):271–295; ICNIRP. 2004, Guidelines on to temperature around 1500 K. The preheating is made by apply- limits of exposure to ultraviolet radiation of wavelengths between ing a constant stand-by filament current through the cathode (less 180 nm and 400 nm (incoherent optical radiation). Health Phys. than 1 A). C 87(2):171–186; ICNIRP. 2000. Revision of the guidelines on limits The variation of the anode current (the thermal electrons fly- of exposure to laser radiation of wavelengths between 400 nm and ing from cathode to anode) is achieved by changing the tempera- 1.4 µm. Health Phys. 79(4):431–440; Sihota, R. and R. Tandon. ture of the cathode, which in turn is achieved by changing the 2011. Parsons' Diseases of the Eye, Elsevier, India. filament current (If). The density of the thermal emission current is described with the Richardson equation: Cathode (of an x-ray tube) (Diagnostic Radiology) The cathode is the negatively charged J = A T e w kT 0 0· 2· - / source of electrons in the x-ray tube. In almost all contemporary tubes, the cathode assembly (Figure C.16) consists of a heated where tungsten wire (filament), which produces thermal electrons. The J0 is density of the emission current emission of thermal electrons depends greatly on the temperature T is temperature of the emitter (in Kelvin degrees) of the cathode. In order to obtain an anode current on the order k and w are constants (k – Boltzmann constant, w – work func- of several amperes (for heavy x-ray exposure), the cathode must tion, for tungsten = 4.5 eV) be heated to some 2000–2700 K. Due to this reason, the material A0 is constant depending on the material of the emitter (for used for preparation of x-ray cathodes is normally tungsten. Its tungsten = 60 A/cm2/K2) high melting point (3410°C) and its ability to be drawn into very thin wire were the most important parameters to be considered The electrons emitted from the heated cathode form an elec- for this choice. Normally, the cathode is a long (∼5–20 mm), thin tron cloud around it. This cloud is called ‘space charge’. When (∼0.1–0.3 mm) tungsten wire in spiral coil with diameter on the the electrons leave the filament, the cathode loses part of its order of 0.2 mm. One of the ends of the filament is also connected negative charge and becomes more ‘positive’. This attracts back to the negative side of the high voltage. The electrical resistance some of the electrons. Normally, an equilibrium state exists of cathode

filament is relatively high and changes from approxi- between the number of electrons emitted and these attracted mately 0.1–0.3 Ω when cold, to some 2–6 Ω when heated above back. The cloud remains near the filament until high voltage 2000 K (ohmic heating). (from 20–150 kV) is applied between the cathode and the anode. Thus, the thermal electrons are accelerated towards the anode by the electric field between the cathode and anode, forming the anode current. The area of the anode, which is bombarded by the 1 thermal electrons and produces x-rays, is called actual (or ther- mal) focal spot. The size of the actual focal spot depends on the size of the cathode wire (in fact, the size of the electron beam generated by it). Most of the x-ray tubes have two foci – one 2 small filament wire (up to 1 mm long) – known as ‘fine focus’ and one bigger filament wire (∼2–3 mm long) – known as the ‘broad focus’. The foci (the filament coils) can be placed side by side (as in the Figure C.16); in a line (one above the other); in different depths (one behind the other), etc. (there are various designs). 3 Normally, the beam of thermal electrons produced by the cath- ode filament is quite spread, resulting in an increased area of the focal spot. This enlarged size of the source of radiation blurs the x-ray image. In order to focus the beam of thermal electrons and to decrease the space charge effect, the cathode filament is placed in a focusing cup (a half-pipe groove, known also as Wehnelt elec- trode, or Wehnelt cylinder). The focusing cap is specially shaped and is made of molybdenum, nickel or steel, because of their poor thermionic emission. The cup can be equipotential with the cath- ode, or under negative potential (Wehnelt electrode, used in grid FIGURE C.16 Cathode of an x-ray tube: (1) filament of the broad focus control x-ray tubes). (in its focusing cup); (2) filament of the fine focus (in its focusing cup) Related Articles: Filament circuit, Filament heating, Wehnelt and (3) cathode side of the x-ray tube (photo from a broken x-ray tube). electrode, X-ray tube, Focal spot, Stationary anode, Rotation (Courtesy of EMERALD project, www .emerald2 .eu) anode Cathode ray tube 148 Cavitation Cathode ray tube of CT systems. Depending on the model, they contain a number of (Diagnostic Radiology) A cathode ray tube (CRT) is an evacuated inserts that can be used to evaluate various parameters including: electronic tube (in glass envelope) with a cathode (electron source) at one end and a fluorescent screen at the other. The electrons from • uniformity the cathode are focused into a small beam and accelerated to the • noise fluorescent screen by a set of electrodes, including an anode oper- • Hounsfield numbers ating at a high voltage relative to the cathode. • spatial resolution A bright spot is produced where the electron beam strikes the fluorescent screen. Catphan® phantoms consist of cylindrical modules, each for a C The beam can be deflected and the bright spot moved over the different purpose, that can be stacked (Figure C.17e). screen surface either by a potential applied to electrodes within the Figure C.17a–d (Husby et al. 2017) shows a set of images of tube or currents applied to magnetic coils located around the tube. a Catphan® 600 phantom, showing the range of capabilities it A CRT is the component used in an oscilloscope to display offers to assess different aspects of image quality. various waveforms applied to deflect the beam, usually in the ver- Abbreviation: CT = Computed Tomography. tical direction, while the beam is being scanned in the horizontal Related Articles: Phantom, CT, Image noise, Spatial resolu- direction. tion, Contrast, Hounsfield number, Quality assurance CRTs are used for the display of video (television) images. The Further Reading: Husby, E., E. D. Svendsen, H. K. Andersen beam is scanned over the surface of the screen in a pattern of hori- and A. C. T. Martinsen. 2017. 100 days with scans of the same zontal lines that are progressively moved in the vertical direction Catphan phantom on the same CT scanner. J. Appl. Clin. Med. to cover the image area. Phys. 18(6):224–231. At each beam location within the image, the intensity of the elec- Hyperlinks: Catphan® is open-source. tron beam is modulated by the video signal transmitted from the camera to control the brightness of each spot in the displayed image. These days the diagnostic TV monitors with CRT are being Cavitation replaced by flat-panel displays. (Ultrasound) Cavitation is a general term that covers a wide Hyperlink: The cathode ray tube site: http://members .chello range of phenomena all having as a common factor that a cavity .nl/∼h.dijkstra19/page3 .ht ml is formed within a liquid, whether the cavity is empty, or con- taining gas or vapour. Formation of cavities may be, for example Cathode rays as a result of explosions or boiling. Acoustic cavitation is, in the (Diagnostic Radiology) Cathode rays are electrons emitted from context of medical ultrasound, formation of a cavity in response a cathode in an evacuated tube such as an x-ray tube or a cathode to an acoustic field. Such cavities are formed from pre-existing ray (CR) tube. microscopic bubbles or bubble nuclei with a size on the order of Related Article: Cathode ray tube microns or also from impurities in the liquid that can act as a seed Hyperlink: Wikipedia: http://en .wikipedia .org /wiki /Cathode_ for cavitation. rays There are two distinct categories of cavitation. In stable cavitation, a bubble undergoes periodic pulsations in an acous- Catphan phantom tic field. In transient cavitation, which occurs at higher acous- (Diagnostic Radiology) The Catphan® phantoms are a family of tic pressures, bubbles reach a maximum size and then rapidly phantoms used for characterisation of the imaging performances collapse. The rapid collapse is a high-energy event and carries FIGURE C.17 Images of a Catphan® 600 module, showing: (a), the sensitometry module; (b) the high-resolution module, (c) the low-contrast module, (d) the uniformity module, (e) Aligning of the CATPHAN phantom. Cavity-gas calibration factor 149 CCD array, scanner potentially destructive effects such as, for instance, breaking where cell membranes. The sound energy (sound pressure) needs to Awall is a factor that corrects for attenuation and scatter in the be above a certain threshold level for a bubble to go from stable wall and build-up cap to transient cavitation. In order to give ultrasound operators an Aion is the ion collection efficiency in the user’s chamber at the indication of the degree of possible risk of transient cavitation, time of Co60 exposure calibration at a primary dosim- the mechanical index has been introduced. It should be noted etry laboratory that this index only gives indication of the likelihood of cavita- k is the charge produced in air per unit mass per unit exposure tion and the effect responsible for tissue damage may be some (2.58 × 10−4 C/kg/R) other mechanism. Further, the model for this index assumes a βwall is the quotient of absorbed dose divided by the collision spherical nucleus, and other effects may occur for other shapes fraction of kerma in the chamber wall, 1.005 of nuclei. (L /r)wall gas is the ratio of the mean restricted collision mass C stopping power of the wall material to that of the gas (m a en /r) ir Cavity-gas calibration factor wall is the ratio of the mean mass energy absorption (Radiotherapy) The cavity-gas calibration factor indicated by coefficient for air to that of the wall for Co-60 gamma N rays gas is the dose to the gas in the ionisation chamber per unit elec- trometer reading. It is constant for all the radiation qualities for which the average energy expended in the production of one ion Calcium tungstate (CaWO4) pair (W) is the same as that for Co-60 gamma rays (33.7 eV). The (Diagnostic Radiology) CaWO4 (crystalline calcium tungstate) absorbed dose to the air in the cavity of an ionisation chamber is the original phosphor used for x-ray intensifying screens is related to the dose to the wall through the Bragg–Gray equa- (since 1896). This material can be found naturally (scheelite), tion and the dose to the wall material can be related to the dose but it is its synthetic form that is used for radiographic screen to air in absence of the ionisation chamber, which is also related – films. Calcium tungstate must be free of impurities to pro- to the exposure assuming a condition of electronic equilibrium. duce good fluorescence. It emits light from 350 to 580 nm (with Consequently, dose to the air in the cavity is related to the expo- peak at 430 nm). This violet light is not seen well by human sure calibration factor of an ionisation chamber. eye, but the photographic emulsion of specific radiographic Dgas dose to the gas in the ionisation chamber is directly related films is very sensitive to it. Calcium tungstate has conversion to the charge per unit mass in the gas by efficiency on the order of 15 (while CsI has ∼45 and CdWO4 has ∼40) and has been replaced by new screens using various W ö rare earth elements. gas = æ D Jgas ç ÷ Related Article: Screen film è e ø Further Reading: Curry, T. S., J. E. Dowdey and R. C. Murry. where 1990. Christensen’s Physics of Diagnostic Radiology, Lea and D Febiger, Philadelphia, PA. gas is the dose to the gas (Gy) W/e is the quotient of the average energy expended to produce an ion pair divided by the electronic charge; for room CBF (cerebral blood flow) air, W/e = 33.7 J/C (Magnetic Resonance) See Cerebral blood flow (CBF) Jgas is assumed to be corrected for ion recombination CBV (cerebral blood volume) As the response of the electrometer is also directly related to (Magnetic Resonance) See Cerebral blood volume (CBV) Jgas, the quotient of Dgas by the electrometer reading M is a con- stant, which depends on the dimensions and composition of the CCD array, scanner ionisation chamber. The ratio D (Diagnostic Radiology) The core component of an x-ray film gas/M is referred to as cavity-gas calibration factor N scanner (or any film/document scanner) is the charge-coupled gas, i.e. device (CCD) array. The array includes a number of sensors, transferring the information through the CCD mechanism. The A N ion sensors are usually photodiodes, detecting the intensity of the gas = Dgas M scanning laser beam, passing through the transparent x-ray where film (in the horizontal direction). The number and size of sen- N sors determine the spatial resolution in the horizontal direc- gas is the dose to the gas in the chamber per electrometer reading (Gy/C or Gy/scale division) tion, measured initially in dots per inch (DPI), or pixels per A inch (PPI), which can be transferred into lp/mm (line pairs per ion is the ionisation collection efficiency at the time of calibra- tion at a primary dosimetry laboratory millimetre). For example, a CCD array with 2550 sensors will be neces- N sary for scanning a horizontal area of 8.5 inches (21.59 cm) with gas is unique to each ionisation chamber and does not depend on the composition of the dosimetry phantom. It is applicable to spatial resolution of 300 dpi (i.e. 5.9 lp/mm) – 8.5 × 300 = 2550. all ionising radiations for which W/e has the value quoted earlier. This example is for scanning a film in greyscale. N If the film scanner is used for colour, the CCD array will need gas can be calculated using Nx, exposure calibration factor (R/C or R/scale division), uncorrected for ion recombination from three parallel sets of sensors (in the example above 2550 × 3 = 7650 sensors) each covered with different transparent filter (red, blue and green). k (W /e) Aion Awallb wall N See the article on X-ray film scanner. gas = Nx ( / )wall ( )air L r men /r Related Articles: X-ray film scanner, Film digitisers, CCD gas wall (Charge-coupled device) CCD coupling (charge-coupled device coupling) 150 CDMAM CCD coupling (charge-coupled device coupling) Different versions of the phantom were produced. (Diagnostic Radiology) Digital imaging tends to replace the tradi- CDMAM Version 3.2 was developed for screen-film detectors; tional film-based techniques. Digital technologies

can be divided the discs were arranged in 16 rows and 16 columns with diameters into two broad categories, computed radiography (CR) and digi- ranging from 0.1 to 3.2 mm and thickness ranging from 0.05 to 1.6 tal radiography (DR) systems, while DR systems can be further μm. Within a row, the disc diameter is constant, with logarithmi- divided into two broad categories: cally increasing thickness. In CDMAM Version 3.4, which is more suitable for digital • Indirect DR, where the x-ray photons are converted to mammography, slightly different diameter and thickness ranges light using a phosphor, and for gold discs were introduced (diameter from 0.06 to 2 mm and C • Direct DR systems on the other hand eliminate the in thickness from 0.03 to 2 μm). stage of light photon creation and using photoconduc- CDMAM Version 4.0 contains 672 gold discs; 21 different tors, such as selenium produce an electrical charge. diameters were introduced, ranging from 0.08 to 2 mm. Every single diameter has its own optimised gold thickness range with Indirect conversion flat panel systems use optical coupling of an 16 different steps. imaging screen to an active matrix array. The CCD camera is a Regardless the version, two identical gold discs are placed in solid-state device composed of many discrete photo conducting each cell of the matrix: one is always in the central position, with cells. Optical light from the output phosphor is converted to elec- the other in a randomly selected corner. trons in the amorphous silicon photo conducting layer of the CCD. The CDMAM phantom is sold with four PMMA slabs, which Coupling refers to the process of effectively transferring the are used for the simulations of different breast thickness. Each optical light to the CCD in order to be converted to a signal. slab is 10 mm thick and has the same dimension as the phantom. However, this process introduces an inherent limitation, because To acquire a CDMAM image, the phantom is put on the breast the efficiency of collecting the light from the screen by the imag- support in combination with PMMA slabs (generally, 2 cm thick ing device is generally rather poor and the coupling efficiency plates below and 2 cm thick plates above the test object, but other can be very low, with the corresponding signal being lost. This is even worse when demagnification (M) is required, as the coupling efficiency drops by M2. Two common ways of CCD coupling are (a) lens coupling and (b) fibre-optic coupling (Figure C.18). Related Articles: Digital detectors, Charge-coupled device (CCD) Further Readings: International Atomic Energy Agency. 2014. Diagnostic Radiology Physics: A Handbook for Teachers and Students, Vienna, Austria; Tate, M. W., S. M. Gruner and E. F. Eikenberry. 1997. Coupling format variations in x-ray detec- tors. Rev. Sci. Instrum. 68:47. CDMAM (Diagnostic Radiology) CDMAM is a commercial phantom for image quality assessment in mammography systems, manufac- tured by Artinis (Nijmegen, Netherlands) (Figure C.19). It contains a 0.5 mm aluminium base with gold discs (99.99% pure gold) of varying thicknesses and diameters, inserted in a 5 mm thick Plexiglass cover. FIGURE C.19 CDMAM. FIGURE C.18 Two common ways of CCD coupling – (a) lens coupling; (b) fiber-optic coupling. CDRAD phantom 151 Cell survival plate combinations can be adopted). To perform the analysis, the acquisitions of at least eight images is required with little move- ment of the phantom between different exposures. CDMAM was designed to derive contrast-detail visibility threshold information. Human readout of the images is time-consuming and affected by inter-reader and intra-reader variability and by a learning effect. Therefore, analysis is generally performed with the auto- matic software CDCOM. The threshold thickness for differ- ent gold disc diameters is detected and the contrast-detail (CD) curves is calculated. C Hyperlink: www .artinis .com CDRAD phantom (Diagnostic Radiology) The CDRAD is a contrast-detail phan- tom, replicating the range of contrasts observed across a range of diagnostic planar x-ray imaging. This phantom consists of a 10 mm thick Plexiglas tablet con- taining a 15 × 15 matrix of squares with ‘hole’ details, varying exponentially in depth in one direction and exponentially in diameter in the perpendicular direction. Lead-painted engravings mark out the matrix structure, as well as the detail diameters and FIGURE C.20 Example of a contrast detail curve obtained using the thicknesses corresponding to each square in the matrix. CDRAD phantom. For the three largest diameter sizes, each square contains one detail. For the remaining detail diameters, each square contains a detail placed within the centre and a detail placed within a ran- cells from a single parent cell. It is divided into five sub-phases: dom corner. This random detail allows verification of the detail prophase, prometaphase, metaphase, anaphase and telophase). detection, without creating a pattern that could be easily deduced The cell cycle follows an orderly sequence of phases between one or recognised. mitosis (M phase) to the next. These phases are the S phase, dur- The phantom is imaged by being placed symmetrically within ing which DNA synthesis occurs, and the G1 and G2 phases, gaps PMMA tablets, of a thickness replicating typical patient sizes. before and after the S phase, respectively. Additionally after mito- The resulting image is scored by comparison to a reference sche- sis, cells may spend some time in a ‘resting’ phase G0. matic. From this, a contrast details curve expressing the contrast A set of cyclin-dependent kinases (CDK) have a major role in threshold for each diameter size can be created. An example of a the regulation of the cell cycle and act via a series of cell cycle contrast details curve is displayed in Figure C.20. checkpoints set at various stages of the cell cycle. If DNA damage To increase statistics and reduce subjectivity, at least three is detected, for example the checkpoint will either delay the cell images should be obtained under the same conditions and evalu- cycle until the damage is repaired or, if repair is not possible, tar- ated independently by three observers. Alternatively, a software get the cell for destruction via programmed cell death (apoptosis, package may be used to automatically evaluate the images. also sometimes referred to as cell suicide). Related Articles: Contrast detail Further Readings: Pascoal, A., C. P. Lawinski, I. Honey and P. Blake. 2005. Evaluation of a software package for auto- Cell proliferation mated quality assessment of contrast detail images – comparison (Radiotherapy) Cell proliferation refers to the increase in the with subjective visual assessment. Phys. Med. Biol. 50(23):5743; number of cells as a result of growth and cell division. Cancer Thijssen, M. A., K. R. Bijkerk and R. J. Van der Burght. 1998. cells demonstrate loss of proliferative control, whereby a group of Manual CDRAD-phantom type 2.0. Project Quality Assurance cells display uncontrolled growth. in Radiology. Section Clinical Physics, Dept. of Radiology. St. Related Article: Cell cycle Radboud, The Netherlands: University Hospital Nijmegen. Cell survival Ceiling mount unit (Radiotherapy) Cell survival, and its opposite cell death, have (Diagnostic Radiology) X-ray stand mounted on the ceiling (usu- different meanings depending on the context. For non-prolifer- ally without the generator). For simple radiography units, the ating cells such as nerve and muscle, death can be defined as ceiling unit holds only the x-ray tube on a telescopic column. the loss of a specific function. For proliferating cells such as Ceiling-mounted complex C-arm units (used mainly in angio- stem cells, the appropriate definition is the loss of capacity for graphic devices) hold the tube plus the image intensifier. sustained proliferation. This is sometimes called reproductive Related Article: C-arm death and is the end point measured with cells cultured in vitro. A survivor that has retained its ability to reproduce and thus can Cell cycle proliferate indefinitely to produce a large clone or colony is said (Radiotherapy) The cell cycle is the series of events that take place to be clonogenic. in a cell, leading to its replication, and it is defined as the interval Related Articles: Cell cycle, Cell proliferation between the midpoint of mitosis in a cell and the midpoint of the Further Reading: Hall, E. J. and A. J. Giaccia. 2006. subsequent mitosis in both its daughter cells (mitosis is the pro- Radiobiology for the Radiologist, 6th edn., Lippincott Williams cess of cell division resulting in the production of two daughter & Wilkins, Philadelphia, PA. Cell survival curve 152 Centre of rotation (see p. 211 for clinical application) Dose (Gy) systems field of view, FOV. Due to certain system properties, information close to the edges is not suitable for imaging or acti- 0 10 20 30 40 vation quantification. The portion of the FOV used for these pur- poses is called useful field of view, UFOV. During the quality control of the camera system, the resolution, linearity and unifor- mity are measured over the UFOV and also over the central field of view, CFOV. The CFOV is defined as 75% of UFOV and is the smallest of the three FOVs. Further Reading: Cherry, S. R., J. A. Sorenson and M. E. C Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA, p. 249. Related Articles: Field of view (FOV), PET, SPECT, Useful field of view Central processing unit (CPU) (General) The central processing unit (CPU) is the computing FIGURE C.21 Shape of survival curve for mammalian cells exposed to radiation. The fraction of cells surviving is plotted on a logarithmic scale part of a computer system, also called processor. Its main com- against the radiation dose on a linear scale. ponents are the arithmetic logic unit (ALU) and a control unit. Usually, the CPU is contained in a single chip. Today, all CPUs are microprocessors (MPU) – a CPU miniaturised in a single Cell survival curve chip. Some medical imaging equipment has a special imaging (Radiotherapy) A plot of surviving fraction against radiation dose processor (Imager), which is an additional CPU used specifically is called a cell survival curve and is usually presented in the form for the acquisition and processing of images. shown in Figure C.21 for x- and γ-rays, with dose plotted on a Hyperlink: www .pcmag .com /encyclopedia _term/ linear scale and surviving fraction on a logarithmic scale. A cell’s radiosensitivity can be determined from cell survival Central ray of x-ray beam curves. The shape of survival curves is commonly described by (Diagnostic Radiology) The central ray of an x-ray beam (also the linear-quadratic model with the ratio of its two parameters, α known as central beam) is the ray with maximal intensity focussed and β, often used as a measure of a cell’s radiosensitivity. at the centre of the detector (x-ray beam). Normally, this ray is For further information, see the article on Linear-quadratic perpendicular to the cathode–anode line. (LQ) model. See the article on Anode heel effect (Figure A.51). Related Articles: Alpha–beta ratio, Cell survival, Dose Related Article: Anode heel effect response model, Linear-quadratic (LQ) model, Probability of cell survival, Radiosensitivity Central volume theorem (Nuclear Medicine) The central volume theorem is used to calcu- CEMA (Converted energy per unit mass) late the cerebral flood flow (CBF) from the relation CBF equals (Radiation Protection) See Converted energy per unit mass the cerebral blood volume (CBV) divided by the mean transit time (CEMA) (MTT), or Centigray (cGy) CBV (Radiation Protection) The SI unit of absorbed dose is the Gray CBF = MTT (Gy). It replaced the rad as the unit of absorbed dose. The equiva- lence between the two quantities is The MTT is the average time the radiopharmaceutical takes to pass through the tissue being studied. Cerebral blood volume 1Gy = 100rad (CBV) and blood flow (CBF) are parameters derived from the first Centigray, i.e. 1/100 Gy (0.01 Gy) is not a normal quantity used pass of the radiopharmaceutical. in the SI system. However, it has been adopted for use in radio- therapy because it is numerically equivalent to the rad: Centre of rotation (see p. 211 for clinical application) (Nuclear Medicine) The centre of rotation (COR) is a stationary 1cGy = 1rad point at the centre of a tomographic imaging system. In a SPECT imaging system, this point correlates to the midpoint of the detec- The adoption of this unit in radiotherapy was made to avoid any tor’s circular orbit around the patient. COR should coincide with confusion in dose

prescription. the mechanical centre of rotation; any deviations could induce Related Articles: Absorbed dose, Gray, Radiation absorbed image artefacts, for example ring artefacts or additional blurring. dose (rad) In a PET system, the COR is the same as the midpoint of the detector circle. A part of the camera system routine quality con- Central beam trol is measuring and correcting for any deviation of COR from (Diagnostic Radiology) See Central ray of x-ray beam the mechanical centre of rotation. Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Central field of view (CFOV) Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, (Nuclear Medicine) An image system is able to depict an area or Philadelphia, PA, pp. 321–322. a volume of a certain size. This size is referred to as the image Related Articles: PET, SPECT Surviving fraction Centric sampling 153 Cephalometric radiography Centric sampling craniostat) used to guarantee exact positioning of the head (Magnetic Resonance) Centric sampling Figure C.22 refers to the (Figure C.23). order in which lines of k-space are acquired. The centric sampling Although also posterior-anterior and axial views are men- scheme can be applied in the phase-encoding direction for 2D tioned in literature, the lateral radiography is predominately used. pulse sequences and in the phase- or slice-encoding direction for The radiogram is required to perform cephalometric analysis 3D pulse sequences. For 3D, although many variants exist, the (‘cephalo’ means ‘head’, ‘metric’ means ‘measurement’). filling of k-space starts at the centre (kz = 0), typically followed The landmarks can be joined by lines to form axes, vectors, by adjacent lines in an interleaved manner (kz = −1, 1, −2, 2…) as angles and planes. The basic elements of analysis are angles shown in Figure C.22 (slice encoding example). and distances; the measurements obtained on the radiogram are Centric sampling is applied when the contrast is changing dur- compared with reference values for the standard patient. As an ing the k-space loop, so that conventional linear sampling would example – on the skull image, several important cephalometric C result in a loss of contrast and/or SNR. landmarks can be identified (for example, S is the midpoint of Different sampling patterns can be used in contrast-enhanced sella turcica) (Figure C.24). angiography applications in order to control the sampling of the There are a number of specific methods for analysing the data k-space relative to the arterial (early) and venous (late) phases of from cephalometric radiography. Nowadays, commercial soft- the contrast agent bolus. Centric sampling results in early sam- ware to realise cephalometric tracings is available. pling of the k-space centre, which means that the image contrast Cephalometric analysis provides clinical indications in ortho- is highly dependent on the amount and location of contrast agent dontics and orthognathic surgery. In orthodontics, it guarantees at the beginning of image acquisition. Sampling of k-space in a support for planning of orthodontic treatments and monitoring combination with the passage of the contrast agent also acts as treatment progress. In orthognathic surgery, it is used in preopera- a low-pass filter, which gives some blurring of the image of the tive evaluation of skeletal and soft tissue patterns and to assist in vessels along different directions depending on the sampling treatment planning, postoperative appraisal of the results of sur- pattern. gery and long-term follow-up studies. When a 3D sequence with two phase-encoding directions (ky Cephalometric radiography is sometimes described by the and kz) is used for contrast-enhanced angiography, only the outer jargon phrase ‘ceph x-ray’. Some of the skull measurements can slower loop (corresponding to TR) is normally considered for be obtained also with a panoramic radiograph (a.k.a. ortho pan centric sampling. Centric sampling of the faster inner loop can tomography [OPG], sometimes described by the jargon phrase in some applications be used to control fat saturation. It is also ‘panorex x-ray’). For this reason, sometimes the OPG method is increasingly used in T1-w magnetisation-prepared 3D sequences. also called cephalometric radiography. Related Articles: Bolus, Contrast-enhanced angiography, Related Articles: Ortho pan tomography (OPG) k-space Hyperlinks: https :/ /en .wiki pedia .org/ wiki/ Cepha lomet ric _ a nalys is Cephalometric radiography (Diagnostic Radiology) Cephalometric radiography is a radio- graphic technique of the skull used by dentists, orthodontics and oral and maxillofacial surgeons to obtain standardised and repro- ducible 2D images of the craniofacial region. The technique is used to measure specific sizes, distances, etc., in the head of the patient. A cephalometric x-ray system (often available in conjunc- tion with a panoramic x-ray system and in some cases also with a dental CBCT scanner), includes an x-ray source, an image receptor and a head-positioning device (called cephalostat or kz (slice) kz = 2 kz = 1 t = 0, kz = 0 kz = –1 kz = –2 kx (read) ky (pahse) FIGURE C.22 k-space sampling pattern for a 3D pulse sequence with centric sampling in the slice direction. The ky–kz positions are sampled consecutively with a time interval corresponding to the rep- FIGURE C.23 An example of panoramic x-ray system and cephalomet- etition time (TR). ric x-ray system. Ceramic capacitor 154 Ceramics C FIGURE C.24 Cephalometric landmarks. Ceramic capacitor cables for the anode and cathode (both at one side of the tube) (General) See Capacitor and a third one at the anode stem behind the anode (under the rotor winding). Ceramic x-ray tube with double bearings Related Articles: Anode, Rotating anode, Metal x-ray tube, (Diagnostic Radiology) The first ceramic x-ray tube with dou- Bearing ble bearings was developed by Philips (Super-Rotalix-Ceramic tube). This rotating anode tube has an unusual design – the Ceramics anode disc is supported on both sides (Figure C.25). This way (General) The word ceramic may be used as a noun or an the anode stem has two bearings (one at each end), which makes adjective, referring to inorganic and non-metallic materials the rotation more stable (in comparison with ordinary x-ray formed by the solidification of a molten substance with the tubes where there is only one bearing at the anode stem – under action of heat. Ceramic materials tend to be hard, porous and the rotor, behind the anode). The x-ray tube with double bear- brittle, and they have a crystalline or amorphous structure. ings is very powerful – i.e. the anode temperature is high and Non-crystalline ceramics, such as glasses, tend to be formed good cooling is necessary. This requires a metal tube housing from melts and are shaped by casting or blowing. Crystalline with internal ceramic coating. This tube has three high-voltage ceramic materials are made by reaction in situ, such as cement ceramic insulators (aluminium oxide) – two for the high-voltage and concrete, or by sintering shaped powders to form a solid body, such as pottery. Ceramics can be categorised as structural (e.g. bricks and roof tiles), refractory (e.g. kiln linings and gas fire radiants), whiteware (e.g. sanitary ware, tableware and wall tiles) and technical (e.g. gas burner nozzles, biomedical implants and jet engine turbine blades). Technical ceramics can be further classified into the fol- lowing: oxides, such as alumina; non-oxides, such as silicides; and composites, which are particulate reinforced combinations of oxides and non-oxides. Rotor Ceramics, such as transition metal oxides, can be semiconduc- Anode tors, which are useful for surge protection applications and gas sensors. Some ceramics display superconductivity at extremely low temperatures or ‘high’ temperatures (>30 K). Other ceram- ics, such as quartz, exhibit piezoelectricity, the generation of an electric potential when under mechanical stress. Piezoelectric ceramics are used for accurate measurements, as crystal oscilla- tors in electric circuits of watches and computers, and as sensors including ultrasound transducers. Piezoelectric materials can also be pyroelectric, meaning that they produce an electric potential with changes in temperature, for use as motion sensors. In addi- tion, pyroelectric materials can also be ferroelectric, meaning that an electric dipole can be oriented or reversed by applying an elec- trostatic field. This property can be used to store information in ferroelectric capacitors. Cathode Ceramics have a wide range of applications due to their spe- cific properties. They are used for knife blades as they remain FIGURE C.25 Schematic cross section of ceramic x-ray tube with dou- sharper for longer than steel blades; however, they are more ble bearings. (Courtesy of Philips Super-Rotalix-Ceramic tube.) likely to break because they are brittle. Some ceramics, such Cerebral blood flow (CBF) 155 Cerebral blood volume (CBV) as alumina, are used in armoured vests and military airplanes, function R(t) (i.e. the fraction of tracer that remains in the tissue due to their low weight. Ceramic balls often replace steel in ball at a time t following an instantaneous arterial bolus input) to the bearings as they are harder, experience less deformation, build initial (or maximal) value of R(t): up less heat, are more chemically resistant and electrically insu- lating. However, the main disadvantage of ceramics is that they ¥ are more expensive. Ceramics can be used in engines, as they do not require a cooling system, leading to greater fuel efficiency ò R (t )dt MTT = 0 max é although they are not widely used in engines as precision and ëR (t )ùû durability are difficult to achieve and the high degree of imperfec- tions lead to a high risk of equipment failure. Ceramics are also Hence, the CBF can be calculated as used for producing watch cases, as they are lightweight, scratch- C resistant and durable. CBV ×max éR Medical Applications: Ceramics are widely used in medicine, (t )ù CBF = ë û ¥ for example bio-ceramics for dental implants and synthetic bones. ò R (t )dt Hydroxyapatite, the mineral in bone, can be made synthetically 0 and formed into ceramic materials. They are used to make ortho- It should be noted that an instantaneous arterial bolus input is typ- paedic implants that bond readily to body tissues without rejec- ically not achievable in practical tracer experiments. Normally, tion or inflammation. Such ceramics are often porous, lacking an approximation to the true arterial input function (AIF) is mea- strength, and so usually coat metal orthopaedic implants or are sured and R(t) is retrieved by deconvolution. used as bone fillers. They are being developed for gene delivery Related Articles: Perfusion imaging, Dynamic susceptibil- and tissue engineering scaffolds. ity contrast MRI, Arterial spin labelling, Arterial input function Ceramic piezoelectric sensors, typically lead zirconate titanate (AIF), Cerebral blood volume (CBV), Mean transit time (MTT) (PZT), are used in ultrasonic transducers for medical imaging. Further Readings: Leenders, K. L. et al. 1990. Cerebral blood Ultrasound transducers are also used in physiotherapy, lithotripsy flow, blood volume and oxygen utilization. Normal values and to obliterate kidney stones and HIFU (high-intensity focused effect of age. Brain 113:27–47; Meier, P. and K. L. Zierler. 1954. ultrasound) for destruction of pathogenic tissue. On the theory of the indicator-dilution method for measurement Related Articles: Quartz, High-intensity focused ultrasound of blood flow and volume. J. Appl. Physiol. 6:731–744; Zierler, K. (HIFU), PZT, Transducer L. 1962. Theoretical basis of indicator-dilation methods for mea- suring flow and volume. Circ. Res. 10:393–407. Cerebral blood flow (CBF) (Magnetic Resonance) The cerebral blood flow (CBF) is the rate Cerebral blood volume (CBV) at which blood passes through the capillary network in brain (Magnetic Resonance) The cerebral blood volume (CBV) is the tissue, typically expressed as volume of blood per unit of time. volume of circulating blood in the brain tissue. The blood vol- Traditionally, this quantity is also known as blood perfusion. ume is the sum of the plasma volume and erythrocyte volume. A common unit for regional CBF (often denoted rCBF) is mL/ The regional CBV (sometimes denoted rCBV) is often given in (min 100 g), i.e. millilitres of blood per minute per 100 g of tis- units of mL/100 g, i.e. millilitres of blood per 100 g of tissue. sue. Through exchange processes, taking place at the interface According to Leenders et al. (1990), the regional CBV in normal between the capillary system and tissue, the blood carries oxygen grey matter is on the order of 5–6 mL/100 g while the regional and nutrients to the brain tissue and removes carbon dioxide and CBV in normal white matter is on the order of 2–3 mL/100 g. other waste products. Hence, the CBF is extremely important for The determination of the rCBV is based

on the principles of maintaining tissue viability. According to Leenders et al. (1990), tracer kinetics for non-diffusible tracers (Meier and Zierler, 1954; the CBF of normal grey matter is on the order of 50–60 mL/(min Zierler, 1962) and relies on the assumption that in the presence of 100 g) and the CBF of normal white matter is on the order of an intact blood brain barrier (BBB), the contrast material remains 20–25 mL/(min. 100 g). MRI-based methods for assessment of intravascular. When an intravascular plasma tracer is employed, CBF are dynamic susceptibility contrast MRI (DSC-MRI) and as in dynamic susceptibility contrast MRI, regional CBV can be arterial spin labelling (ASL). derived from the time integral of the tissue concentration of con- The central volume theorem states that the CBF is related trast agent (C) in combination with the time integral of the cor- to the cerebral blood volume (CBV) and the mean transit time responding arterial concentration of contrast agent (C (MTT) according to the following expression: art), i.e. the arterial input function (AIF): CBV CBF = ¥ ¥ MTT (1- Hlarge ) C t dt 1 Hlarge ò ( ) ( - ) CBV = 0 ò C (t )dt 0 ¥ = ¥ The determination of the CBV and CBF is based on the princi- r(1- Hsmall ) ples of tracer kinetics for non-diffusible tracers Meier and Zierler ò Cart (t )dt r(1- Hsmall ) F 0 ò AI (t )dt 0 (1954), Zierler (1962) and relies on the assumption that in the presence of an intact BBB, the contrast material remains intra- In order to obtain whole-blood volume in units of mL/100 g, the vascular. When such an intravascular tracer is employed (as in correction factor [1 − Hlarge]/[ρ(1 − Hsmall)] is introduced, where DSC-MRI), the CBV can be derived from the time integral of the Hlarge and Hsmall are the haematocrit values in large and small ves- tissue concentration-versus-time curve, normalised to the time sels, respectively, and ρ is the brain density. integral of the corresponding arterial concentration-versus-time Finally, the central volume theorem states that the CBV is curve. According to Zierler’s area-to-height relation, the MTT can related to the cerebral blood flow (CBF) and the mean transit time be calculated as the ratio of the time integral of the tissue residue (MTT) according to the relationship CBV = CBF·MTT. Cerebrospinal fluid (CSF) 156 Certificate of conformity Related Articles: Arterial input function (AIF), Dynamic sus- Ĉerenkov radiation ceptibility contrast MRI, Cerebral blood flow (CBF), Mean transit (Radiation Protection) The energy lost when charged parti- time (MTT) cles (normally electrons) are slowed down within a transparent Further Readings: Leenders, K. L. et al. 1990. Cerebral blood absorbing medium is sometimes given off in the form of UV and flow, blood volume and oxygen utilization. Normal values and visible light, mainly at the blue end of the spectrum. This emit- effect of age. Brain 113:27–47; Meier, P. and K. L. Zierler. 1954. ted light is called Ĉerenkov radiation (after the Russian scientist On the theory of the indicator-dilution method for measurement Pavel Ĉerenkov who first characterised it). It is most often seen of blood flow and volume. J. Appl. Physiol. 6:731–744; Zierler, K. in the cooling tanks in nuclear reactor plants and food irradiation L. 1962. Theoretical basis of indicator-dilation methods for mea- facilities. C suring flow and volume. Circ. Res. 10:393–407. Related Article: Ĉerenkov effect Cerebrospinal fluid (CSF) Cerrobend® (General) Cerebrospinal fluid (CSF) is a watery fluid that is con- (Radiotherapy) Frequently, part of the radiation field has to be tinuously produced and absorbed and that flows in the ventricles shielded to avoid irradiating underlying sensitive structures. When within the brain and around the surface of the brain and spinal individual blocks are designed, low-melting-point alloys are used. cord. It has three main functions. It protects the brain and spi- Popular products currently in use are Cerrobend® and Ostalloy®, nal cord from trauma, supplies nutrients to nervous system tissue which are alloys of lead, cadmium, bismuth and tin. They have a and removes waste products from cerebral metabolism. It plays a melting point of 69°C and 70°C, respectively, although they are special role in T2-weighted MRI where fluid is bright compared usually worked at temperatures of 90°C and 95°C, respectively, to both soft tissues and bone and provides significant diagnostic since they are easier to pour and mould at these temperatures. information (Figure C.26). Cerrobend has a density of 9.4 g/cm3 at 20°C (∼85% of lead den- sity); it consists of 50% bismuth, 26.7% lead, 13.3% tin and 10% Ĉerenkov effect cadmium. It is important that the Cerrobend is poured slowly to (Radiation Protection) Charged particle radiation (beta particles, prevent formation of air bubbles. alpha particles) has kinetic energy given by the classical equation: Newer materials such as MCP 96® (melting point 96°C, alloy of lead, bismuth and tin) do not contain cadmium, which has rela- E = ½mv2 tively high toxicity. Related Articles: Low-melting-point alloy, Block design, Although relativity states that the velocity of such particles given Custom blocking by this equation can never be greater than the speed of light in a Further Reading: Podgorsak, E. B. 2003. Review of Radiation vacuum (the constant c), the kinetic energy of such particles emit- Oncology Physics: A Handbook for Teachers and Students, ted due to radioactive decay may imply that their velocity is faster International Atomic Energy Agency, Vienna, Austria. than the speed of light for the medium through which the particles are travelling (related to c by the refractive index of the material). The particles rapidly lose kinetic energy through interactions with Certificate of conformity electric fields around the atoms of the material, until their velocity (General) Document issued by a certifying body (or notified is reduced to sub-light speeds for that material. body) that guarantees the compliance of a certain product with The energy lost in this interaction process is in the form of the specified requirements and particular regulations, usually UV and visible light, mainly at the blue end of the spectrum, regarding safety, and also performance since 2020 in Europe. and is called Ĉerenkov radiation, after the Russian scientist Depending on the medical device risk class, its conformity Pavel Ĉerenkov who first characterised it. It is most often seen with the regulations is declared by apposite bodies by affixing a in the cooling tanks in nuclear reactor plants and food irradiation specific mark or obtaining a specific authorisation (e.g. CE mark facilities. for Europe, or FDA authorisation for the USA). Hereby, more details on the CE marking are given. The CE marking is affixed to the medical device and accom- panying documents, demonstrating that the manufacturer has complied with the applicable directive and regulations. In order to obtain the CE mark there are different steps (shown in Figures C.27 and C.28): 1. Check if the product is a medical device and what type of medical device. 2. Classify the medical device. 3. Meet the requirements specified in the regulations: in order to do so, the manufacturer can ‘respect a certain harmonised standard or some common specifications that can guarantee the respect of all, or part of, the essential requirements; if there is no harmonised stan- dard of reference, the manufacturer has to prove how each essential requirement has been respected’ provid- ing a checklist. 4. Find and follow the right conformity assessment FIGURE C.26 MR brain scan visualising the CSF. procedures (CAP) (see Figure C.27): CAPs allow Certificate of conformity 157 C ertificate of conformity C FIGURE C.27 A diagram illustrating the necessary steps to obtain the CE mark for a medical device. FIGURE C.28 A diagram illustrating the necessary steps to obtain the CE mark for an In vitro diagnostic medical device. manufacturers to demonstrate the conformity of implantable medical devices, some in vitro diagnos- their device with the regulation. The choice of the tic medical devices and medical devices. A quality CAP depends on the classification of the device. In management system (QMS) is required for all devices particular, low-risk products generally allow self- (except Class I non-sterile/non-measuring). Most certification. Higher risk products will require the companies apply EN ISO 13485 to comply with QMS services of a third party, namely an EU notified body. requirements. The latter is an accredited test laboratory based in the 5. Assemble the technical documentation: it contains EU. Its intervention is compulsory for all the active all the important information to support the claims of Certification 158 Characteristic curve compliance with the directive requirements. It must be charge by the electrodes and processed by the electrometer, and available to national surveillance authorities. environmental factors such as the ambient air temperature and 6. Affix the CE Mark: it is affixed to the device and any pressure. document. When a notified body intervenes, its ID The ideal ionisation chamber will have a consistent (or ‘flat’) number must appear below or next to the CE mark. response to the incident radiation irrespective of changes in any 7. Draw up a declaration of conformity: it is a one-page of these parameters. However, in reality, ionisation chambers document in which the manufacturer declares the con- do not have a flat response and will need to be characterised to formity of his product with the essential requirements compensate for variations due to changes in each parameter. Such and the safety and performance requirements. All characterisation may be done by the manufacturer or as part of C devices must have it. the calibration process, comparing the response of the chamber, for example to well-defined radiation quantities and energies at Related Articles: Medical device; Life cycle of equipment precisely measured ambient temperatures and pressures. Further Readings: Medical Device Regulations 2017/745, Related Articles: Ionisation chamber, Proportional counter, In-Vitro Diagnostics Regulations 2017/746, FDA Regulation of Geiger–Müller (GM) counters Medical Devices. Pecchia, L., D. Piaggio, R. Castaldo, L. Radice Further Readings: Bevelacqua, J. J. 2004. Contemporary and N. Pallikarakis. Medical device regulation and assessment: Health Physics. Problems and Solution, Wiley-VCH GmbH & New challenges for biomedical engineers. In: De Moni, E., Co. KGaA, Weinheim, Germany, pp. 219–221; Knoll, G. F. 2000. A. Menciassi and A. Redaelli. eds. Advanced Bioengineering Radiation Detection and Measurement, 3rd edn., John Wiley & Methods, Technologies and Tools in Surgery and Therapy, Patron Sons, Inc., New York, p. 114. Editore, pp. 177–192; Pokvić, L. G., A. Bošnjaković and M. Peklić Vitt. 2020. European Union medical device legislation, vigilance, Characteristic curve and metrology system. In: Clinical Engineering Handbook, (Diagnostic Radiology) A characteristic curve for radiography Academic Press. film is the graph representing the relationship between opti- cal density and exposure. It is also known as a H and D curve Certification named after the Swiss chemist Ferdinand Hurter (1844–1898) and (General) Certification (professional certification or qualifi- English chemist Vero C. Driffield (1848–1915), who developed cation) is a designation earned by a professional to assure his this concept for photographic emulsions (Figure C.29). qualification to perform specific job. Usually, certifications are A specific feature of the characteristic curve is the logarithmic issued by specific professional body/panel, which assesses the scale used for exposure, usually showing the relative exposure. candidate. The main role of such a body is to safeguard the The slope of the curve at each point represents the contrast public interest. For example, the UK Institute of Physics and transfer characteristics of the film. A limitation of film is that its Engineering in Medicine (IPEM) is licensed by the Engineering Council to register Chartered Engineers and by the Science Council to register Chartered Scientists. Usually, a certified pro- D 3.0 max fessional is listed in a Professional Register and can use post- nominal letters (e.g. CSci – Chartered Scientist, used by many Medical Physicists in the UK). Shoulder 2.5 CEST (Chemical exchange saturation imaging) Film contrast (Magnetic Resonance) See Chemical exchange saturation imag- ing (CEST) 50% Exposure 2.0 contrast CFI (colour flow imaging) (Ultrasound) See Colour flow imaging (CFI) 1.5 Chamber response (Radiation Protection) Chamber response refers to the sensitiv- ity of an ionisation chamber (whether it is operating as an ioni- sation chamber, proportional counter or Geiger–Müller counter) 1.0 in detecting the incident radiation, normally measured in terms of the size of the output pulse amplitude from the chamber. A chamber being able to respond to – i.e. to

detect lower radiation doses/dose rates will give a larger pulse amplitude and is said to 0.5 be more sensitive. Toe Chamber response may also refer to the way the displayed Base + fog density reading on an electrometer of the detection of radiation by an ionisation chamber is affected by a number of parameters. These —1 —1 —1 —1 —1 —1 64 32 16 8 4 2 1 2 4 8 16 32 64 include radiation-related factors such as the type, energy and flu- ence rate of the radiation that affect the total charge liberated Relative exposure within the chamber. Other factors affecting the response include chamber-related factors such as the shape, geometry and angular FIGURE C.29 Typical characteristic curve on radiographic film. response of the chamber affecting the collection of the liberated (Courtesy of Sprawls Foundation, www .sprawls .org) Optical density Characteristic function 159 Charge measurement mode contrast transfer is reduced at both low and high exposures. This Molybdenum anode is represented by the toe and shoulder of the curve. characteristic radiation The toe of the curve at the very low end of the exposure scale 17.6 and 19.7 keV represents density not produced by intended exposure. It is the density of the film base plus density, referred to as fog, which can Molybdenum filter 20 keV come from a variety of sources such as unwanted exposure from the environment, long age of unprocessed film and problems with film development (chemical fog). The straighter portion of the curve with a steep slope is the region that produces image contrast. The slope is determined by a combination of film design characteristics and processing C conditions. The film gamma (the maximum slope) and average gradient are two parameters used to specify the overall slope of the curve. The film contrast factor, defined as the difference in 0 5 10 15 20 25 30 35 density for a doubling of exposure, is used to specify the slope at Photon energy (keV) all points along the curve. FIGURE C.31 Sketch of a typical mammographic x-ray spectrum with Characteristic function molybdenum characteristic radiation (and an additional molybdenum (Diagnostic Radiology) See Characteristic curve K-edge filter). (Courtesy of Sprawls Foundation, www .sprawls .org) Characteristic radiation Charge (Diagnostic Radiology) Characteristic radiation is the type of (General) Electric charge is a basic property of matter carried by x-radiation produced by electron transitions between shells in an some elementary particles and occurs in discrete natural units. atom. The process begins when bombarding electrons enter the This charge is either positive (+ve) or negative (−ve) in nature and x-ray target material or anode and expel electrons from the inner cannot be created or destroyed. Objects that have an excess of one shells with the higher binding energies. These electron vacan- type of charge exert a force on one another when in proximity. cies are then filled from the shells with lower binding energies. Objects that are oppositely charged attract each other, and those Because the electrons are moving to a lower energy state, the dif- with the same charge repel each other. The unit of charge in the ference in energy is expelled as an x-ray photon (Figure C.30). SI system is the coulomb (units: m kg s) and the electron has a The energy of the photon is determined by the specific energy negative charge of 1.60217733 × 10−19 C. difference between the electron shells, which is a ‘characteristic’ Further Reading: Resnick, R. and D. Halliday. Physics, Wiley of the specific material. The typical characteristic x-ray spectrum International Edition, New York. consists of just a few specific photon energies. Related Articles: Coulomb, Force electrostatic Characteristic radiation is a significant proportion of the mam- mographic beam spectrum. Figure C.31 shows a sketch of a typical mammographic x-ray spectrum, produced by molybdenum anode Charge deposition effect (note the two lines of its characteristic radiation). Figure C.31 (Radiotherapy) If a beam of ionising radiation passes through a also shows an additional K-edge metal filter (also molybdenum), material, then charged ions can be produced. These can travel which absorbs most of the x-rays with energy above 20 keV. within the material but it may take some time for recombination Related Article: K-edge metal filter to take place and this might result in a build up of static charge. Therefore, a residual charge may be present in the material for Characteristic x-ray some time after the irradiation has ceased, which could influence subsequent dose measurements. This effect has to be taken into (Diagnostic Radiology) See Characteristic radiation account in dosimetry, and for this reason, it is recommended that electron dosimetry is carried out using slabs of tissue equivalent material instead of larger blocks since it is easier to remove resid- Characteristic x-radiation ual charge from slabs. Related Articles: Charge, Electrical charge, Electron beam dosimetry Charge measurement mode K (Radiation Protection) The charge measurement mode is applied in the read-out of a semiconductor detector. The signal (charge) from the detector passes through a charge-sensitive preampli- fier and then through a linear amplifier, integral or differential discriminator (single-channel analyser) to counter or count-rate L meter. Related Articles: Germanium detector, Pulse-height analys- ers (PHAs) for radiation detectors Atom Further Readings: Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, pp. FIGURE C.30 Sketch showing emission of characteristic radiation. 611, 620–622; Lutz, G. 2007. Semiconductor Radiation Detectors: (Courtesy of Sprawls Foundation, www .sprawls .org) Device Physics, Springer-Verlag, Berlin, Germany, p. 190. Relative exposure Charge-coupled device (CCD) 160 Charged-particle disequilibrium Charge-coupled device (CCD) (Diagnostic Radiology) A charge-coupled device is an analogue shift register that is used in many imaging devices. The device was proposed by Eugene Lally at the Jet Propulsion Laboratory in 1961. The first commercial devices were made in 1974. The devices are made so that light falling on the devices deposits charge in an array of pixels. The pixels are then read out in one of several fashions to produce a digital image. In this process, the data are passed from row to row in the C matrix and then is read out – illustrated in the following four images (Figure C.32). CCDs have replaced analogue traditional video camera tubes in almost all radiology applications and are main sensors in digi- tal photography devices (some modifications being also used in medical imaging). The CCD devices are better, cheaper and less likely to need maintenance (Figure C.33). FIGURE C.33 Image of a CCD chip. Abbreviation: CCD = Charge-coupled device. Hyperlink: en .wi kiped ia .or g /wik i /Cha rge -c oup le d _dev ice Charge-sensitive preamplifier (General) A charge-sensitive preamplifier (or amplifier) has an output that is proportional to the electric charge applied to its input. In radiation sensors, the amount of radiation falling on a sen- sor in a given time produces a proportional build up of electric charge in the sensor. For this to be read out and amplified, the charge can be sensed in two ways: 1. The charge (Q) collects within the sensor and generates Readout a proportional voltage Vc due to its internal capacitance C, where Vc = Q/C. This signal voltage must then be amplified without draining away the charge – requiring an amplifier with both low drift and with exceptionally high input impedance. Such amplifiers are sometimes known as ‘Electrometer amplifiers’. 2. The charge Q is allowed to flow into the amplifier as it is generated, where the flow of charge (a current) may be integrated on a separate capacitor and the resultant voltage amplified and buffered. First row is readout Whichever technique is used, the amplifier must be low-drift, and some additional circuitry is needed to reset the amplifier to a baseline value by effectively discharging the capacitance holding the charge. Charged particles (Radiation Protection) A charged particle is, in physics, either a subatomic particle with an electric charge or an ion. A subatomic particle is one smaller than an atom and there are two broad cat- egories – those that are composite, such as protons, which them- Second row is readout selves consist of smaller particles (quarks), and fundamental particles, which have no substructure. In medical physics, the most common charged particles are electrons (beta particles or negatrons), positrons, protons, alpha particles. Neutrons and neutrinos are subatomic particles but have no charge. Related Articles: Alpha particles, Beta particles Charged-particle disequilibrium (Radiotherapy) If a volume V containing a medium, homoge- Last row is readout neous in density and atomic composition, is uniformly irradiated by photons in any small volume inside V, the number, the energy FIGURE C.32 Sequence of readout of a CCD chip. and the direction of secondary electrons entering and leaving Charged-particle equilibrium 161 C heck source the small volume will be the same if the photon attenuation is Charged-particle equilibrium is not established for thin layers of negligible. Photons first transfer energy to electrons, which in materials irradiated by photons as the maximum range of the sec- turn deposit the energy in the medium and the energy carried ondary electrons produced is larger than the material dimensions in and out of the small volume by secondary electrons is equal. so that part of the electron energy is lost simply by the electrons This condition inside the small volume is indicated as charged- leaking out of the material. particle equilibrium (CPE). In practice, the CPE is not possible Charged-particle equilibrium is also not established in a bulky at any point since the photon fluence is changing through photon block of high atomic number Z material, with a gamma ray source interactions. distributed all through the volume of the block. The block dimen- In regions where the photon fluence varies, causing a dif- sions are large with respect to the range of the Compton electrons ference between the total secondary charged particle energy produced by the gamma rays but small relative to the attenuation entering and exiting the volume, CPE does not exist. Cases of length of the gamma rays. The atomic number of the material is C charged-particle disequilibrium are sufficiently high so that an appreciable amount of energy of the electrons is transformed into Bremsstrahlung radiation, which is 1. Regions in the vicinity of a point radiation source where then lost to the material and escapes from the block. the photon fluence is changing rapidly with distance. Abbreviation: CPE = Charged-particle equilibrium. 2. Regions in close proximity to boundaries between Related Articles: Charged-particle disequilibrium, Collision media of different composition and where the photon kerma, Build-up region, Electron maximum range, Electron prac- energy is high enough that the range of secondary elec- tical range, Percentage depth dose trons is no longer negligible compared to the mean free path of the photons. Charged-particle therapy 3. Regions where there is the presence of inhomogeneous (Radiotherapy) The term charged-particle therapy is generally electric or magnetic field. used to mean therapy with charged hadrons – specifically protons or light ions (rather than using electrons, which are not hadrons). An evident situation of charged particle disequilibrium is in Related Articles: Hadron therapy, Heavy particle beams, Ion the build-up region of a beam where the dose increases before therapy, Proton therapy, Carbon ion therapy decreasing exponentially. The uniformity of the photon fluence is also not satisfied near the edges of a finite beam at distances Chartered scientist between the beam edge and the point under consideration larger (General) A scientist who has obtained such professional qualifi- than the maximum secondary electron range. This particular cation, proving his/her high competence and effective leadership disequilibrium is called lateral disequilibrium and it is obtained in the science field. Chartered scientists use their specific knowl- at the central axis of the beam when the lateral electron range edge of the field and their scientific understanding to advance sci- exceeds the beam radius. This effect is particularly evident in ence and technology. small field size such as in stereotactic radiosurgery. The chartered status, an assurance for fellow scientists, col- A transient charged-particle equilibrium (TCPE) exists at all leagues, employers and customers, is awarded by the Science points within a volume in which the dose D is proportional to col- Council if the scientist meets all the requirements. If a scientist lision kerma Kcoll, while for CPE region D = Kcoll. This situation is becomes a chartered

scientist, he/she becomes entitled to use obtained on the axis of a broad high-energy photon beam where CSci as a title, but he/she needs to meet continuing professional the photon fluence is not uniform because of the geometric diver- requirements each year in order to keep the status. gence and the photon attenuation. This title originated in the United Kingdom, but it is currently Abbreviations: CPE = Charged-particle equilibrium and accepted throughout Europe thanks to Directive 2005/36/EC. TCPE = Transient charged-particle equilibrium. Further Readings: ‘Chartered scientist’. Wikipedia, Related Articles: Charged-particle equilibrium, Collision Wikimedia Foundation, 5 November 2017, en .wi kiped ia .or g /wik kerma, Build-up region, Electron maximum range, Electron prac- i /Cha rtere d _ Sci entis t; ‘CSci in Europe’. The Science Council, tical range, Percentage depth dose scien cecou ncil. org /s cient ists- scien ce -te chnic ians/ benef its -o f -pro fessi onal- regis trati on /cs ci -in -euro pe/; ‘Chartered Scientist Charged-particle equilibrium (CSci)’. The Science Council, scien cecou ncil. org /s cient ists- scien (Radiotherapy) Charged-particle equilibrium (CPE) is estab- ce -te chnic ians/ which -prof essio nal -a ward- is -ri ght -f or -me /csci /; lished in a volume when the energy carried into the volume by Directive 2005/36/EC. the charged particles is balanced by the energy carried out of the volume by the charged particles. Check source If a volume V containing a medium, homogeneous in density (Radiotherapy) A radioactive check source is recommended and atomic composition, is uniformly irradiated by photons in for operational and constancy checks of the dosimetry systems any small volume inside V, the number, the energy and the direc- (ionisation chamber and measuring electrometer) used for the tion of secondary electrons entering and leaving the small volume calibration of the radiation beams. The check source permits will be the same if the photon attenuation is negligible. Photons an exposure of the dosimetry system in a fix and reproducible first transfer energy to electrons, which in turn deposit the energy geometry. Generally, the test device contains a beta-emitting 90Sr in the medium and the energy carried in and out of the small /90Y radioactive source because of the radioisotope long half- volume by secondary electrons is equal. This condition inside the life. The typical initial activity of the source is approximately 30 small volume is indicated as CPE. MBq housed in a shielded container. The check source permits When CPE is established, the collision kerma Kc is equal to an exposure of the dosimetry system in a fixed and reproduc- the dose D: ible geometry (Figure C.34 for cylindrical chamber calibration set up). Generally, the test device contains a beta-emitting 90Sr D = Kc (CPE) /90Y radioactive source because of the radioisotope long half-life. Chemical dosimetry 162 C hemical dosimetry The typical initial activity of the source is approximately 30 MBq radioactive decay of the check source using the 28.78 year half- housed in a shielded container. The check sources are available for life of 90Sr. If the measured exposure value is within ±0.5% of the cylindrical and parallel-plane ionisation chambers (Figure C.35). expected or standard value, the ionisation chamber and electrom- The ionisation chamber and electrometer performances are eter combination is considered reliable (Figure C.36). checked by using a stopwatch to determine equal exposure in the An alternative method for operational and constancy checks of isotope device. The average reading of the dosimetry system is the dosimetry systems is to use 60Co equipment, when available, corrected for temperature and barometric pressure and then com- performing the exposure measurements, in air or in a phantom, pared with an expected value that is the reading obtained on occa- at an accurately reproducible distance from the source in a fixed sion of the calibration of the dosimetry system corrected for the irradiation geometry. C Chemical dosimetry (Radiotherapy) The objective of chemical dosimetry is to deter- Ionisation mine the absorbed dose in an irradiated material from the chemi- chamber cal changes produced by the radiation. Any system, in which a well-characterised and measurable change takes place in a chemical property, may in principle be used as a chemical dosim- Shielding eter. However, the chemical system, as with any other type of dosimeter, must meet a number of requirements and therefore, even if a great number of chemical systems have been studied, it is possible to characterise a single universal chemical system. 90Sr 90Sr Solid, liquid and gaseous systems have been introduced and their source source application depends on the specific studied problem. Examples of solid dosimeter are plastic film dosimeters, systems incorpo- rating organic dyes, alanine dosimeters and photographic films. Most liquid dosimeters are based on dilute aqueous solutions of various compounds. After the irradiation, the aqueous chemical dosimeters can be analysed by titration or light absorption. One of FIGURE C.34 Diagram of the Sr90 check source for a cylindrical ionisa- the most studied liquid chemical systems is the Fricke dosimeter tion chamber. in which ferrous ions in a sulphate solution are oxidised by the action of radiation. Other liquid dosimeters are based on ceric sulphate, oxalic acid or on a combination of ferrous sulphate and cupric sulphate. A recent advancement of chemical dosimetry is the development of gel dosimetric systems in which the nuclear magnetic resonance (NMR) relaxation properties of irradiated gels infused with conventional Fricke dosimetry solutions are used to determine the Fricke gel absorbed dose. Fe2+ ions dis- persed throughout the gel matrix are converted by radiation to Fe3+ ions, with a corresponding change in the paramagnetic prop- erties that may be quantified using NMR relaxation measurements or optical measurement techniques. Because of the limitations of those dosimeters related with the ion diffusion in the gel, alter- native polymer gel dosimeters have been proposed. Polymer gel monomers, which are usually dispersed in an aqueous gel matrix, undergo a polymerisation reaction as a function of absorbed dose. The formation of polymers induced by the radiation influences the NMR relaxation properties and results in other physical changes that may be used to quantify the absorbed dose. As the polymeri- sation is inhibited by oxygen, all free oxygen has to be removed FIGURE C.35 Radioactive check sources for cylindrical and parallel- from these gels and this makes their use troublesome. Recently, plane ionisation chambers. a new polymer gel dosimeter formulation has been introduced in +0.5 % + + + + + + + + + + + + + + + + + + + + + –0.5% FIGURE C.36 Chronological plot of the results of the checks. Standard value Chemical exchange 163 C hemical impurity which oxygen is bound in a metallo-organic complex, thus remov- ing the problem of the oxygen inhibition. Quantitative techniques for measuring dose distributions in these dosimeters include also optical and x-ray computer tomographies, vibrational spectros- copy and ultrasound. Abbreviation: NMR = Nuclear magnetic resonance. Chemical exchange (Radiation Protection) Chemical exchange with ionic resins (ion- exchange method) is used in nuclear medicine for the removal FIGURE C.38 Image in an acutely ischemic rat brain showing a reduc- tion in cerebral blood flow (left), and change in the apparent diffusion C of some elements from the mixture (e.g. radioactive waste). A sample of radioactive solution passes through ion-exchange resin. coefficient (centre), and a reduction in pH as measured by CEST (right). As a result, the purified solution is separated from the radioactive waste (Figure C.37). There are two kinds of ion-exchange resins: characteristics of this spectrum can be displayed as an image. CEST imaging can be done to visualise endogenous macromol- Cation-exchange: R-H + Na+ → R-Na + H+ ecules, or can be done with exogenous contrast agents (usually Anion-exchange: R-OH + Cl− → R-Cl + OH− paramagnetic agents) specifically added to enhance the CEST effect. Monitoring tissue pH and detection of glucosaminogly- cans (gagCEST) are two exciting applications of CEST imaging The chemical exchange with ionic resins (ion-exchange method) in vivo (Figure C.38). is very useful in radiochemistry, nuclear medicine and in the Related Articles: Magnetisation transfer contrast (MTC), Off- removal of some elements from the mixture. resonance, Radiofrequency (RF) pulse Related Articles: Radioactive materials, Radioactive sources, Further Readings: Ward, K. M., A. H. Aletras and R. S. Radioactive waste Balaban. 2000. A new class of contrast agents for MRI based on Further Reading: Saha, G. P. 2001. Physics and Radiobiology proton chemical exchange dependent saturation transfer (CEST). of Nuclear Medicine, 2nd edn., Springer-Verlag, New York. J. Magn. Reson. 143:79–87; van Zijl, P. C. and N. N. Yadav. 2011. Chemical exchange saturation transfer (CEST): What is in a name Chemical exchange saturation transfer (CEST) and what isn’t? Magn. Reson. Med. 65(4):927–948. (Magnetic Resonance) Chemical exchange saturation transfer (CEST) is a technique where RF energy is transmitted at fre- quencies different from the water resonance. Certain macromol- Chemical impurity ecules can absorb this off-resonance RF energy and transfer it (Nuclear Medicine) Chemical impurity is the amount of unde- through chemical exchange to the bulk water protons. The effect sirable chemical constituents in a radiopharmaceutical. This of this transfer is an attenuation on the water signal. This transfer measurement is important to ensure that the presence of any effect allows CEST to image low concentration macromolecules chemical material with potential physiologic, pharmacologic or by observing the degree of signal suppression on the much larger toxic effects is at or below an acceptable limit. The testing should water signal, thereby enhancing imaging sensitivity to these be performed on any chemical substance that is used or formed macromolecules. The degree of signal suppression seen is related during the synthesis or production of radiopharmaceuticals, for to the RF power applied, the RF offset frequency, the macro- example chemical impurities, unlabelled components, reagents molecular concentration and the macromolecule-water interac- and by-products. tion rates (which are indicative of the type of macromolecule). An example of a chemical impurity significant for nuclear Often the RF pre-pulses are varied over a range of frequencies medicine pharmacy is the aluminium ion, Al3+, that may be pres- and the amount of signal reduction at each offset frequency is ent in the elute from a 99Mo → 99mTc generator. Measurement of calculated relative to the signal without the off-resonance RF Al3+ concentration is usually performed by the use of a colouri- pre-pulses. This allows a CEST spectrum to be generated and metric spot test and may be carried out regularly. Another example is trace metals such as Fe in 111InCl solu- tions, which can significantly reduce the labelling efficiency of 111 Radioactive solution In – biomolecules and blood cells. A third example is carrier iodine in sodium–radioiodine solutions, which can compete with the radioactive iodine dur- ing iodination processes, resulting in poor labelling efficiency or interference with the uptake in the thyroid of tracer amounts of Radioactive waste radioiodide. Membrane Related Articles: GMP, Quality control, Radionuclide purity, Radiochemical purity, Biological purity Further Readings: European Pharmacopeia, European Directorate for the Quality of Medicines (EDQM), Council of Europe [http: / /www .edqm .eu /s ite /H omepa ge -62 8 .htm l, accessed on 31 July 2012]; Kowalsky, R. J. and S. W. Falen. 2004. Radiopharmaceuticals Purified solution in Nuclear Pharmacy and Nuclear Medicine, 2nd edn., American Pharmacists Association, Washington, DC; Zolle, I. ed. 2007. FIGURE C.37 Example of a chemical exchange with use of an ion- Technetium-99m Pharmaceuticals – Preparation and Quality exchange resin (membrane). Control in Nuclear Medicine, Springer, Heidelberg, Germany. Chemical quenching 164 Chemical shift artefact Chemical quenching By applying a frequency-selective 90° RF pulse and afterwards (Radiation Protection) Quenching in liquid scintillation detectors applying spoiler gradients to destroy the phase coherence of the causes the reduction of counting efficiency as a result of diminish- signal, the magnetisation originating from the nuclei with the ing the amplitude and number of scintillations. There are several selected frequency is effectively reduced (saturated). Afterwards, types of quenching: chemical, colour, dilution and optical. The the imaging sequence follows immediately and thus the images chemical quenching is caused by chemical composition of the will only show signal from the non-saturated nuclei. The disad- radioactive sample added to the liquid scintillation solution for vantage of the technique is that very homogeneous magnetic B0 counting. The chemical composition of the sample can change and B1 fields are required. either the energy transfer from the detected radiation

to the scin- When the technique is applied to saturate the fat signal, it is C tillating solution or the transfer of light from the scintillating solu- referred to as FATSAT. tion to the photomultiplier (PM) tube. Related Articles: Chemical shift, Fat saturation (FATSAT), Quenching should be taken into account to obtain accu- RF pulse rate counting of samples. There are three methods used for this purpose: Chemical shift (Magnetic Resonance) The chemical shift is generally defined 1. Internal standard method – sample is measured first, independent of field strength as δ in parts per million (ppm): then a known amount of the same radioisotope is added and the sample is measured again. - w 2. Channels ratio method – two pulse-height analysers w d = 0 ×106 w0 (PHA) are applied to describe the shape of the quench curve. The ratio of the counts (Figure C.39) where ω0 is the Larmor frequency of a reference substance ω is the resonant frequency of the line of interest C R = 1 C2 The chemical shift is due to the charged electron cloud that cir- culates around the nucleus, inducing a small but finite local mag- will decrease with an increase in quenching. netic field that opposes the applied magnetic field. The nucleus 3. External standards method – first the sample is mea- is said to be shielded and the extent of the shielding is given by sured and then with Cs-137 or Am-241 source, placed the shielding constant (or more general, the rotationally averaged close to the sample. The correction factor is calculated chemical shift tensor) σ. The effective field at the nucleus is gener- as the ratio of both counts. ally less than the applied field by a fraction σ; thus, the effective field Beff at the nucleus is Further Readings: Saha, G. P. 2001. Physics and Radiobiology of Nuclear Medicine, Springer, New York; Thornton, S. T. and Beff = B0 (1- s) A. Rex. 2000. Modern Physics for Scientists and Engineering, Saunders College Publishing, Philadelphia, PA. σ is related to the Larmor frequency ω0 as follows: Chemical selective saturation (CHEMSAT) g (Magnetic Resonance) This technique to reduce unwanted sig- w0 = B0 (1- s) 2p nals relies on the difference in the Larmor frequencies between nuclei (e.g. hydrogen) in different compounds, one example being and to chemical shift as follows: the so-called chemical shift between hydrogen in water and fat. d = 106 (sref - ssample ) The frequency at which a given nucleus comes to resonance depends therefore not only on the strength of the applied field and C1 the gyromagnetic ratio of the nucleus but also on the molecular environment of the nucleus. Therefore, for a given nucleus, it is C2 observed that the atoms in different chemical environments give MR signals at different characteristic value of the applied field. Related Article: MRS Chemical shift artefact (Magnetic Resonance) Chemical shift is caused by the inherent differences in the chemical environment of protons in water mole- cules and those in fat molecules. When subject to a magnetic field, protons in a fat molecule will have a lower resonant frequency than those in water molecules due to its molecular environment. MR systems are most frequently tuned to the resonant frequency of Pulse height water. When frequency encoding is applied to tissue, it causes the fat signal to be shifted a number of voxels from its actual position FIGURE C.39 Example of the channels ratio method used to eliminate in the frequency-encoding direction. The misregistered fat signal chemical quenching. appears in the form of light or dark bands across the image. Relative intensity Chemical shift imaging (CSI) 165 Chemotaxis targeting The amount of chemical shift artefact experienced for a partic- CSI is a technique with a long history, dating back to the 1970s ular image depends mainly on the receive bandwidth used. There before the development of 2D Fourier MRI. However, hardware is a balance between increasing bandwidth to minimise chemical and pulse sequence limitations precluded widespread implemen- shift effects and reducing signal to noise ratio as more noise is tation on clinical MRI systems until the 1990s. Not least among included with the desired signal. the requirements, given the prevalence of proton (1H) spectros- The chemical shift of the magnet depends on the strength of copy, is excellent static field uniformity, allowing water suppres- the static magnetic field, B0. The chemical shift between water sion over a large volume. This necessitates shimming prior to and fat for a 1.5 T magnet is 3.5 ppm. This typically translates to each examination and effectively limits applications of the tech- a shift on the order of a few voxels. nique to a few regions of the body, such as the brain and prostate. Another type of chemical artefact, called the phase cancella- This problem is less serious for CSI using nuclei other than hydro- tion artefact, exists in voxels containing both fat and water. This gen. Because time constraints limit the number of phase-encod- C appears as a black outline on the image between the fat and water ing steps that can be performed in CSI, data sampling is sparse boundaries. It is caused by the fat and water dephasing after the and the point spread function correspondingly poor. This leads to initial excitation due to their difference in resonant frequencies. To ‘Fourier bleed’ of signal between voxels. A particularly serious reduce this artefact, in-phase spin-echo pulse sequences are used. manifestation of this in brain CSI is bleed of intense lipid signals Abbreviation: ppm = Parts per million. from the margin of the skull into voxels within the brain, where Related Article: Frequency encoding they can obscure metabolite signals of interest. For this reason, CSI is often combined with spatial presaturation bands angled so Chemical shift imaging (CSI) as to eliminate this lipid signal. (Magnetic Resonance) Chemical shift imaging (CSI), otherwise CSI data can be displayed in the form of arrays of spectra, but known as magnetic resonance spectroscopic imaging (MRSI), it is often helpful to extract the signal intensity of each resonance combines spatial encoding techniques drawn from MRI with from each spectrum and display these signal intensities as a metab- magnetic resonance spectroscopy (MRS) in order to produce olite map or in the form of metabolite signal ratios, often using one-, two- or three-dimensional arrays of spectra. colour overlay on a structural MR image. However, while such The advantage of CSI lies in its ability to collect data from mul- images are useful qualitatively, it is better to regard the spectra as tiple voxels simultaneously. Thus, it is not necessary to choose the the base information for quantification since derived parametric optimal volume of interest (VOI) location prior to signal acquisi- maps may be subject to a range of errors (Figures C.41 and C.42). tion, and it is possible to map variations of metabolite signals within Related Articles: Chemical shift, Frequency encoding, organs and between diseased and normal-appearing regions. Magnetic resonance imaging, Magnetic resonance spectroscopy, Conventional CSI uses similar pulse sequences to those Phase encoding, Single voxel spectroscopy employed in MRI, except that frequency encoding is avoided Further Reading: Keevil, S. F. 2006. Spatial localization because imposition of a gradient during signal acquisition would in nuclear magnetic resonance spectroscopy. Phys. Med. Biol. preclude extraction of spectroscopic information using Fourier 51:R579–R636. analysis. Therefore, two-dimensional (2D) CSI uses slice selec- tion followed by phase encoding on two orthogonal axes, and Chemotaxis targeting three-dimensional CSI requires phase encoding on all three axes. (Nuclear Medicine) Chemotaxis refers to the process in which These requirements make CSI a slow and/or low-resolution tech- bodily cells, bacteria or multicellular organisms direct their nique, although adoption of fast imaging techniques developed in the context of MRI has done much to redress this in recent years (Figure C.40). 90° 180° RF Gs Gy Gx Echo FIGURE C.40 Two-dimensional spin-echo CSI pulse sequence. (From FIGURE C.41 Array of proton spectra from the brain. (From Keevil, S. Keevil, S. F., Phys. Med. Biol., 51, R579, 2006.) F., Phys. Med. Biol., 51, R579, 2006.) Chest radiography 166 C hirp Physical contrast in chest Mass Air C Air Calcium Receptor FIGURE C.43 Contrast formation in chest radiography. (Courtesy of Sprawls Foundation, www .sprawls .org) and also produces good penetration and visibility through the ribs (Figure C.44). For film chest radiography, wide latitude (or dynamic range) film is generally used. This reduces the potential of the wide den- sity range in the body producing both areas of over- and underex- FIGURE C.42 NAA/Cr ratio map, showing also outer volume suppres- posure within the image. sion bands. (From Keevil, S. F., Phys. Med. Biol., 51, R579, 2006.) Digital radiography with wide dynamic range receptors and the capability to digitally process the images and apply window- ing reduces the problems produced by the high physical contrast movement according to certain chemicals in their environment. within the chest. A number of new investigations of chest imaging The chemotaxis process is used to study inflammatory processes point out that compared with high kV classical film radiography, in which radiolabelled leukocytes are directed to the inflamma- digital radiography produces better images with lower kV (on the tory area by the chemical products produced there. order of 80 kVp). Related Article: Abdominal imaging Chest radiography Further Readings: Beutel, J., H. L. Kundel, R. L. Van Metter, (Diagnostic Radiology) Chest radiography is one of the most fre- eds. 2000. Handbook of Medical Imaging: Volume 1, Physics quently performed x-ray examinations. Chest radiography pres- and Psychophysics, SPIE Press, Bellingham, WA; Honey, I. D., ents a wide dynamic range from bones to air-containing lungs. A. Mackenzie and D. S. Evans. 2005. Investigation of optimum The patient-to-patient variability (anatomic noise) in this region is energies for chest imaging using filmscreen and computed radi- also significant. The high dynamic range of contemporary digital ography. Br. J. Radiol. 78:422–427. x-ray systems presents far better results than screen-film radiogra- Hyperlink: Sprawls foundation: www .sprawls .org /resources phy, but tomographic methods such as CT and MRI are far more effective in detection of low-density lung lesions laying under Chirp high-density bones (ribs cover ∼2/3 of the lung). Cardiac exami- (Ultrasound) A chirp is a frequency-modulated pulse that can be nations in the region are made with angiographic equipment. used in coded excitation schemes. A simple form of chirp is a lin- Physical Contrast in the Chest: The lungs that are partially ear continuous-wave chirp of the form cos(ωt + γt2/2), where the filled with air form a low-density background that appears dark in the usual radiographic display. Most other structures (ribs, heart, abnormal masses within the lungs, etc.) that are more dense than the lungs appear as bright areas. The chest radiograph is a dis- X-ray beam spectrum for chest play of the variations in physical density that are within the region (Figure C.43). General Imaging Techniques for Chest: Conventional chest radiography is performed with an x-ray spectrum containing Heavy filtration high-energy photons to enhance penetration through the body. A potential problem in chest radiography is that the very high physical contrast within the chest, described earlier, can pro- KV 120 duce a wide range of exposure values to the image receptor that extend beyond the receptor’s range of good contrast transfer. This must be addressed both through the optimisation of the x-ray spectrum and the selection of receptor and display contrast characteristics. 0 20 40 60 80 100 120 An x-ray spectrum that is generally optimum for chest imag- Photon energy (keV) ing is produced with a relatively high KV value and more beam filtration than used for most other radiographic procedures. This FIGURE C.44 X-ray beam filtration for chest radiography. (Courtesy of reduces the high contrast between the low- and high-density areas Sprawls Foundation, www .sprawls .org) Quantity Chi-square test 167 Chromium-51 [51Cr] parameter γ determines the rate at which the frequency changes. HO – CH2 – CH2 –N+ (CH3)3 After passing the received signal through a matched filter (which in this case has an impulse response that essentially is a time- FIGURE C.45 Molecular structure of choline. reversed version of the chirp), a compressed waveform is obtained. However, the resulting compressed waveform is not a desired delta function from a single reflector, but it has a main lobe and side lobes The

3.22 ppm trimethylamine resonance frequently, but some- (here called range side lobes, as they appear in the range direc- what loosely, attributed to Cho in a proton spectrum in fact arises tion). Much effort has been devoted to reduce the range side lobes from a number of related choline-containing compounds: phos- using various windowing techniques. Difficulties in this process- phocholine, glycerophosphocholine and betaine, as well as Cho ing arise because of the frequency-dependent attenuation of ultra- itself. The peak is therefore more properly ascribed to choline- sound in tissue, causing the centre frequency of the returned signal containing compounds, or total choline (tCho). C to decrease with depth, as well as distorting the spectrum. Further, Choline-containing compounds play a role in membrane syn- the time gain compensation will cause the waveform reflected at a thesis, and consequently, the tCho peak is elevated in malignant fixed depth to vary in amplitude. The gain is corrected with time, tumours due to rapid cell division (Figure C.46). and as the waveform is long, the later part of the waveform will Chromium-51 [51 be increased in amplitude. These problems make it unlikely that Cr] a simple filter with sufficiently low-range lobes could be realised. (Nuclear Medicine) Chi-square test Element: Chromium (General) The Chi-square χ2 test is a statistical test useful to Isotopes: 18 < N < 38 determine the presence of a random error other than the expected Atomic number (Z): 24 variation of a Poisson distribution in a set of measurements. Mass number (N): 27 Examples of measurements where this test is useful are in the Symbol: 51Cr determination of faulty instrumentations and the normal variation Production: between patients, animals, etc. The first step of the test is to obtain Usually reactor 50Cr(n, γ)51Cr Daughter: 51 a series of measurements, which should be at least 20. The second V step is to compute the mean value of the measurements: Half-life: 27.7 days Decay mode: EC – decay n Radiation: Gamma, conversion electrons, Auger electrons, N å N = i characteristic x-ray photons n i=1 Gamma energy: 320 keV (10%) Dose rate from 1 0.06 µSv/h at 30 cm (point source); 0.0054 µSv/h at and the quantity: MBq: 1 m (10 mL vial) Absorption (HVL): ≈2 mm lead (note the very low photon yield) n (N N å i )2 - (n -1)SD2 Biological half-life: 8 h (30%), 10–160 days c2 = = N N Critical organ: Liver, plasmaproteins i=1 ALImin (50 mSv): 700 MBq The χ2 square value can be referred to a graph provided in most Absorbed dose: 0.6 mGy/MBq liver statistics text books. Locate the P-value closest to the position with Effective dose: 0.038 mSv/MBq (oral); 0.021 mSv/MBq coordinates n and χ2. The P-value is the probability that random (inhalation) variation from a Poisson distribution would equal or exceed the χ2-value. A P-value close to 0.5 suggests that the calculated χ2- value is in the middle range, expected for a Poisson distribution. A lower P-value indicates that there is a high probability that there is an additional source of error present. A higher P-value indi- 0.8 cates that there is less variation than expected from a Poisson distri- bution but should also be a cause for concern. Typically, a P-value between 0.05 and 0.95 is acceptable. P-values above 0.99 and below 0.6 0.01 indicate that there is likely to be a fault in the measurement – CH3 of tCho, 3.22 ppm system and P-value above 0.95 and below 0.05 indicates that the measurements should be considered inconclusive and repeated. 0.4 χ2 is in itself a statistically variable quantity and should there- fore be treated with caution. The standard deviation for n ~20 is about 25% and for n ~100, it is 15%. If values are close to the 0.2 critical χ2 values, the experiment is recommended to be repeated. Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, 0.0 Philadelphia, PA, p. 387. 4 3 2 1 Choline ppm (Magnetic Resonance) Choline (Cho) is a chemical compound that is detected in in vivo proton (1H) NMR spectra of a number FIGURE C.46 1H NMR spectrum of the human brain showing tCho of organs (Figure C.45). resonance. CIF (contrast improvement factor) 168 C ine MRI Contemporary angiographic systems do not use cine film any- 24 7S3 more, but directly record the dynamic digital image in the system 7/2– 27.7025 d memory. Cr Related Articles: Cineangiography, Cineradiography, 51 Chromium 24Cr Cinefluoroscopy, Angiography 51.9961 EC [Ar]3d54s Cine loop 6.7665 (Diagnostic Radiology) Cine loop is a method of observation of cine films, where a section of the film is repeatedly observed. C The method is common in x-ray angiographic cardio examina- Clinical Applications: 51Cr is labelled to the chelate EDTA tions, where movement of the cardiac muscle has to be exam- useful for glomerular filtration rate measurements (blood sam- ined at a specific period of the cardiac cycle. The method is also ples) in vitro. 51Cr-labelled RBCs have commonly been used for used (often with the same name) in digital radiography, where a measurements of red cell mass, survival studies of RBC and plate- sequence of digital frames is looped and after the end of the last lets and plasma volume. selected image, the first image is again addressed and the whole Further Readings: Annals of the ICRP. 1987. Radiation Dose sequence repeats. to Patients from Radiopharmaceuticals, Biokinetic models and Related Articles: Fluoroscopy, Cinefluoroscopy, Data, ICRP Publication 53, Vol. 18, Pergamon Press, Oxford, Cineangiography, Digital subtraction angiography 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 Cine MRI .lu. se /nu clear data/ toi/]; Firestone, R. B. 1999. Table of Isotopes, (Magnetic Resonance) Cine MRI techniques are designed to 8th edn., Update with CD-ROM [http://ie .lbl .gov /toi .html]; acquire a series of individual MRI image frames and display Kowalsky, R. J. and S. W. Falen 2004. Radiopharmaceuticals in them sequentially as a movie. The primary application of cine Nuclear Pharmacy and Nuclear Medicine, 2nd edn., American MRI is in cardiac imaging where a cine acquisition is used to Pharmacists Association, Washington, DC. display a movie of the heart beating over a single cardiac cycle (see Figure C.47). CIF (contrast improvement factor) The basic building block of any movie is a frame. The dura- (Diagnostic Radiology) See Contrast improvement factor (CIF) tion of a frame determines the temporal resolution of the move. Typical frame durations useful in cardiac cine MRI are on the Cine film order of 50 ms. (Diagnostic Radiology) Cine film is mainly used in cineangiog- In a real-time cine acquisition, each acquired frame corre- raphy to record angiographic x-ray examination with cine camera sponds to a single snapshot of the heart at that instant. Real-time (taking the image from the output of the image intensifier through acquisitions are technically demanding as k-space for each frame special optics). This roll film is mainly with two widths: 35 mm must be completely filled in ∼50 ms, while still achieving ade- (more often) and 16 mm. The wider film covers a larger area and quate resolution and SNR. Through technical developments such is mainly associated with greater dose. The film is processed by as the application of parallel imaging techniques, real-time car- special processing machines and is further viewed by projector diac for routine clinical cardiac applications is likely to become (as movie film). feasible. To date however, most cine MRI acquired in clinical FIGURE C.47 Cine MRI images of short axis of heart. Cineangiography 169 Cineradiography Cineangiography systems used in cardioangiography use syn- chronisation of the imaging with the electrocardiography of the patient. This allows recording at different phases of the cardiac cycle. Contemporary systems do not use cine film, but directly record the digital image in the system memory. Related Articles: Fluoroscopy, Biplane cine system, Cineradio- graphy, Cinefluoroscopy, Angiography, Dual-energy imaging Cinefluoroscopy (Diagnostic Radiology) Cinefluoroscopy is a system/method C allowing the recording of dynamic images from fluoroscopy to FIGURE C.48 In prospective gated cine MRI, the ECG R-wave initiates cine film. During cinefluoroscopy, the x-ray generator usually data acquisition. Data for each frame are acquired at the same phase of the operates in continuous fluoroscopic mode. The cine camera is cardiac cycle in each beat. attached to the output of the image intensifier (II) through special optics. This can be fibre optic type or using lenses (often called tandem optic). The latter uses semi-transparent mirror to split the practice is not real time. While the final movie appears to show light output from the II to two beams – one going to the TV moni- the time course of a single beat, the actual data acquired to gener- tor, the other one to the cine camera (and often a third output to ate each frame of the movie are collected over many heartbeats. spot-film radiography) – see Figure C.49. This type of acquisition is illustrated in Figure C.48. Each The cine film used in such a system is 16–35 mm. As these sys- movie frame has a corresponding k-space that must be filled. In tems are mainly used for cardioangiography, the 35 mm camera a real-time acquisition, each k-space frame would be filled com- and film are preferred as these have better resolution. However, pletely in a single frame duration and the entire movie collected in the patient dose in all cine recording x-ray systems is very high, a single heartbeat. More generally, only a few lines in k-space will as the quick exposure of the film requires higher light intensity, be filled for each frame in every heartbeat. The number of lines hence higher x-ray beam intensity. filled per heartbeat per frame is called the views per segment. The cine camera uses an electronic synchronous motor to Segmented k-space acquisitions allow total scan times that are move the cine film. The number of frames per second is based within the breath-holding ability of most patients – an important on the power supply of the camera. For example, if 60 Hz is used, consideration in reducing movement artefacts. the camera can record with speed up to 60 fps, or, respectively, An ECG trace is acquired during scanning to provide a tempo- 30 or 15 fps. ral reference to the phases of the beating heart. In prospective gat- Video recording of the fluoroscopic image had been used ing, the ECG directly triggers acquisition to ensure that k-space in the past, but its quality is lower than the cine film quality. lines for each movie frame are acquired at the same phase of the Contemporary digital fluoroscopic units (with II or with flat panel cardiac cycle in each heartbeat. In retrospective gating, the MRI detector) record the digital image directly and allow easy image data are acquired continually and rebinned appropriately into manipulation and record. Their quality is rapidly moving to out- k-space frames after the scan by reference to the recorded ECG. perform cine image quality. The main sequence type used in cardiac cine imaging is bal- Related Articles: Fluoroscopy, Biplane cine system, anced FFE (i.e. a true FISP), which provides good inherent blood/ Cineradiography, Cineangiography myocardium contrast. Prior to the availability of balanced FFE sequences, spoiled gradient-echo sequences (e.g. FLASH-type Cineradiography sequences) were used. Blood/myocardium contrast with spoiled (Diagnostic Radiology) Cineradiography is most often used to gradient-echo sequences is achieved through the bright blood record angiographic x-ray examinations. In this method, the x-ray time-of-flight technique. Balanced techniques have the advantage of faster acquisition times and independence of contrast on blood velocity. As such, balanced sequences work at faster frame rates (∼4 ms per k-space line) compared to FLASH techniques (∼8 ms per line), and with slow flow. Further Reading: Lee, V. S. 2006. Cardiovascular MRI, Physical Principles to Practical Protocols, Lippincott Williams & Wilkins, Philadelphia, PA. Related Articles: Balanced FFE, Fast low angle shot (FLASH), Fast imaging with steady-state precession (FISP) 2 1 Cineangiography (Diagnostic Radiology) Cineangiography is the method of record- ing angiographic x-ray examination. As this system records 3 images of blood vessels filled with contrast media (most often iodine-based), the x-ray spectrum has to comply with the absorp- tion K-edge of iodine. Cineangiography can use either the fluoro- scopic or

radiographic mode of operation of the x-ray generator. If the radiographic mode is used, the kV of the x-ray pulses can vary FIGURE C.49 Typical cinefluoroscopic x-ray equipment (C-arm). At above or below the K-edge, thus allowing dual-energy subtraction the end of the image intensifier are mounted on the (1) TV camera, (2) the (hence better contrast resolution). Cine camera and (3) the Cut-film camera. Frame 1 Frame 2 Frame n Frame 1 Frame 2 Frame n Frame 1 Frame 2 Frame n Circuit breaker 170 Circularly polarised (CP) generator operates at radiographic mode, producing short x-ray Circular orbit pulses. These pulses are synchronised with the shutter of the cine (Nuclear Medicine) Circular orbit describes circular detector camera. The whole system can also be synchronised with the elec- motion during a SPECT acquisition. trocardiogram of the patient (in case of cardioangiography – one In a circular detector orbit, the detector will move circularly of the most often used examinations with cineradiography). The around the patient with a radius defined by the anatomy furthest x-ray systems for cineradiography require special x-ray generator. from the centre of rotation (see Figure C.51). During this mode, the camera records static images with speed Circular orbits are satisfactory for SPECT imaging of reason- synchronised with the frequency of the mains. The speed of the ably circular anatomy such as the brain, but cause a loss of spa- image sequence can reach up to 120 fps. The camera recording tial resolution when imaging irregular regions, such as the body, C these sequences can be either classical high-speed cine camera, because the circular orbit will cause the detector to be a large or digital camera (most often with CCD sensor). The patient dose distance from the body surface in certain views. during cineradiography is quite high. For example, a 23 cm image For this reason, a non-circular, body contouring orbit is usu- intensifier would need entrance exposure on the order of 10–15 ally used when performing SPECT of structures within the body. μR/s. Related Articles: Body contouring orbit, Centre of rotation, Cineradiography should not be mistaken with spot-film radiog- Single-photon emission computed tomography (SPECT) raphy of fluoroscopic images, which normally has very low speed Further Reading: Bushberg, J. T., J. A. Siebert, E. M. (4–8 fps) and is used mainly in peripheral (limbs) angiography. Leidholdt and J. M. Boone. 2002.; The Essential Physics of Related Articles: Fluoroscopy, Cinefluoroscopy, Medical Imaging, 2nd edn. Cineangiography Further Reading: Bushberg, J. T., J. A. Seibert, E. M. Circular polarisation Leidholdt and J. M. Boone. 2002. The Essential Physics of (Magnetic Resonance) See Circularly polarised (CP) Medical Imaging, 2nd edn., Lippincott Williams & Wilkins, Philadelphia, PA. Circularly polarised (CP) Circuit breaker (Magnetic Resonance) If the B1 field vector (i.e. the magnetic (General) Circuit breaker is an electrical switch, which interrupts component of the transmitted RF pulse) at a point in space under- the electric current in a circuit when the current becomes too goes constant rotation in a plane, then the field is said to be circu- high, caused by overload or short circuit. Unlike a fuse, a circuit larly polarised at that point. breaker can be reset (either manually or automatically) to resume A circularly polarised B1 field is physically equivalent to the normal operation after it has been tripped. When the current is summation of two orthogonal, linearly polarised fields in phase high enough to operate a trigger device in a breaker, a pair of con- quadrature (Figure C.52): tacts conducting the current are separated by preloaded springs or other mechanism. Generally, a circuit breaker registers the cur- rent either by the current’s heating effect or by the magnetism it creates in passing through a small coil. Circuit breakers are made in varying sizes, from small devices that protect an individual household appliance up to large switchgear designed to protect high-voltage circuits feeding an entire city. Circuit(s), electrical (General) An electrical circuit is a term that refers to any inter- connected network of electrical components. Current can only flow when a path or circuit exists between the two terminals of any power source. A circuit diagram or schematic is used to describe the com- ponents and how they are interconnected. Solid lines represent the interconnection of components using a set of internationally agreed symbols for the commonly occurring devices, and their properties are usually displayed in numbers and units alongside each component (Figure C.50). Related Articles: Resistor, Capacitor, Diode Further Reading: ANSI standard Y32 (also known as IEEE Std 315); IEC 60617 (also known as British Standard BS 3939) FIGURE C.51 Circular orbit. A Battery Switch Resistor Capacitor Inductor AC source Diode Transistor Motor Meter FIGURE C.50 Common symbols used in electrical circuits. Circularly polarised coil (CP coil) 171 Clinical target volume (CTV) y medical devices and medical locations, with clinical engineering practitioners delivering functionalities within or to the healthcare systems. Accordingly, clinical engineers are required to combine z engineering and managerial principles to healthcare technology, x B|| in order to optimise and manage the delivery of healthcare in a safe and efficient way. Saide Calil reports in Miniati et al. (2015) that the term was Direction of coined by the cardiologist Cesar A. Caceres (considered the father propagation B of clinical engineering) in the second half of the twentieth cen- B1 tury, when more complex technology was designed and started populating the healthcare systems, hence the requirement of spe- C cialised personnel. FIGURE C.52 Circularly polarised B1 field. A circularly polarised field Although many common tasks of clinical engineering can be decomposes into two orthogonal, linearly polarised fields in quadrature identified globally, the specific roles of clinical engineers are geo- (denoted B⊥ and B∥ in the diagram). graphically dependent and fast changing. In fact, the most fre- quent tasks worldwide associated with clinical engineering are health technology management and servicing. Education is also a B = B recurrent task among the clinical engineers of the Western world, 1sinwt y + B1 coswt x along with information technology as opposed to the Eastern world, where there is an emphasis on a more proactive role in Related Articles: B1, Linearly polarised, CP coils medical procedures. Conversely, the USA and Canada, followed by Europe, also highlight the importance of the role of clinical Circularly polarised coil (CP coil) engineers in risk management. (Magnetic Resonance) See Quadrature coil The fast evolution of medical devices/locations, the growing importance of healthcare technologies in medicine and the fact Classic (coherent) scattering that hospitals are more and more involved in research and teach- (Radiation Protection) See Coherent scattering ing are reshaping the profile of this profession. In fact, it is more common to see clinical engineers working with innovative tools Classified workers such as artificial intelligence, big data and the internet of things, (Radiation Protection) The term ‘classified workers’ refers to and reducing the gap between clinical engineering practitioners those workers who are employed in work activities involving (historically associated with medical device maintenance and ser- normal exposure to ionising radiation. The classification of such vicing) and bioengineers (historically considered as the branch of workers is based on the evaluation of the individual activity (and biomedical engineering more focused on research and academia). relative risk) and measurements of the work situation. Detailed Related Articles: Standards; Medical equipment manage- working conditions, responsibilities, duties and monitoring ment; Medical device; Maintenance; Risk management arrangements are clearly specified in the national set of radiation Further Readings: Miniati, R., E. Iadanza and F. Dori. protection standards. The limits only refer to work activities and 2015. Clinical Engineering: From Devices to Systems, shall not include medical exposures and background. Academic Press. Definition of Clinical Engineer by American Related Articles: Dose limits, Occupational exposures College of Clinical Engineering (ACCE). Definition of Clinical Engineer by IFMBE CED. Definition of Clinical Engineer Clearance of tracers by Associazione Italiana Ingegneri Clinici (AIIC). Iadanza, (Nuclear Medicine) This entry refers to the clearance of tracers Ernesto. 2020. Clinical engineering. In: Clinical Engineering from blood in the capillaries into tissue. The amount of tracer Handbook. Academic Press. extracted from the blood is the product of the extraction fraction Hyperlinks: https://ced .ifmbe .org/ and the blood flow through the capillaries. The unit is the same as for a flow, i.e. mL/min or mL/min/g. The increase in tracer Clinical equipoise extraction due to an increase in blood flow offsets the decrease (Radiotherapy) Clinical equipoise is defined as a state of uncer- in extraction factor; hence, the net tracer delivery is proportional tainty within the expert medical community about the relative to blood flow. merits of different interventions. In practice this may mean that In tracer kinetic modelling, clearance is also used to describe insufficient evidence has been collected to demonstrate to the the clearance of tracer from tissue to blood. community that one intervention is inferior. It may also mean that Further Reading: Cherry, S. R., J. A. Sorenson and M. E. different experts would prefer different interventions for the same Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, patient. Clinical equipoise is generally seen as an ethically neces- Philadelphia, PA, p.387. sary condition for a randomised clinical trial. Related Articles: Tracer kinetic modelling, Tracer, Tracer Within radiotherapy, examples of randomised clinical trials delivery where the concept of equipoise has been relevant include com- parisons of different fractionation schemes and comparisons of Clinical engineering charged particle therapy with conventional photon radiotherapy. (General) Clinical engineering is a multidisciplinary profes- Related Articles: Fractionation, Charged particle therapy sion utilising science, technology, engineering and mathematics (STEM) knowledge, methods and tools in medicine. Clinical Clinical target volume (CTV) engineering is a branch of biomedical engineering particu- (Radiotherapy) The clinical target volume (CTV) describes the full larly involved in the regulation, management and assessment of extent of malignant growth to be treated with radiotherapy. This Clinical trial endpoints 172 Coaxial cable • Progression-free survival: The time between (a) treat- ment* aimed at controlling or shrinking cancer and (b) signs that the cancer is growing back, or death. • Response rate: The percentage of patients whose cancer shrinks or disappears. • Quality of life: A measure of patient life satisfaction. • Patient-reported outcomes: Health perception scores from validated questionnaires completed by patients Gross tumour volume without physician involvement. C Clinical target volume Planning target volume *timepoints other than treatment, e.g. diagnosis or trial randomi- sation point, may also be considered. Treated volume Related Articles: Quality-adjusted life years (QALYs), Patient- Irradiated volume reported outcome measures (PROMs) FIGURE C.53 Definition of target volumes as in ICRU 50. Clock frequency (General) See Clock pulse includes any subclinical microscopic growth that is not visible with Clock pulse standard imaging techniques. It is often determined by adding a (General) A clock pulse is an electric signal used to coordinate margin to the gross tumour volume (GTV), the margin based on the actions of two or more digital electronic circuits. The clock anatomy and biological considerations. Functional imaging tech- pulse can be either in high or low state. The electronic circuits niques such as PET may add information on the extent of the margin may synchronise at the rising edge, falling edge or both in the needed. The CTV may include regional lymph nodes. Margins will rising and in the falling edges of the clock pulse. Clock pulses are be added to the CTV to create the planning target volume (PTV). continuous, precisely spaced changes in voltage, also known as The use of CTV as a planning volume was proposed by the clock signals. Most often, they are in the form of a square wave ICRU in Report 50 (with addendum 62). This report provides a with a 50% duty cycle, usually with a fixed, constant frequency. common framework on prescribing, recording and reporting therapies, with the aim to improve the consistency and inter-site Clutter comparability. It details the minimum set of data required to be (Ultrasound) Clutter is a term used to describe unwanted echoes able to adequately assess treatments without having to return to from electronic systems. The term is most commonly used not the original centre for extra information (Figure C.53). only in radar applications but also to describe noise in B-mode Abbreviation: CTV = Clinical target volume. and Doppler imaging. Related Articles: ICRU, Clinical target volume (CTV), B-mode clutter has been described as specifically arising from Planning target

volume (PTV), Treated volume, Irradiated volume grating or side lobes, although a more general description might Further Readings: ICRU (International Commission on include electronic noise and extraneous echoes from poor beam Radiation Units and Measurements). 1993. Prescribing, reporting control. It has been described as the background grey level when and recording photon beam therapy. ICRU Report 50, Washington, imaging a fluid-filled target where there should be no direct scat- DC; ICRU (International Commission on Radiation Units and tering or reflection. Measurements). 1999. Prescribing, recording and reporting pho- Clutter in colour flow imaging arises from the high-intensity ton beam therapy (Supplement to ICRU Report 50). ICRU Report signals from the vessel wall and surrounding tissue. These are 62, Washington, DC. usually relatively slow-moving structures and filters are used to remove the high intensity but low frequency shift echoes. Clinical trial endpoints However, these filters will also remove signals from low-velocity (Radiotherapy) Before a clinical trial begins, researchers must blood flow and engineers face challenges in applying and setting define a relevant set of endpoints: metrics by which the success or filters appropriate to specific applications. otherwise of the trial will be measured. Related Articles: Wall filter, Grating lobes, Side lobes A clinical trial’s primary endpoint should describe the infor- Further Reading: Evans, D. H. and W. N. McDicken. 2000. mation that the research team are most interested in finding out Doppler Ultrasound: Physics, Instrumentation and Signal (e.g. overall survival). Secondary endpoints should describe infor- Processing, 2nd edn., Wiley, New York, pp. 246–250. mation that is not the main focus of the trial, but that researchers are still interested in finding out (e.g. the incidence of a specific side effect). Sample size is generally determined as the number of CMUT trial subjects required to detect a clinically meaningful difference (Ultrasound) See Capacitive micromachined ultrasound in the primary endpoint, at a given statistical significance level. transducer. Some common endpoints for clinical trials of cancer treat- ments include: CNR (Contrast to noise ratio) (Diagnostic Radiology) See Contrast to noise ratio (CNR) • Overall survival: A total measure of how long patients live after treatment*. Coaxial cable • Disease-free survival: The time between (a) treatment* (General) A coaxial cable is a cable consisting of two conduc- aimed at curing cancer and (b) signs that the cancer has tors laid concentrically along the same axis. The inner con- returned, or death. ductor is surrounded by a tubular flexible dielectric insulator, Cobalt 173 Coded aperture which is in turn surrounded by fine woven wire, or a thin Cobalt-60 teletherapy units developed in the 1950s provided metallic foil acting as a shield. The whole cable is wrapped in a the first practical megavoltage treatment units. At that time, this protective plastic sheathing. The term coaxial comes from the type of treatment offered a significant step towards high-energy inner conductor and the outer shield sharing the same geomet- treatment and meant that 60Co teletherapy was at the forefront of ric axis. The current in the inner conductor draws the current radiotherapy for many years. The advantages of cobalt units are in the outer conductor towards the centre rather than letting that they have a high-energy gamma ray emission, a long half- it dissipate outwards and thus the associated current flow is life and high specific activity and simple means of production. restricted to the adjacent surfaces of the inner and outer con- Although linear accelerators have taken over in most centres ductors. As a result, coaxial cable has very low radiation losses because of the sophisticated treatment capability, cobalt units are and low susceptibility to external interference and effectively still in use in many places through the world because they are far guides signals with low emission along the length of the cable. simpler, easier to maintain and cheaper to purchase (Figures C.54 C Coaxial cables are often used as a transmission line for radio through C.56). frequency signals. Because they can carry a large number of Related Articles: Caesium, Gamma knife signals simultaneously, coaxial cables are widely used in cable Further Reading: Podgorsak, E. B. 2003. Review of Radiation television systems. Oncology Physics: A Handbook for Teachers and Students, International Atomic Energy Agency, Vienna, Austria. Cobalt (General) Cobalt (Co) is a transition metal with atomic number Coded aperture 27. It is a ferromagnetic metal and found in compounds in nature. (Nuclear Medicine) Pinhole apertures with several holes Cobalt has many isotopes ranging from 50Co to 73Co; 59Co is sta- arranged according to specific binary arrays are collectively ble. The remaining are radioisotopes, of which 60Co and 57Co are called coded apertures. Coded apertures were originally used in commonly used in medical physics. astronomy, but have recently been introduced in nuclear medi- 60Co in Radiotherapy: Cobalt-60 has a radioactive half-life of cine applications, mostly for preclinical imaging. The advan- 5.27 years. It decays into 60Ni with the emission of beta electrons tage over conventional pinholes is an increased efficiency due with a maximum energy of 320 keV, and two gamma rays with to a large open fraction without a loss in spatial resolution. The energies of 1.17 and 1.33 MeV. The electrons are absorbed into the source capsule, while the high-energy gamma rays can be used for radiation therapy. The 5.27 year half-life means that cobalt sources need replacing at regular intervals. 57Co in Nuclear Medicine – Schilling’s Test: Cobalt-57 has a radioactive half-life of 271.8 days. It decays into 57Fe by electron capture, which also emits gamma rays at 122.1 keV. It is used in nuclear medicine for the Schilling test, which can help to deter- mine the uptake of vitamin B12. Related Article: Cobalt unit Cobalt Gray Equivalent (Radiotherapy) In order to simplify dose prescriptions, histori- cally proton therapy doses were sometimes quoted in Cobalt Gray Equivalent (CGE) ‘units’. The standard proton RBE of 1.1 (well described by Paganetti et al. in 2002), was applied as a multipli- cative factor to physical doses in proton therapy, resulting in a CGE prescription. For example, a physical proton dose of 1.82 Gy would correspond to 1.82 × 1.1 = 2 CGE. Rather than CGE, the FIGURE C.54 A typical cobalt-60 treatment unit showing the treatment preferred term is now DRBE. couch and the control pendant suspended from the ceiling. Abbreviations: CGE = Cobalt Gray equivalent, RBE = Relative biological effectiveness. Related Articles: RBE Further Reading: Paganetti, H., A. Niemierko, M. Shielding Ancukiewicz, L. Gerweck, M. Goitein, J. Loeffler and H. Suit. 2002. Relative biological effectiveness (RBE) values for proton beam therapy. Int. J. Radiat. Oncol. Biol. Phys. 53(2):407–421. Cobalt unit Source (Radiotherapy) The 60Co radionuclide in a cobalt unit decays with a half-life of 5.26 years into 60Ni with the emission of elec- trons (beta particles), with a maximum energy of 320 keV, and Beam two gamma rays with energies of 1.17 and 1.33 MeV. The emit- ted gamma rays constitute the therapy beam, while the electrons are absorbed in the cobalt source or the source capsule, where FIGURE C.55 Cobalt sources can be deployed by rotation where the they produce relatively low energy and essentially negligible source is rotated from a safe position. The source can also be deployed by Bremsstrahlung x-rays and characteristic x-rays. translation to a position where exposure can be achieved. Coded aperture tomography 174 Coded aperture tomography Source Geometric field margin 100% Collimator C 50% Penumbra Penumbra Treatment couch Distance from central axis Geometric field size FIGURE C.56 Shows that the physical size of the source leads to a widening of the penumbra compared to that achievable on a linear accelerator. signal to noise ratio depends on the source spatial distribution and on the specific coded aperture pattern used. In a coded aper- ture imaging setting, assuming that the source consists of several point sources, each point source casts a shadow of the pattern on the detector. The size of the projected shadow depends on the geometrical setting (object to aperture, object to detector and aperture to detector distance). Each coded aperture pattern is associated with a decoding mask, and in order to decode the projected overlapped shadows, the acquired image is convolved with a decoding matrix. The dimensions of the decoding mask are also dependent on the geometrical settings. Common coded apertures used in nuclear medicine applications include uni- formly redundant arrays (URA) and modified uniformly redun- dant arrays (MURA). Due to practical fabrication constraints, a row and column representing solid material are introduced between each row/column in the mask. This is called a ‘no two holes touching’ (NTHT) technique. A third family of patterns, a special case of MURA, has a spe- cial property, namely that if rotated 90°, the pattern is the original FIGURE C.57 A 31 by 31 no two holes touching (NTHT) MURA patterns anti-mask, i.e. each hole is now an open pinhole and vice anti-mask coded aperture. Notice that at 90° rotation open elements are versa (Figure C.57). A reconstructed image can be obtained by replaced by closed elements and vice versa. The exception is the central summation of two images, one acquired with the original mask element in the centre of rotation. and one with the anti-mask. Due to the nature of the coded aper- ture arrays, the method is subjected to image artefacts. These artefacts can be avoided by choosing a less artefact prone array papers but is not commonly used in the current clinical environ- and using an optimised setup. Artefacts can also be compensated ment. In the technique, a number of pinholes are drilled in a solid for in the preprocessing prior to image reconstruction. plate that is placed in front of the scintillation crystal of a gamma Related Articles: Signal to noise ratio (SNR), Pinhole camera. Depending on the details of the technique, the pinholes collimator can be random, pseudorandom or structured. There is also a very elegant technique in which a Fresnel zone plate is used. Coded aperture tomography For a short time before single-photon emission computed (Nuclear Medicine) Coded aperture tomography is a technique tomography became common, a seven-pinhole technique that is that was developed to do limited angle tomography with a gamma somewhat related to coded aperture tomography was used for camera. The method has been the subject of many scientific tomographic imaging of the myocardium using Tl-201. Percentage depth dose Central axis Coded excitation 175 C oincidence timing window Related Article: Single-photon emission computed tomogra- Scaler phy (SPECT) Single channel analyser Coded excitation (Ultrasound) Coded excitation is a generic term for strategies to increase the signal to noise ratio (SNR) in imaging or Doppler Coincidence circuit modalities. In general, increasing the emitted energy can increase Amplifier Amplifier the SNR. There are, however, limits to the energies that can be emitted, as peak intensities are limited by various regulations for Preamplifier PM tube Scintillator PM tube Preamplifier + sample safety reasons. In addition, increasing the peak intensities will result in increased non-linear propagation where energy is lost to C higher harmonics that are quickly attenuated. A possible remedy FIGURE C.58 A scheme of coincidence circuit for liquid scintillation counter. is to emit longer pulses coded in a binary form or as a frequency- modulated chirp. As the shape and properties of the transmitted pulse are known, correlation-based methods can be used to extract coincidence system of counting reduces the background noise spatial scattering information. In radar systems, the improvement due to environmental and cosmic radiation, as well as PM-tubes in SNR can be a factor of a thousand or more, but is more mod- noise. est in ultrasound systems. Examples of such techniques that have Further Readings: Knoll, G. F. 2000. Radiation Detection been suggested for medical ultrasound are chirps or binary codes and Measurement, John Wiley & Sons, Inc., New York; Saha, G. such as Golay codes. P. 2001. Physics and Radiobiology of Nuclear Medicine, Springer, Heidelberg, Germany. Coherence (General) Coherence manifests itself in the interference between waves. The interference is stationary, i.e. temporally and spa- Coincidence detection in PET systems, true tially constant, and depends on the properties of the involved (Nuclear Medicine) See True coincidences waves. Waves interfere in either a constructive (add) or destruc- tive (subtract) manner depending on their phase relative to one Coincidence imaging another. (Nuclear Medicine) Coincidence imaging describes an imaging scenario with two opposed detectors providing near-simultaneous Coherence detection of

two detection events, most notably the annihilation (Non-Ionising Radiation) See Laser photons resulting from positron emission. Typically, the events fall within 0.5–20 ns of each other. The coincidence data are used to position the annihilation event during the process of image Coherent scattering reconstruction. (Radiation Protection) Coherent scattering refers to an interac- tion between a photon and an atom by which the direction of the incident photon is changed, but its energy remains the same. Coincidence summing of pulse-height spectrum Although in the interaction energy is initially transferred to the (Nuclear Medicine) Measurements of radiation from radionu- atom, it is subsequently given back to the emerging photon, even clides that emit multiple photons may sometimes result in events though the direction of the photon has changed. This process may that are incorrect. In some decay schemes, multiple photons can also be called elastic scattering. be emitted in cascade (or promptly) because they are part of the Related Article: Elastic scattering same decay. If two or more photons reach the detector and inter- act, then it is impossible to distinguish between them, leading to Coherent source a summation of the imparted energy. Examples of such radionu- (General) In physics, coherent source refers to a source that radi- clides are 111-In (emitting 245 and 172 keV) and 60-Co (1330 and ates photons with identical phase and frequency, meaning there 1170 keV). Coincidence summing can also appear in PET systems is no phase shift. The most common use of coherent sources is in where an event is registered if two annihilation photons strike the fields of lasers. two detectors at the same time. The assumption is then that the annihilation has occurred along a straight line defined by the two Coincidence circuit for liquid scintillation counter detectors. Recent PET systems try to take into account the time (Radiation Protection) The low-energy beta radiation emitted by, difference between the two photons due to the distance from the for example, H-3 or by C-14, maximum energy of 18 and 159 keV, centre of the system as information to construct images of less respectively, is measured using a liquid scintillation counter. The noise. Although these two photons come from the same decay, sample, prepared often in an aqueous solution, is introduced to they arrive at the detectors at different times because of the dif- the liquid scintillator in small quantity to prevent a large reduc- ferent distances. tion of the scintillation output (quenching). The counting effi- ciency is close to 100% if the photomultiplier (PM) tube noise and Coincidence timing window cosmic ray detection is eliminated. Therefore, the liquid scintil- (Nuclear Medicine) After annihilation, two photons are emitted lation counting system consists of two photomultipliers placed on in almost opposite directions. These two photons will arrive near- opposite sides of a scintillator vial and connected in a coincidence simultaneously to their opposite detector. The time difference circuit (Figure C.58). between their detection will be a function of the distance from The coincidence circuit enables pulses passing to the scaler the centre of rotation (COR). When an event is registered in a only if they are registered simultaneously by two PMs. The detector, the coincidence processor expects a registration in one Cold spot 176 Collection region of the opposite detectors. Since the speed of light is known and collection time is on the order of μs. The electrons move to the constant, it is possible to calculate the width of the time inter- collecting electrode. However, the charged particles may disap- val in which the corresponding photon should be expected. The pear from the active volume of the ion chamber as a result of ion maximum expected time interval is called the coincidence timing recombination or diffusion. window, which typically is 6–12 ns. In order to register a coinci- The collection efficiency of an ionisation chamber operating dence, there must be two recorded events; one in a detector, which under specified conditions is the ratio of the measured current will open a time window and another event in an opposite detector to the ideal saturation current, i.e. the number of ions collected within the time window. by the electrodes to the number of the ions created originally by If the count rates are too high, i.e. high administrated dose, radiation in the ion chamber volume. C there is an imminent risk that there are two or more events reg- If the applied voltage is high enough to suppress recombina- istered in an opposite detector within the coincidence timing tion, then all of the originally created charged particles are col- window. The coincidence processor might have difficulties in lected. The measurement of the ionisation current collected by the finding the true event and therefore misplacing the LOR. These electrodes enables the determination of radiation dose. coincidences are called random coincidences (more information Related Articles: Gas-filled radiation detectors, Ionisation in separate article Event type in PET). chamber Related Article: Event type in PET Further Readings: Bevelacqua, J. J. 2004. Contemporary Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Health Physics. Problems and Solution, Wiley-VCH GmbH Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, & Co. KGaA, Weinheim, Germany, p. 219; Knoll, G. F. 2000. Philadelphia, PA, pp. 218–222. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, p. 133. Cold spot (Radiotherapy) See Hot and cold spots Collection region (Radiation Protection) The term ‘collection region’ may apply COLIPA to heat or charge. This article deals with charge collection (Non-Ionising Radiation) COLIPA is the European Cosmetic region, which applies to radiation detectors including ionisation Association, which defines guidelines and standards for the test- chambers. ing of sunscreen creams. Such standards are also generally used One of the very important properties of an ion chamber is that by photobiologists in current phototesting practice, where patient every ion and free electron, produced by ionising radiation in the light sensitivity is tested by means of artificial light sources which gas, should be collected by the internal or outer electrode, follow- are designed to simulate solar radiation. ing the electric field lines. Related Articles: Solar simulator In a parallel-plate ionisation chamber, the electric field is uni- Further Readings: Matts, P. J. et al. 2010. The COLIPA in form and its strength E is equal to vitro UVA method: A standard and reproducible measure of sun- screen UVA protection. Int. J. Cosmet. Sci. 32(1):35–46; www .e V esc .e uropa .eu /e n /pol icies /poli cy -ar eas /e nterp rise/ datab ase -s elf -a E = d nd -co -regu latio n -ini tiati ves /1 27. where Collection efficiency V is the voltage applied to the electrode (Radiation Protection) When a gas-filled ion chamber is irradi- d is the distance between an inner and outer electrode ated, the gas in the chamber is ionised. The application of an elec- tric field to the electrodes causes the movement of the charged In a cylindrical ionisation chamber, the electric field lines are particles (electrons and ions) towards the electrodes. As the volt- radial and the strength E of the electric field depends on the dis- age difference between the electrodes is increased from zero to tance r according to the following expression: a high value, the current collected increases with voltage until reaching an asymptotic value – the saturation current for the given V radiation intensity. The free electrons and negative ions move in E = r ´ ln (b/a) the opposite direction than positive ions. The mobility μ of the charged particles, defined as where V is the voltage applied to the electrode p r is the distance from the anode (inner electrode) m = v ´ E b is the radius of an outer electrode (cathode) a is the radius of the anode where v is the drift velocity of the particles The collection region is an area in which the electric force F p is the gas pressure in the detector acting to the charge q E is the electric field strength F = q ´ E is almost constant in the ranges of the usually applied electric field and gas pressure. Its value is on the order of 10−4 m2 atm/ will cause its collection on the suitable electrode. The geometry of (V s), which corresponds to a drift velocity of 1 m/s for p = 1 atm the charge collection can be 2π or 4π. and E = 104 V/m and an ion transit time of ∼10 ms. Free electrons Related Articles: Parallel plate ionisation chamber, have the mobility about 103 greater than ions and therefore its Cylindrical ionisation chamber Collective dose 177 Collimation Further Reading: Knoll, G. F. 2000. Radiation Detection and colleagues in low and middle income countries. J. Med. Phys. Int. Measurement, 3rd edn., John Wiley & Sons, Inc., New York, p. 1(1):11–15. 133. Collimation Collective dose (Radiotherapy) To be clinically useful, radiation beams must be (Radiation Protection) The collective dose (usually meant to collimated. In a typical modern medical linac, the photon beam imply the collective effective dose) describes the total risk to a collimation is achieved in a number of ways as listed later. For group of individuals. electron beams, electron beam applicators are also used to create When a population is exposed to ionising radiation, each a clinically useful beam, in addition to the primary and secondary individual will receive an individual effective dose. Although collimators. C statistically it is possible to ascribe a risk of late stochastic radi- ation effects to that individual from this dose, it is much more 1. Primary collimator meaningful, especially in epidemiology, to look at the dose and This defines the largest available circular field size risk to the population as a whole by summing the mean effec- and is a conical opening machined into a tungsten tive dose to any individual over the total number of individuals shielding block, with the sides of the conical opening exposed. projecting on to edges of the target on one end of the Thus, mathematically, the collective dose S is the mean effec- block and on to the flattening filter on the other end. tive dose to an individual multiplied by the number of individuals 2. Secondary movable beam-defining collimators in the group: These consist of four blocks, two forming the upper and two forming the lower jaws of the collimator. They S = åEiNi can provide rectangular or square fields at the linac iso- centre, with sides on the order of few millimetres up to i 40 cm. Modern linacs incorporate independent (asym- where metric) jaws that can provide asymmetric fields, most Ei is the mean effective dose to an individual commonly one-half or three quarter blocked fields in Ni is the total number of individuals in the group which one or two beam edges, respectively, are coinci- dent with the beam central axis. The unit of collective dose is man Sieverts (manSv). 3. Multileaf collimator (MLC) Related Article: Effective dose These are a relatively recent addition to linac dose delivery technology. In principle, the idea behind a College of Medical Physics (ICTP) multileaf (multi-element) collimator (MLC) is simple; (General) The Abdus Salam International Centre for Theoretical however, building a reliable MLC system presents a Physics (ICTP) College of Medical Physics (focussed on medi- substantial technological challenge. The number of cal imaging) is a program of the ICTP, Trieste, Italy, first organ- leaves in commercial MLCs is steadily increasing, ised in 1988 by A. Benini and L. Bertocchi. Since 1988 the and models with 120 leaves (60 pairs) covering fields ICTP Medical Physics Colleges have been in 1990, 1992, 1994, up to 40 × 40 cm2 and requiring 120 individually 1996, 1999, 2002, 2004, 2006, 2008, 2010, 2012, 2014, 2016 and computer-controlled motors and control circuits are 2018. These colleges have educated around 1200 young medi- currently available. MLCs are becoming invaluable cal physicists from over 100 low- and middle-income countries in supplying intensity-modulated fields in confor- (31% of all students are women). Over the years the college mal radiotherapy, either in the step and shoot mode directors have included: A. Benini, L. Bertocchi, J. Cameron, F. or in a continuous dynamic mode. The MLC may be De Guerrini, S. Mascarenhas, R. Cesareo, P. Sprawls, J. Chela- attached

to the linear accelerator as a tertiary colli- Flores, S. Tabakov, G. D. Frey, F. Milano and M. De Denaro. mator or may be incorporated into one set of the sec- The faculty of lecturers has included eminent specialists from ondary collimators. many countries. A major college objective is to develop the students as edu- Collimator Scatter Factor: Radiation is scattered from the cators who can create effective programmes in medical physics collimators. In the case of the beam-defining collimators, this (imaging and radiation safety) within their countries. varies with field size because as the collimators open, they pres- Many of the students from the ICTP College of Medical ent more surface area to the beam, which leads to more scattered Physics have become respected medical physicists and leaders radiation. This is why output measured without a scattering phan- in their countries and have established academic departments tom, sometimes referred to as in-air, increases with field size. and societies and have become professors, heads of department This factor can be determined by measuring the change in out- and officers of their societies. They have active roles in the fur- put versus field size ‘in-air’ with a build-up cap large enough to ther professional development and healthcare provision in their provide electronic equilibrium. These values are normalised to a countries and in various international projects, including the 10 × 10 field. Multilingual Medical Physics Dictionary. Many college students Collimator Rotation Angle: Some collimators can be rotated have organised university programmes and short courses in their in order to provide a better fit for the radiation field to the area countries. of treatment. The degree of rotation is defined by the collimator Related Article: ICTP (International Centre for Theoretical rotation angle. Physics) Related Articles: Multileaf collimator, Electron applicator Further Reading: Tabakov, S., P. Sprawls, A. Benini, L. Further Readings: Bentel, G. C. 1992. Radiation Therapy Bertocchi, F. Milano and M. DeDenaro. 2019. International Planning, McGraw-Hill, New York; Podgorsak, E. B. 2003. College of Medical Physics at ICTP – 30 years support for Review of Radiation Oncology Physics: A Handbook for Teachers Collimator 178 Collimator design and Students, International Atomic Energy Agency, Vienna, the collimator surface may pass through. By using a collimator, it Austria; Walter, M. 2003. Textbook of Radiotherapy Radiation is possible to acquire images that represent the activity distribu- Physics, Therapy and Oncology, 6th edn., Churchill Livingstone, tion since the signal recorded in a detector element is assumed Edinburgh, UK; Williams, J. R. and D. Thwaites. 2000. to have originated somewhere along a line perpendicular to the Radiotherapy Physics in Practice, Oxford Medical Publications, detector element (Figure C.59). Oxford, UK. The shape, width and length of these holes determine the col- limator efficiency and collimator spatial resolution. The purpose Collimator of the collimator is to allow photons with a certain direction to (Nuclear Medicine) A collimator is a device that filters gamma reach the detector and at the same time stop other photons that C and x-rays so that only photons with a direction perpendicular to deviate from this direction. A collimator can be designed to fulfill a certain purpose, i.e. a pinhole collimator is used to get a magnified image of small organs. Detector For further details about different collimator designs, see Collimator design. Related Articles: Collimator design, Spatial resolution SPECT, SPECT Collimator (Radiotherapy) See Collimation Collimator design (Nuclear Medicine) Four main types of collimators are used in Detector obtaining the desired image resolution using a gamma camera. These are as follows: 1. Pinhole 2. Parallel-hole 3. Diverging 4. Converging The way these four collimators are constructed can be seen in FIGURE C.59 The collimator filters out every photon that has an Figure C.60. FOV is the field of view and S signifies the direction oblique angle of incidence. of the scan. Crystal b Crystal b a a Parallel hole Fov Fov Diverging hole Parallel hole Diverging hole Crystal b Crystal b a Aperture Fov a Converging hole Pin-hole Converging hole Fov Pinhole FIGURE C.60 The four most common collimator designs used in emission tomography. Collimator rotation angle 179 Collision kerma Pinhole Collimator: Pinhole collimators produce an inverted, size A and the collimator scatter factor. In the case of the beam- magnified image. The pinhole collimator is a small opening (a defining collimators, this varies with field size because as the col- few mm) in a high absorbing metal and is placed at the end of a limators open, they present more surface area to the beam, which metal cone, typically 20–25 cm from the detector. leads to more scattered radiation. This is why output measured The image is magnified when the distance between the detec- without a scattering phantom, sometimes referred to as in-air, tor and the face of the collimator f, is greater than the source to increases with field size. detector distance b, so the source must be placed close to the col- The collimator scatter factor, SC, is the ratio of the output ‘in- limator to get a magnification. When placed close to the collima- air’ for a given field size to that for the reference field: tor, the source is magnified but the camera can only attain a small imaged area. Because of the characteristics of the pinhole colli- S ( A E ) = ( A ) X ( A,E ) D ( A,E mator, it is used for the magnification of small organs (e.g. thyroid ) c , CF ,E = (1 ) = C X 0,E D (10,E ) and heart) and for small animal imaging. Parallel-Hole Collimator: The parallel-hole collimator is the most commonly used collimator for gamma cameras and it is This factor can be determined by measuring the change in output described by Cherry et al. (2003) as the ‘workhorse’ collimator. versus field size ‘in-air’ with an ionisation chamber plus a build- The collimator consists of a number of parallel holes separated by up cap large enough to provide electronic equilibrium. These val- lead septa. The septal walls prevent photons from crossing over ues are normalised to a 10 × 10 field. The measurement set-up is from one hole to another. shown in Figure C.61. Diverging Collimator: The holes in a diverging collimator This is otherwise known as the collimator factor (CF), or the diverge from a point typically 40–50 cm behind the collimator. air output factor, or the relative exposure factor (REF). A camera with a diverging collimator produces a minified, non- Related Articles: Peak scatter factor, Scatter factor, Tissue– inverted image. Using such a collimator makes it possible to proj- air ratio, Collimation ect sources distributed over a large area onto the detector. Converging Collimator: This collimator can project the mag- Collision kerma nified images of the source onto the detector when the source is (Radiotherapy) A major part of the initial kinetic energy of elec- located between the convergence point and twice the length of the trons in low-atomic-number materials (e.g. air, water, soft tissue) collimator. The holes typically converge at a point 40–50 cm in is expended by inelastic collisions (ionisation and excitation) with front of the collimator. These collimators can be used to maxi- atomic electrons. Only a small part is expended in the radiative mise the utilisation for a camera with a large detector area when collisions with atomic nuclei (Bremsstrahlung, electron–posi- imaging a small organ. tron annihilation). The total kerma (K) can thus be divided into Further Reading: Cherry, S. R., J. A. Sorenson and M. E. two parts: the collision kerma (Kcoll) and the radiative (radiation) Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, kerma (Krad): Philadelphia, PA, pp. 218–222. Related Articles: Collimator parameters, Spatial resolution SPECT, SPECT K = Kcoll + Krad The fraction of energy lost through the radiative processes is Collimator rotation angle known as the Bremsstrahlung fraction (g) and therefore the frac- (Radiotherapy) See Collimation tion due to collision interactions is (1 − g). So it is possible to relate the collision kerma to the total kerma in the following expression: Collimator scatter factor (Radiotherapy) Radiation scattered from the collimators makes a significant contribution to the total dose, and depends on field Kcoll = K (1- g) Source Source d = SSD p p dmax A 10 × 10 FIGURE C.61 Set up for measuring the collimator scatter factor. Collision mass stopping power 180 Collision mass stopping power For monoenergetic photons, the relationship between the collision c. e. p. kerma (Kcoll) and the photon energy fluence (Ψ) at that point in the medium is given by D and Kcol, β = 1 æ m Kcol = en ö Yç r ÷ è ø where (μen/ρ) is the mass energy absorption coefficient. For polyenergetic beams, the collision kerma can be expressed D C as a relationship between the photon energy fluence spectrum β < 1 β > 1 ΨE(E) and the mass energy absorption coefficient as follows: Kcol Buildup Equilibrium E region region max æ en ö Kcol = ò m YE (E )ç ÷dE r è ø 0 Depth or wall thickness that is FIGURE C.62 Relation between absorbed dose and collision kerma for a megavoltage photon beam. The point designated as c.e.p. is the centre of æ m electron production. (From Loevinger, R., Med. Phys., 8, 1981.) Kco = en ö l Yç r ÷ è ø where Related Articles: Kerma, Air Kerma, Exposure, Ψ is the photon energy fluence (integral distribution) Bremss trahlung (men /r) is the mass energy absorption coefficient averaged over Further Readings: Khan, F. M. 1994. The Physics of all photon energies, weighted by the spectral distribu- Radiation Therapy, 2nd edn., Lippincott Williams & Wilkins, tion of the photon energy fluence Baltimore, MD; Loevinger, R. 1981. A formalism for calculation of absorbed dose to a medium from photon and electron beams. It is clear that while the kerma occurs at a particular point, the Med. Phys. 8(1):1–12. absorbed dose (absorption of energy by the material) will take place at a different location due to the finite range of the second- Collision mass stopping power ary electrons produced. The collision kerma can be related to the (Radiation Protection) Mass stopping power is defined as the absorbed dose, since the radiative photons will mostly escape the change (loss) in energy of the particle per unit length of a parti- volume of interest, by the following approximation: cle’s path, i.e. dE/dx. The total mass stopping power (Stot) is equal to the sum of the collision mass stopping power (Scoll) and the D radiation mass stopping power (Srad): b = Kcol Stot = Scoll + Srad (C.3) By assuming all radiative photons escape the volume of inter- est, then β is approximately 1. In the build-up region of an x-ray The interaction of electrons with the electric field of the nucleus photon beam, β is less than 1, at the depth of maximum dose, to produce Bremsstrahlung gives rise to radiation mass stopping dmax (charged particle equilibrium) β is equal to 1, and at depths power. This process is more significant at high electron energies beyond this point (transient charged-particle equilibrium), β is and for media of high atomic number. greater than 1. The relationship between collision kerma and The collision mass stopping power itself is the change in absorbed dose at depths beyond dmax is almost constant because energy per unit path length due to elastic collisions with the the average energy of electrons generated does not change notice- atomic electrons of a medium. Collision energy losses decrease ably with increasing depth. This is illustrated in Figure C.62. very slowly in materials with atomic number, due to the greater Collision kerma may also be expressed in respect to exposure binding energy possessed by these materials. Collisions with by the following equation: atomic electrons cause excitation and ionisation of the atoms within the medium that a certain particle is travelling through. ( æ W Where the collisions lead to the total energy of the incident par- K air ö col ) = X air ç è e ÷ ticle being absorbed at the point of the interaction, or close by ø without the energy being transferred away, then it is referred to where as the restricted collisional mass stopping power. This concept is Wair/e is the average energy expended in air per

ion pair formed the most useful at a microscopic level in describing how energy is (the current most accurate value for this is 33.97 J/C) deposited and leads to absorbed dose. X is exposure (units C/Kg) given by the following relation: The collision mass stopping power is equal to the sum of the mass stopping power associated with soft collisions (Ssc) and that dQ associated with hard collisions (Shc): X = dm Scoll = Ssc + Shc (C.4) where dQ is the absolute value of the total change of ions of one sign produced in air when all the electrons and positrons pro- Soft collisions account for around 50% of all collisions. They duced by photons in mass dm of air are completely stopped in air. occur when an outer electron is dragged off by a passing charged Absorbed dose and kerma Collision sensor 181 Colour flow imaging (CFI) particle, causing excitation and ionisation. The excitations that significant (expensive) damage. The reasons for such an incident result from soft collisions give rise to low-level energy transitions could be due to either incorrect operation of the equipment or and associated Ĉerenkov radiation. Hard collisions result from simply electric failure of one or more controlling motors, causing interaction between a charged particle and a single electron. This them not to stop when required. interaction may lead to the ejection of a delta ray (high-energy Therefore, it is important for the safety of both personnel and electron) and characteristic x-rays (Figure C.63). equipment that some means of avoiding, or at least minimising Total mass stopping power varies for electrons and protons. the damage, is in place. The simplest collision sensor employed Electrons lose energy steadily and have a definite range. For is a ‘touch ring’ located externally on the head of the gantry (and electrons, change in energy varies more significantly with length also employed on some EPID arm systems). The ring contains travelled (depth in tissue). Protons exhibit a Bragg peak, where a microswitch held in the closed position such that when a tiny they react more significantly. This effect is used to the physicist’s deflection in the ring occurs, the switch opens and interrupts the C advantage in radiotherapy: treatments are planned so that the pro- power supply to all drive motors. The operator must then reset the ton Bragg peak, and its increased reaction rate, coincides with the switch to the closed position before continuing with any motion. treatment area (Figure C.64). Similar techniques using inductive proximity sensing circuits are Related Articles: Stopping power, Linear stopping power, also used. Bethe–Bloch equation, Bremsstrahlung, Delta ray, Ĉerenkov An alternative solution employed in some departments is to use radiation simple software to calculate combinations that would potentially cause a collision or crush a patient. The user enters the values and Collision sensor can quickly confirm whether such a situation would occur. This (Nuclear Medicine) This is a sensor used to prevent heavy camera solution removes the necessity of an external ‘touch ring’, but it equipment hitting a patient or other equipment. The sensors can does not take into account the variability in size and position of be placed on the mobile parts of the detector. One example is the patients and any other additional objects placed in the room. collimator collision sensor on a gamma camera, which registers Related Articles: Interlock, Interlocking device when the collimator strikes an object and stops the motion, pre- venting the camera moving too close to the patient. Collisional energy loss (Radiation Protection) Collisional energy loss occurs as a result Collision sensor of soft and hard collisions between particle radiation, or the sec- (Radiotherapy) As a result of the ability of various components ondary electrons caused by ionising radiation, and atoms within of a linac to move independently (gantry, couch, EPID), there is the absorber. For a definition of these interactions, see Collision a possibility that a patient, or indeed an operator of the equip- mass stopping power. ment, could be crushed, and equally there is a chance of a col- Related Articles: Secondary electrons, Secondary ionisations, lision between the moveable components, which could result in Collision mass stopping power Colour Doppler (Ultrasound) The common shorthand for colour flow imag- ing where estimates of the mean velocity are overlaid onto the B-mode image as colour-coded pixels. The mean velocity is found b b from an ensemble of pulses (4–16) and the mean phase shift that δ-ray has occurred between them, by what is referred to as the auto-cor- relator approach (see Autocorrelation). A more extensive descrip- a tion of the technique can be found under the article Colour flow imaging. a Colour flow imaging (CFI) Soft collisions: b >> a Hard collisions: b ≈ a (Ultrasound) Colour flow imaging (CFI) conventionally refers to the depiction of a colour-coded map of flow superimposed on a B-mode image. Colour is used to depict areas of the image in FIGURE C.63 Soft and hard collisions. which movement is detected by ‘Doppler’ methods. Commercial colour flow ultrasound scanners were introduced in the late 1980s when autocorrelation methods were used to produce 2D colour flow maps at clinically useful frame rates and spatial resolutions. Bragg peak The image (Figure C.65) shows a conventional colour flow image of flow in renal vessels. Colour is usually portrayed as red hues in one direction relative to the beam, blue hues in the oppo- site direction, but other colour maps may be used. Change in hue, for example from red to orange, represents changes in velocity vector and is in turn dependent on the speed and direction of the (a) Length (b) Length blood relative to the beam. The colour image depicts small ves- sels, which may not be evident on B-mode because of their small size. Because colour flow signals are highest when flow is aligned to the ultrasound beam, whilst in B-mode, contrast is best when FIGURE C.64 Electron and proton mass stopping power ranges: (a) structures are orthogonal to the beam. Colour can be used to iden- electron mass stopping power and (b) proton mass stopping power. tify vessels for pulsed-wave Doppler investigation. Energy Energy Colour saturation 182 C omforters and carers Columnar caesium iodide (Diagnostic Radiology) See Caesium iodide Combining cancer therapies (Radiotherapy) Up to three stages of cancer therapy can be deliv- ered with curative intent: 1. Neoadjuvant cancer therapy, used before the primary treatment to make it easier or more effective. C 2. Primary cancer therapy, used to eliminate the tumour as far as possible. 3. Adjuvant therapy, used after the primary treatment to eliminate any remaining cancer cells. Curative treatments may also be concurrent, i.e. delivered at the FIGURE C.65 Colour flow image of an aorta with flow in the right renal same time as each other. artery as red towards the probe and flow in the left renal artery (blue) Surgery and radiotherapy are the main types of primary can- away from the transducer. cer therapy. Additional therapies that may be given as neoadju- vant, adjuvant or concurrent treatment include: The colour flow image is very dependent on instrument fac- • Chemotherapy: uses drugs to kill cancer cells through- tors, including pulse repetition frequency, gain, wall filter, trans- out the body. mit frequency, size of the area under investigation (colour area), • Hormone or endocrine therapy: blocks/lowers hormone line density used and the ensemble length. The need for separate levels to slow/halt cancer growth. pulses for colour reduces frame rate of the ultrasound image and • Molecularly targeted therapy: targets specific genetic the trade-offs between frame rate, colour spatial resolution and mutations that help cancer cells to survive and grow. sensitivity are complex. These factors are partly under operator • Immunotherapy: enhances the body’s immune response control and partly optimised by the scanner in order to ensure to help eliminate the cancer. an acceptable image. As the colour map is, in effect, a separate • Cryoablation: kills cancer cells with cold. image from the B-mode, there are compromises made in depict- • Radiofrequency ablation: kills cancer cells by using ing both in the final image, with the possibility of one scan mask- electrical energy to heat them. ing information from the other. Narrowband and wideband techniques are now used commer- Related Articles: Palliative treatment, Radiofrequency abla- cially for CFI. Other CFI displays used are power Doppler, colour tion, Cryoablation M-mode imaging and 4D CFI. Comet tail Colour saturation (Ultrasound) The lung is normally considered poorly accessible (Diagnostic Radiology) See Hue, saturation, luminance (HSL) using ultrasound because ultrasound is heavily attenuated by air. Air also produces reverberation artefacts under the lung surface. Colour sensitivity This however does not mean that ultrasound is a poor diagnostic (General) The colour sensitivity of the human eye depends on vari- tool for conditions associated with the lung. ous parameters, most notably the luminance. Usually it is accepted The ‘comet tail artefact’ is an echographic image obtained that a normal untrained eye could distinguish between 150 and 250 with a cardiac ultrasound probe positioned over the chest and different grey shades, while the number of distinguishable colours consists of multiple lighter beams (‘comet tails’) fanning out in this case can be from 100,000 to several millions. from the lung’s surface. These ‘comet tails’ originate from As medical imaging relies mainly on greyscale images, water-thickened interlobular septa. The presence and the num- research has shown that a well-trained eye could distinguish ber of ‘comet tail’ images provide information on interstitial around 870 just noticeable differences (JND) of grey. The pixels pulmonary edema, making ultrasonography an attractive, easy- of contemporary digital medical images use 16 bits, of which usu- to-use, bedside diagnostic tool for assessing extravascular lung ally 12 bits are used to record the image contrast, and the other water (oedema or pulmonary congestion). The ‘comet tail’ four bits are used for supporting information (for example text or artefact is a well-known and widely used marker of pulmonary graphs displayed over the image). The 12 bits present 4096 levels oedema. of grey (212) which is more than enough for human vision. A spe- cial windowing technique is applied to adjust this large number of Comforters and carers grey levels to the less-sensitive human eye. (Radiation Protection) ‘Comforter and carer’ is defined in Related Articles: Visual acuity, Retina, Cones, Matrix size, Regulation 2(1) of the Ionising Radiation Regulations 1999 (IRR Window 99) as an individual who knowingly and willingly incurs an expo- Further Readings: Elert, G. ed. The Physics Factbook, sure to ionising radiation resulting from the support and comfort Number of Colors Distinguishable by the Human Eye, http: / /hyp of another person who is undergoing or who has undergone any ertex tbook .com/ facts /2006 /Jenn iferL eon g. shtml ; Kimpe, T. and medical exposure. T. Tuytschaever. 2007. Increasing the number of gray shades in When a person attending with a patient who is a relative or medical display systems – how much is enough? J. Digit. Imaging friend is expected to receive an exposure to ionising radiation, 20(4):422–432. whether at the time of the imaging investigation or radiation Commissioning 183 Compensation therapy, or as a consequence of tending to the patient afterwards, Compartment he or she may receive a radiation dose in excess of the annual (Nuclear Medicine) A compartment is a space in which substances public dose limit (1 mSv). In such cases, the person should be or tracers are distributed uniformly and in a specified form or designated as a comforter and carer. forms. The rate at which substance/tracer is transported out of a Comforters and carers are not subject to dose limits and can compartment is proportional to the amount in the compartment. enter controlled areas. However, the radiation exposure to the A tracer compartmental model is a mathematical description comforter and carer should only be conducted in accordance with of the transport/reaction pathways of a tracer in terms of intercon- suitable written arrangements designed to restrict radiation expo- nected compartments. sures to previously established constraints and so far as is reason- Further Reading: Strand, S.-E., P. Zanzonico and T. K. ably practicable. A reasonable constraint used is to limit the dose Johnson. 1993. Pharmacokinetic modeling. Med. Phys. 20: to the comforter and carer to no more than 5 mSv in any 5 year 515–527. C period. It

is essential in designating a person as a comforter and carer Compartment models to ensure that they are given sufficient information and instruction (Nuclear Medicine) A compartment is defined by a volume or about the hazards and risks and that they therefore give informed space within which the tracer is rapidly distributed uniformly over consent. the volume, which means that there are no significant concentra- tion gradients. The spatial extent of a compartment is sometimes Commissioning defined by clear physical boundaries, for example the intravas- (Radiation Protection) Commissioning is a term to describe the cular blood pool, reactants and products in a chemical reaction testing processes required to bring a new piece of equipment into and substances that are separated by a membrane. Other compart- clinical use. Commissioning involves familiarisation with new ments have different restrictions, for example the tracer can be equipment, in order to understand how to undertake the routine metabolised or trapped in one of two different cell populations in procedures (both clinical and non-clinical) for which it has been an organ, and thus the two cell populations represent two different purchased. compartments. Some aspects of commissioning may be undertaken at accep- Additionally, two different radioisotopes do not necessarily tance testing, for example baseline values will be determined for share the same compartments. For example, labelled red blood comparing with future quality control checks. The clinical com- cells are distributed intravascularly while thallium is distributed missioning should begin with applications training, where spe- both intravascularly and extravascularly. Therefore, a compart- cialists employed through the supplier train local operators on ment model (i.e. number, interrelationships, organisation and equipment use, and lead to a comprehensive understanding of the definition of compartments) must be calculated for each tracer. equipment’s uses and limitations. Operating procedures should Related Articles: Tracers, Analogue tracers be written to describe all the necessary operational functions and Further Reading: Cherry, S. R., J. A. Sorenson and M. E. features of the new equipment in order for it to be brought safely Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, and effectively into clinical use. Philadelphia, PA, p. 380. Committed dose Compensating filters (Radiation Protection) Following intake to the body of a radioac- (Radiotherapy) A compensating filter can be positioned in the tive substance, biokinetic models describe the uptake and expul- path of the beam to modify the dose distribution. These filters are sion of the radioactive substance over time in individual organs typically used for more complex situations where a simple wedge and tissues. It is necessary to know both the radionuclide and the or an additional layer of tissue equivalent material (bolus) is not chemical form to carry out this biokinetic analysis. The combina- suitable. tion of such information, together with the physical decay (half- The filters are manufactured from a high-density material and life) of the substance can then be used to determine the radiation positioned on an accessory tray in the head of the linac. The fil- dose received by the person over the time that the substance ters are extremely time-consuming to produce for each individual remains in the body. This is called the committed dose to the field and patient. They typically consist of small blocks of mate- person, defined by the International Commission on Radiological rial of varying thickness across the whole field. Protection (1991). The availability of IMRT with MLC results in the reduction of For each tissue or organ, T, there will be a committed equiva- the use of such filters. lent dose HT(τ) based on the activity accumulated within the tis- Abbreviations: IMRT = Intensity-modulated radiation ther- sue, the type(s) of radiation emitted by the substance, and the time apy and MLC = Multileaf collimator. it remains there, known as the integration time τ. If this time is Related Articles: Compensation, Compensator, Wedge filter, not specified for a particular radioactive substance and pathway of Bolus intake, then it is taken to be 50 years for adult intake, or 70 years Further Reading: Williams, J. R. and D. I. Thwaites. 2000. for children. Radiotherapy Physics in Practice, 2nd edn., Oxford University If the committed equivalent doses are multiplied by the rel- Press, Oxford, UK. evant tissue weighting factors and summed for the whole body, then we have an analogous quantity to effective dose – i.e. com- Compensating wedge mitted effective dose E (τ). (Radiotherapy) See Wedge Related Articles: Equivalent dose, Effective dose, Radiation weighting factor, Tissue weighting factor Compensation Further Reading: 1990 Recommendations of the International (Radiotherapy) The term compensation refers to the process Commission on Radiological Protection ICRP, Publication 60, of modifying a dose distribution to take account of an oblique Ann. ICRP 21(1–3), 1991. patient surface across the field. This is achieved by placing a Compensator 184 Compound filter compensator (compensating filter, wedge filter, bolus) in the where beam path. IMRT can also be said to provide a form of compensation. r = z = x2 - y2 Abbreviation: IMRT = Intensity-modulated radiation therapy. Related Articles: Compensator, Compensating filter, Bolus, q æ y = (z) = arg tan-1 ö ç è x ÷ Wedge filter ø Compensator where (Radiotherapy) A compensator is used to modify the dose r is called the modulus of z C distribution whenever the beam is incident on a non-uniform, θ is called the argument or phase of z or oblique, patient surface. For simple geometric situations, a compensator can be a wedge filter or an additional layer of Complication-free tumour control tissue equivalent material (bolus). However, for more complex (Radiotherapy) The aim of radiotherapy is to achieve compli- problems, a compensating filter must be uniquely made for each cation-free tumour control, i.e. deliver enough radiation to the field. tumour to destroy it without irradiating normal tissue to a dose IMRT can also be said to provide a form of compensation. that will lead to serious complications. For more information on Abbreviation: IMRT = Intensity modulated radiation therapy. the effectiveness of radiotherapy treatment, see the article on Related Articles: Compensating filter, Compensation, Bolus, Therapeutic effect. Wedge filter Related Article: Therapeutic effect Competent or regulatory authority Composite (Radiation Protection) Under the IAEA Basic Safety Standards, (Nuclear Medicine) A composite material refers to a substance a regulatory authority is defined as ‘an authority or a system of consisting of two or more materials. The two materials have dif- authorities designated by the government of a state as having legal ferent physical properties and remain separate on a macroscopic authority for conducting the regulatory process, including issuing level. Composite materials can be made to be both lightweight authorisations, and thereby regulating nuclear, radiation, radioac- and strong. Composites also offer a low attenuating alternative tive waste and transport safety’. for patient couches used in computed tomography and different Under the EURATOM Basic Safety Standards Directive, a emission tomography modalities. competent authority means the government department or agency Related Article: Couch (patient) designated by an EU member state as having the legal responsi- bility for regulating and enforcing the requirements of national legislation designed to meet the directive. Composite transducer As suggested in the IAEA definition, in some countries (Ultrasound) PZT, lead zirconate titanate, is the most commonly the role of enforcing the requirements of national legislation used transducer material. This material is a synthetic ceramic, for the use of ionising radiation and for radiation protection which can be produced in various versions for different applica- may be split between more than one regulatory or competent tions requiring particular properties. However, the mismatch in authority. acoustic impedance between PZT (3.0 × 107 kg/m2/s1) and soft Related Articles: Basic safety standards (BSS), IAEA, tissue (1.6 × 106 kg/m2/s) is a significant disadvantage for the EURATOM transmission of ultrasound energy into tissue and for the ability of transducers to produce short acoustic pulses necessary for high- resolution imaging. The mismatch will cause internal reverbera- Complex number tions within the transducer element. (Nuclear Medicine) A complex number, z, is an extension of the One way of reducing this mismatch is to use a composite real number system and can be expressed in the form of transducer where the PZT elements are embedded in a polymer matrix with lower acoustic impedance. The advantages of a com- z = x + yi posite transducer are better matched impedance, broad bandwidth and high sensitivity. where i is defined as Related Articles: Transducer, Backing material, Matching layer, PZT i2 = -1 Further Reading: Cobbold, R. S. C. 2007. Foundations of Biomedical Ultrasound, Oxford University Press, Oxford, NY. where y and x are real numbers Compound filter yi and x are called the imaginary and real part of the complex (Radiotherapy) In the kilovolt range, radiotherapy is normally number, respectively delivered using metal filters positioned in the beam to modify the radiation quality. In this way, softer (lower energy) x-ray photons One of the useful properties of complex numbers is that one that would irradiate the superficial layers of the skin and would can obtain negative real numbers by squaring complex numbers. not contribute to the dose at the desired point of delivery are Alternatives to the Cartesian representation of a complex number removed from the beam. are the polar or exponential representation A compound or composite filter uses two or more materials that complement each other in their absorbing abilities and are constructed with higher atomic number (Z number) closest to the z = reiq = rcosq + irsinq tube and with lower atomic number closest to the patient. Compound imaging 185 Compound imaging As the beam passes through the filter, most filtration occurs B C in the higher atomic number material and the lower atomic num- ber second layer of the filter absorbs any characteristic radiation D emerging from the first. With a copper and aluminium compound A filter, photoelectric attenuation in copper produces characteristic radiation with an energy of about 8 keV; this would be energetic enough to reach the patient and increase skin doses. However, the aluminium positioned on the patient side of the copper absorbs this characteristic radiation. Aluminium’s own characteristic radiation has a low energy (1.5 keV), and this is absorbed in the air gap between the patient and filter. C Related Articles: Skin sparing, Kilovoltage (kV) Further Reading: Williams, J. R. and D. Thwaites. 2000. Radiotherapy Physics in Practice, Oxford Medical Publications, Oxford, UK. Compound imaging FIGURE C.66 By using images acquired from different views, an (Ultrasound) Compound imaging uses multiple acquired images extended image is produced with echoes obtained from several directions to produce the final image. In current commercial usage, it is usu- where insonated areas of tissue overlap. ally applied to images that are a compound of images obtained from different transmit steering angles or frequencies. Three limitations of ultrasound are as follows: 1. Limited field of view from a single transducer 2. Directional dependence in the image 3. Speckle Compound imaging has, at times, been used to address all these limitations. Early compound scans used multiple images from different scan angles to produce a compounded scan. The advantage of this system is that some tissue interfaces are better seen from par- ticular scan direction and that effect of speckle is reduced since the speckle pattern is different from different scan angles and compounding produces some averaging of the speckle in the final S image. It is obviously imperative to have a clear spatial reference for the image, to align and merge the images produced from dif- Conventional image forming Compound imaging using multiple ferent views. in a linear array. Acoustic beam steering directions. The The system is shown diagrammatically in Figure C.66. An shadowing (S) deep to a region region of acoustic shadowing of attenuation is reduced in size early system using this approach was the Diasonograph® scan- ners from Glasgow in the early 1960s. A more recent variation of the technique is ultrasound tomography whereby tissue can FIGURE C.67 Diagrammatic representation of spatial compounding. be imaged from a circular array of transducers or transduc- ers swept in a circular motion. Compounded images may be obtained from pulse-echo images or transmission ultrasound techniques using speed of sound and attenuation properties. Systems using these techniques have been developed for breast imaging. Current commercial systems can use spatial compounding and frequency compounding to obtain images from a single trans- ducer using different transit directions

(Figure C.67) or different frequencies. This offers the advantage of reducing speckle arte- fact and improving echoes from interfaces over a wider angle than with conventional transmit paths. Since each compounded image requires more transmit pulses, effective frame rate is reduced. If artefacts are part of the diagnostic pattern recognition, for exam- ple in shadowing from microcalcifications, then the use of spatial compounding may not be recommended since it will inherently reduce this effect. In Figures C.68 and C.69, the effect of spatial compounding is demonstrated. FIGURE C.68 Longitudinal image of a kidney. Compound nucleus 186 Compton effect Compression (Ultrasound) The word compression is used in two different ways in diagnostic ultrasound. For compression of the medium by the ultrasound wave, see Rarefaction. Electrical signals can also be compressed. Compression of the dynamic range of the received echo signals in a B-mode scan- ner is of considerable importance. The received signals can have amplitudes in a 60 dB range, from μV to V while images in the display use a 20 dB range of brightness levels. The transformation C from a 60 dB range to a 20 dB range is effected by non-linear amplification where the smaller signals are amplified more than the larger signals. This is called compression. Different applications acquire different degree of compres- sion. This is why many scanners have a control to adjust the non-linear amplification curve, often labelled ‘compression’ or ‘dynamic range’. Related Articles: Rarefaction, Dynamic range FIGURE C.69 Longitudinal image of a kidney using spatial compounding. Compressed sensing (Magnetic Resonance) Compressed sensing (CS) is a technique Compound nucleus for efficiently acquiring and reconstructing an image based on under-sampled data. At its heart, compressed sensing is based (Nuclear Medicine) A compound nucleus is formed in a reac- on the idea that a signal can be recovered with far fewer samples tion in which two nuclei, or a bombarding particle (e.g. a neu- than required by the Nyquist sampling theorem. In signal pro- tron) and a target nucleus, combine into an excited nucleus. The cessing, the Nyquist sampling theorem states that the sampling nucleus lives for a certain time, and the nucleus ‘forgets’ how rate must be twice the highest frequency contained in the signal. it was formed. It decays by ‘evaporation’ of nucleons from the However, given knowledge about a signal’s sparsity, the signal compound nucleus, and by emission of gamma photons. In the may be reconstructed with even fewer samples than the Nyquist case where the bombarding particle is a neutron, the process may sampling theorem requires. The term for this combination of be illustrated by incoherent sampling exploiting signal sparsity is referred to as compressed sensing. n + AX ® A-1Y * ® AX* + n Þ AX* A Z Z Z Z ® Z X + g MRI k-space data represents a set of sinusoidal signals that are measured. Two things make the use of compressed sensing pos- where sible for MRI reconstruction: incoherent k-space sub-sampling A Z X is the target nucleus (sampling at with density below the Nyquist criteria) and the pres- A+1 * ence of signal sparsity. The incoherent nature of the sampling cre- ZY is the unstable compound nucleus AX* Z is the excited nucleus ates noise in the image rather than structured artefacts such as aliasing. This noise can be removed with a transform (such as a wavelet transform) and a noise thresholding procedure. There is Further Reading: Podgorsăk, E. B. 2006. Radiation Physics then an iterative process where the de-noised image is compared to for Medical Physicists, Springer, Berlin, Germany. the original image and a correction is made. The iterative process continues over a predetermined number of cycles, or more com- Compound scan monly until a predetermined level of data consistency is achieved. (Ultrasound) A scan where individual images obtained from dif- Compressed sensing reduces MRI scan time by allowing image ferent beam directions or at different frequencies are combined to reconstruction using fewer k-space lines, enabling faster acquisi- produce the final image. tion. This produces a high-quality image with reduced scan time. Related Article: Compound imaging Related Articles: k-space, Nyquist Further Reading: Lustig, M., D. Donoho and J. M. Pauly. Compounding 2007. Sparse MRI: The application of compressed sensing for (Ultrasound) Compounding is the process by which individual rapid MR imaging. Magn. Reson. Med. 58(6):1182–1195. images from different beam directions or frequencies are com- Hyperlink: www .s iemen s -hea lthin eers. com /e n -us/ magne tic -r bined to produce a final image. esona nce -i magin g /cli nical -spec ialit ies /c ompre ssed- sensi ng Related Article: Compound imaging Compton effect Compressibility (Radiation Protection) Also known as Compton scatter, incoher- (Ultrasound) Compressibility (κ) is the relative volume change of ent scatter or inelastic scatter. a fluid as a response to pressure change: When a photon strikes an outer orbital or unbound electron, there is some transfer of energy to the electron, but the photon is 1 ¶V k = - not completely stopped. The process may be thought of in terms V ¶p of a billiard-ball type collision. (Figure C.70) The incident photon interacts with a ‘free’ (outer shell) electron. The incident photon The compressibility together with density determines the sound strikes the electron whence the electron and the scattered pho- speed in a fluid. ton are emitted in separate directions. Some of the photon energy Compton electron 187 Computed tomography (CT) is scattered into a new direction, and the electron referred to as Compton or recoil electron is ejected from the atom with kinetic Ejected electron energy equal to the loss of energy of the photon. Also known as inelastic scattering, incoherent scattering or Compton interaction. Related Article: Compton effect + Computational phantoms (Diagnostic Radiology) See Software phantom Incident C photon Computed radiography (CR) (Diagnostic Radiology) Computed radiography (CR) is an x-ray diagnostic radiology imaging process that uses photostimulable phosphor plates. In most cases, the CR plates are a direct replace- ment for film-screen cassettes, so conventional imaging equip- Scattered ment can be used. In the process, photon • A plate is inserted into the cassette holder and exposed FIGURE C.70 Illustration of Compton effect. in the conventional way. • The plate is taken to a reader and read out using laser stimulation. is transferred to the electron and the energy lost by the incident • The plate is erased using bright white light. photon increases as the scattering angle increases (i.e. the angle between the initial direction of the photon and the direction of the The main advantage of CR is that the dynamic range is much scattered photon). greater than screen film. Thus, errors in exposure do not cause as The Compton effect dominates interaction processes at great a degradation of image quality. Resolution is less than screen medium energies (90 keV to 2 MeV in water or human tissues; film, but the values are acceptable in most imaging situations. between 200 keV and 2 MeV, the only interaction is Compton). CR systems use the existing x-ray tube, generator and stand. The process is also known as inelastic scatter. The ejected elec- The only modification of the x-ray system is that the x-ray tron is known as a Compton electron or recoil electron. film cassette is replaced by a storage phosphor cassette. This The likelihood of interaction for the Compton effect varies makes the transfer from classical radiography to CR very cost with 1/E where E is the energy of the incident photon, and with effective. electron density, represented by NAZ/A where NA is Avogadro’s Some sources use the term computed radiography (CR) instead constant representing the number of atoms per unit mass, Z is of the term digital radiography (DR), or vice versa. These two atomic number and A is atomic mass. Hence, Compton effect is terms differ as CR and DR use different detectors (Figure C.71). relatively independent of the Z of the material. CR systems use photostimulable phosphor plates as the detector. Related Articles: Compton scatter, Incoherent scatter, These systems were developed before the DR systems and used Inelastic scatter the phosphor plate as a replacement of the x-ray film. Direct DR uses x-rays transferred to charge by amorphous selenium detec- Compton electron tors, and indirect DR uses phosphor and photodiode or CCD (Radiation Protection) An electron ejected from an atom as a detector. See the specific articles for these systems. result of a Compton interaction with an incident photon. Only Related Article: Storage phosphor some of the energy of the photon is given to the electron, and the photon is scattered. Computed tomography (CT) Related Article: Compton effect (Diagnostic Radiology) Computed tomography, usually referred to as CT, is a medical diagnostic imaging technique, which uses Compton interaction x-rays to create cross-sectional images, or ‘slices’, of a patient’s (Radiation Protection) Interaction between an incident photon anatomy. The medical CT scanner was the invention of British and an outer shell (loosely bound) atomic electron in which only engineer, Sir Godfrey Hounsfield, and the first clinical scans some of the energy of the photon is transferred to the electron. of a patient’s head were obtained in 1972 at Atkinson Morley’s The photon is scattered to a new direction, and the electron is Hospital in Wimbledon, London (Figure C.72). Later, the tech- ejected from the atom with kinetic energy equal to the loss of nique was refined for body imaging. In 1979, Sir Godfrey, and energy of the photon, minus the work function for ejection of the American scientist Allen Cormack, were awarded the Nobel Prize electron. in Medicine for the invention. Also known as inelastic scatter, incoherent scatter, or Compton The basic imaging principle of CT is the same as that of plane- scatter. film radiography, in that it relies on the differences in the x-ray Related Article: Compton effect attenuating properties between materials. CT does not use radio- graphic film to acquire the image information, but digital signals Compton scattering from the detectors are sent to a computer for processing and then (Radiation Protection) Interaction between an incident photon displayed on a monitor. and a loosely bound atomic electron in which only some of the The principal components of a CT scanner are a collimated energy of the photon is transferred to the electron. The photon source of radiation, usually a diagnostic x-ray tube, a radiation Computer-aided detection 188 Computer-aided perception (CAP) Analogue Screen/film Computed Photostimuable radiography phosphor (CR) X-ray Phosphor + detector CCD technologies Indirect x-ray C conversion Digital Phosphor + a-Si/TFT Direct x-ray (flat panel detector) conversion Direct digital radiography Direct x-ray (DDR) Photoconductor + conversion TFT (flat panel detector) FIGURE C.71 Different x-ray imaging detector technologies. (Image courtesy of A. Pascoal.) years, the multislice CT scanner has been introduced, which enables the simultaneous acquisition of a number of ‘slices’ in a single rotation. Related Articles: CT reconstruction, Multislice CT scanner, Helical scanning Computer-aided detection (General) Computer-aided detection (CAD) is a digital technol- ogy used in various algorithms for image detection and recog- nition. CAD has numerous applications in various branches of industry, including healthcare. CAD is used in healthcare pri- (a) (b) marily as part of different assistive tools and algorithms, in the form of computer-aided diagnosis (CADx) to assist medical staff FIGURE C.72 (a) One of the first CT scans of the human brain (1972) during the reading of the diagnostic images. The overall effect is and (b) a modern CT brain scan. (Courtesy of ImPACT, UK, www a reduction in observational oversights, reduction in diagnostic .impactscan .org) errors and optimisation of the diagnostic process both in terms of speed and performance. CAD and CADx technologies are commonly used in parallel detector and a computer. The x-rays emitted pass through the to ensure optimal diagnostic process. Each of the technologies object to be imaged, and the detectors measure those transmitted. however is responsible for the different aspects of the diagnostics. The x-ray source and detector are rotated through a minimum of AD assists image recognition – mainly special related parameters 180° around the object, and attenuation information is obtained like location, shape, dimensions, etc., while computer-aided diag- from different angles, or views. The data are processed by the nosis assists the conclusion making, e.g. defining the type of the computer, enabling a cross-sectional image of the object to be

finding. reconstructed. Currently, most CT scanners use filtered back-pro- CAD and CADx systems combine artificial intelligence, neu- jection for image reconstruction. However, iterative reconstruction ral networks and other contemporary technologies. They serve as methods are being implemented on the latest models of scanners. a valuable assistant for doctors in the diagnostic radiology units. Early CT scanners were of the so-called first-generation Interpretation of images originating from DR modalities like dig- geometry (Figure C.73). An x-ray pencil beam and single detector ital x-ray, CT, MRI, US may be significantly improved with the translate across the field of view to acquire an attenuation profile. help of the detective and assistive technologies. The assembly (x-ray beam–detector) then rotates through small angles, each time acquiring a further profile. Computer-aided diagnosis (CADx) The schematic diagram shows a third-generation CT scanner (General) Computer-aided diagnosis (CADx) is the application of (Figure C.74), in (a) the scan (x–y) plane and (b) the sagittal (y–z) data processing to clinical data and clinical images, with each plane. This design is now almost universally employed in medi- patient’s data being analysed by a computer, which offers a tenta- cal CT scanning. It employs an x-ray fan beam in the x–y plane, tive diagnosis or range of diagnoses with their associated statisti- and an arc of detectors, whose extent is generally large enough to cal probabilities. encompass the whole object being imaged, eliminating the need for translation. The tube and detector assembly rotate around the Computer-aided perception (CAP) z-axis to acquire data for image reconstruction. Along the z-axis, (General) Computer-aided perception is the use of image pro- the x-ray beam is collimated, so that information acquired in a cessing of clinical data to assist clinicians by providing additional single rotation is limited to a narrow ‘slice’ of tissue. In recent visual clues to aid in the diagnosis and treatment of patients. Computer-controlled accelerator 189 Concave target volume 0° Translate X-ray pencil beam Rotate 90° C Detector (a) Distance (b) Distance FIGURE C.73 Principle of operation of a first-generation CT scanner showing two attenuation profiles for a simple object: (a) 0° profile and (b) 90° profile. X-ray tube Gantry Patient Table Arc of detectors Side view of detector bank FIGURE C.74 A schematic diagram of a third-generation CT scanner: (a) scan plane and (b) sagittal plane. (Courtesy of ImPACT, UK, www .impactscan .org) A basic form of aided perception would be to use image pro- (RV system). Treatment parameters are transferred electronically cessing techniques to highlight areas of mammograms, which the from the treatment planning system to the RV system. RV system computer calculates as abnormal, and highlight these automati- holds the patient treatment parameters and checks them against cally for the reporting radiologist. the parameters used in daily set up to confirm that the set up is More complex processes include the generation of ‘virtual correct. reality space’, based on 3D imaging data such as helical CT or Computer Control: In addition, the working configuration and MRI and using sophisticated tissue segmentation algorithms and many internal operating parameters of the linear accelerator are 3D display software to provide a virtual patient, presenting the set and monitored by the system’s own dedicated computer. These information to the clinician in a clearer and more easily compre- parameters are, for example output, beam flatness and symmetry. hended model. If they exceed predetermined tolerances, then software interlocks This ‘virtual reality’ may be projected back into the space will switch the linear accelerator off until parameters are adjusted occupied by the patient, such as in some image-guided therapies, back within tolerance. producing additional 3D visual clues as to the subsurface location Abbreviation: RV system = Record and verify system. of tumour or important blood vessel. Related Article: Interlock Computer-controlled accelerator Concave target volume (Radiotherapy) (Radiotherapy) A concave target volume is defined as one who Networking: To achieve the required dose delivery, modern has at least one interior angle greater than 180°. The opposite of a linear accelerators are computer controlled. The main delivery concave target volume is a convex target volume, see Figure C.75 control mechanism is represented by the record and verify system for theoretical examples. Attenuation Attenuation Concomitant boost 190 Condensed history technique Convex Concave Condensed history technique (Radiotherapy) In Monte Carlo modelling of charged particles, the condensed history (CH) technique is the usual method of approximated particle track simulation. Unlike photons (at radio- therapy energies), which have mean free path lengths between interactions on the order of the typical simulation geometry, FIGURE C.75 Simple examples of convex and concave target volumes. charged particles including electrons undergo huge numbers of continuous interactions, which would be computationally pro- hibitive to model in full. Therefore, it is desirable to ‘condense’ C Concave target volumes can be treated successfully with multiple elastic and semi-elastic interactions into a single history, intensity-modulated radiotherapy, which allows critical structures which still sufficiently approximates the true particle history. within the concavity of the target volume to be spared in the dose To this end, a multiple scattering theory is required, which can distribution. Typical tumour sites and respective critical struc- encompass the effect of multiple small-angle scattering events tures that require concave target volumes include within a single path deflection. The selection criteria for those collisional and radiative events that are sufficiently semi-elastic to • Harynx (spinal cord) be accurately accounted for under a multiple scattering theory are • Prostate (rectum) implemented as simple energy cut-offs. The two separated classes • Skull base (optic pathway structures, brainstem) of interaction are then termed as hard and soft. Energy is modelled as being continually lost/deposited along Concave target volumes cannot be treated successfully with mul- the condensed charged particle track due to soft interactions. Hard tiple beams with no intensity modulation. interactions must each be modelled explicitly, and thus they will Related Articles: Target volume, Conformal radiotherapy, end the current CH step when they occur. Where no hard interac- Beam’s eye view, Intensity-modulated radiotherapy tion occurs across a chosen maximum step length, the CH step is Further Reading: Bortfeld, T., R. Schmiidt-Ullrich, W. de artificially terminated, in order to preserve greater accuracy in the Neve and D. E. Wazar. 2005. Image-Guided IMRT: Concepts and simulated multiple scatter event (see Figure C.76). Clinical Applications, Birkhauser, Basel, Switzerland. If the multiple scattering angle is applied at the end of the CH step, it is likely to overestimate the range of the charged particle, as in reality the particle will fluctuate along some curved path Concomitant boost instead of the straight line assumed under the CH technique. For a (Radiotherapy) This technique maintains a standard fractionation similar reason, the particle position at step end will be incorrectly schedule to a large target volume and delivers a second daily dose displaced transversely. A number of solutions exist to minimise to a small boost volume (a ‘field within field’) on some of the such errors, with one of the simplest being to sample the scatter- treatment days. ing angle at a random position along the CH step, instead of at its end (see figure). This is known as the random hinge method. Concrete Related Articles: Monte Carlo method, Variance reduction (Radiation Protection) See Radiation shielding techniques Further Reading: Seco, J. and F. Verhaegen. 2013. Monte Concurrent therapy Carlo Techniques in Radiation Therapy, CRC Press, Boca (Radiotherapy) See Combining cancer therapies Raton, FL. FIGURE C.76 Representative path of an electron, and its energy deposition under a CH technique. Condenser chamber 191 Conductivity Condenser chamber Sample holder (Radiation Protection) A number of ionisation chambers can be described (in broad terms) as capacitors. These are initially charged to certain potential V1 (prior to the measurement). The ionising radiation gradually discharges the chamber and the resulting decrease of the voltage is used to estimate the integrated Inner electrode ionisation charge. The charge liberated by the radiation is Q = C (V1 -V Outer electrode 2 ) C where C is the capacity of the condenser chamber V1 is the initial charge of the condenser chamber (prior to the measurement) – + – V2 is the voltage after the ionisation radiation has partially dis- charged the chamber (at the end of the measurement) FIGURE C.78 Scheme of the well-type condenser chamber designed for radiation source calibration. Such condenser chambers use a capacitor connected to their central electrode. This capacitor collects the liberated charge. Later a reading device measures the integrated charge in the current from an unknown source is compared with that measured capacitor. This design is convenient for measurements where from a standard one. the cable between the chamber and the electrometer has to Related Articles: Integrating dosimeter, Cylindrical ionisa- be avoided as a source of noise. Often condenser ionisation tion chamber, Ionisation chamber chambers are used to measure low exposure rate, for example Further Readings: Johns, H. E. and J. R. Cunningham. as personal pencil dosimeters (integrating dosimeter, pocket 1983. The Physics of Radiology, 4th edn., Charles C. Thomas, dosimeter). Springfield, IL; Knoll, G. F. 2000. Radiation Detection and Figure C.77 shows a condenser chamber consisting of a cen- Measurement, 3rd edn., John Wiley & Sons, Inc., New York, pp. tral electrode and outer electrode (chamber wall). The chamber 145–146. is filled with air and connected to the electrometer. The chamber is charged and placed in the radiation field for a period of time in which the integrated ionisation charge is measured. Conduction band A typical example of condenser chamber is the so-called (General) The conduction band refers to a range of electron pocket chamber. It is with small size (like a pen) and equipped energy, higher than that of the valence band. It is either unfilled or with a quartz fibre electroscope for reading the dose on an internal partially filled by conduction electrons (Figure C.79). scale. Specific energy is required for each electron to migrate from The ionisation condenser chamber can perform as well-type the valence band to the conduction band. In case an electron (Figure C.78) designed to measure the radioactivity of the source jumps to this higher energy level, it resides there for fractions of in a 4π counting geometry. The walls are made of brass or steel a second, then returning back to the valence band by emitting and the inner electrode of a thin foil of aluminium or copper. the energy in the form of heat, light, or transferring it to another The gas pressure in the chamber can be changed to increase the electron. ion current. This type of condenser chamber is used for radia- Conduction band electrons are free to move around the crystal. tion source calibration, for example of gamma radiation. The ion Upon application of external electric field, the conduction elec- trons accelerate. Conductivity Outer electrode (General) Conductivity, or specifically electric conductivity, is (chamber wall) defined as the ability of a substance to conduct electric current between two points. The conductivity is the reciprocal (inverse) Central electrode E Conduction band Band gap Valence band To electrometer FIGURE C.77 Scheme of condenser air chamber with central electrode connected to electrometer. FIGURE C.79 Conduction band. Cone 192 Cone beam CT of the more commonly encountered term, resistivity (ρ) and Tube at 0° depends on the number of charge carriers in the material and their mobility. It depends on the nature of the substance, its cross- sectional area and its temperature. All substances possess con- ductivity to some degree, but the amount varies widely, ranging from extremely low (insulators such as benzene, glass) to very high (silver, copper and metals in general). Metals are better con- ductors of electricity because of their high free-electron density, while non-metals, such as rubber, are poor conductors and may be C used as electrical insulators. Increasing the cross-sectional area of a given conductor will increase its conductivity since more electrons will be available for conduction. However, increasing the temperature of a metal conductor will reduce its conductiv- ity due to the electrons ‘bumping’ into each other more as they are moving faster. Controversially in a non-metal, an increase in temperature improves conduction because it frees more electrons. Tube at 180° Conductivity measurements are widely used in industry for liq- z-axis uids, which generally consists of ionic compounds dissolved in (a) (b) water. These solutions have conductivities approximately midway between insulators and metallic conductors. The conductivity is FIGURE C.80

(a) Inconsistent projection data in peripheral detectors easily measured by electronic means, and provides information and (b) consistent projection data for central rays (not to scale). (Courtesy about the quality of the water, or the make-up of the solution. of ImPACT, UK, www .impactscan .org) Electrical conductivity is commonly represented with σ and has the SI units of siemens per metre (S/m). Cone (Diagnostic Radiology) A very simple beam restrictor – a metal extender (cylinder cone) attached to the diaphragm of an x-ray equipment. See article on Beam restrictor. Related Articles: Beam restrictor, Diaphragm Cone beam artefact (Diagnostic Radiology) The cone beam artefact occurs on mul- tislice scanners with extended coverage in the z-direction (see (a) (b) Cone beam CT). For scanners acquiring more than four simulta- neous slices, the beam can no longer be assumed to have parallel FIGURE C.81 Image of Teflon rod at angled relative to scan axis: (a) beam (pencil beam) geometry along the z-axis. absence of cone beam artefact and (b) appearance of cone beam artefact. The diverging beam along the z-axis results in inconsistent (Courtesy of ImPACT, UK, www .impactscan .org) projection data from opposing views, and this effect increases with distance from the central ray. Figure C.80a shows (in an exaggerated way) the path of the ray to the outermost detector. Cone beam CT When the x-ray tube is at 0°, the ray traverses a different path (Diagnostic Radiology) Single-slice CT scanners were limited through the patient in comparison to the situation when the tube is to a maximum coverage of 10 mm along the scan axis (z-axis). at 180°. These inconsistencies give rise to the cone beam artefact. Because of the narrow divergence angle of the beam in the The inconsistencies are not present for the central rays in the x-ray z-axis direction, they could be considered as having parallel ray beam as the beam follows the same path through the patient in geometry (Figure C.82a), with the x-ray source, detector and opposing views (Figure C.80b). imaged slice lying in the same plane. Cone beam CT refers to Figure C.81 shows images of a Teflon rod angled with respect the extended coverage of the x-ray beam along the z-axis, which to the horizontal. The images were obtained on a 16-slice scan- is available on multislice scanners. The divergence angle of the ner. The artefact is seen in images from the outer detector rows beam is increased and this angle is referred to as the cone angle (Figure C.81b). The severity of the artefact increases for objects (α) (Figure C.82b). For scanners acquiring more than four simul- away from the centre of the field of view. taneous slices, the beam can no longer be assumed to have paral- To reduce the appearance of cone beam artefacts, multislice lel geometry and they are therefore considered as cone beam CT scanners with more than four slices utilise special reconstruc- scanners. In cone beam CT, off-axis objects are projected onto tion algorithms, which are either adaptations of conventional different detector rows for different tube angles (Figure C.83), filtered back-projection methods or modified Feldkamp-type and the data from one rotation no longer represents a planar reconstructions. slice. Different approaches to image reconstruction are therefore Related Articles: Artefact, Beam hardening, Cone beam required in order to avoid cone beam artefacts. CT, CT reconstruction, Helical artefact, Image artefact, Metal The types of algorithms used in cone beam reconstruction can artefact, Motion artefact, Partial volume effect (artefact), Ring be divided into two types: exact cone beam, 3D algorithms, using artefact modified Feldkamp-type methods, or approximate cone beam, Cones, retina 193 Confidence Limit Collimators Cone angle α Collimators Fan beam Cone beam Beam width Beam width (a) Detectors (b) Detectors C FIGURE C.82 Diagram of CT x-ray beam: (a) on a single-slice scanner and (b) on a multislice scanner with extended coverage along scan axis (z-axis). 180° Projection data • Blue sensitive cone cells (aka S cones, for short) with peak sensitivity around 420–440 nm • Green sensitive cone cells (aka M cones, for medium) with peak sensitivity around 534–555 nm • Red sensitive cone cells (aka L cones, for long) with peak sensitivity around 564–580 nm The cones are mainly concentrated in the macula of the retina and are larger than rods – the cone diameter is about 0.006 mm, while rod diameter is about 0.002 mm. There are approximately six million cones in the retina, the majority of which are found in the fovea. The distribution of the cone cells is roughly accepted as 64% reds, 32% greens and 2% blues (also, their distribution in the ret- ina is not homogeneous). The maximum sensitivity of the human eye is in the region of the green-yellow colour. As the cones are responsible for perception of the high-resolution images, the rep- resentation and transmission of colour is directly related not only to the visual contrast resolution, but also to the spatial resolution. The discovery of the light sensitivity of the cone cells and rod 0° Projection data cells in the human eye retina (G. Wald and P. Brown at Harvard, z-axis and E. MacNichol, W. Marks and W. Dobelle at John Hopkins, 1964) placed the RGB colour model on more solid scientific FIGURE C.83 background. Although the photoreceptors of the retina had been Off-axis objects projected onto different detector rows for different tube angles. (Courtesy of ImPACT, UK, www .impactscan .org) known since the nineteenth century and the light-sensitive protein rhodopsin had been discovered by Franz Boll in 1876, the more detailed understanding of colour perception needed some 100 2D algorithms, such as advanced single-slice rebinning (ASSR) years further research. Despite this, visual perception and colour used by Siemens. processing in the human brain are still not well understood. Cone beam CT scanners include conventional diagnostic mul- Related Articles: Retina, Rods, RGB, Visual acuity, Visual tislice systems acquiring up to 320 simultaneous slices with a perception matrix detector array and also systems for different applications Further Readings: Hecht, E. 1987. Optics, 2nd edn., Addison that use flat-panel detectors (FPDs). The latter include dental CT, Wesley; Rods and Cones, (accessed 28 February 2013), http: / / CT attachments to linear accelerators for image-guided radiother- hyp erphy sics. phy -a str .g su .ed u/ %E2 %80 %8 Chbas e /vis io n /r odcon apy (IGRT) and to C-arm angiography systems. e .htm l; http://en .wikipedia .org /wiki /Retina; Tabakov, S. 2013. Related Articles: Cone beam artefact, CT reconstruction Introduction to vision, colour models and image compression. J. Further Reading: Shaw, C. ed. 2014. Cone Beam Computed Med. Phys. Int. 1:50–55. Tomography (Imaging in Medical Diagnosis and Therapy), CRC Press, ISBN: 978-1439846261. Confidence Limit (Nuclear Medicine) Confidence limits are outer limits making Cones, retina up a confidence interval and give a measure of the reliability of (General) The photoreceptors of the retina are of two types – a single measurement based on statistics from a whole popula- cones and rods. The cones (named this way due to the shape tion. Confidence limits are set with respect to the expected prob- of the cells) are associated with colour vision. There are three ability level desired. The probability that a population contains a types: specific value is called the confidence level. In order to increase Conformal dose distribution 194 Consistency the confidence level, one must widen the confidence interval, i.e. change the confidence limit. Conformal dose distribution (Radiotherapy) A conformal dose distribution is produced whenever each individual beam is shaped to the target volume as viewed from that beam position by the use of blocks, MLCs etc. The dose distribution isodose lines will then neatly surround the target with the appropriate level of dose while minimising the C dose delivered to nearby healthy tissue and adjacent critical struc- tures (organs at risk). Abbreviation: MLC = Multileaf collimator. Related Articles: Multileaf collimator, Tertiary collimator, Conformal radiotherapy, Custom blocking, Dose distribution, Target dose distribution, Critical structures, Organs at risk Conformal radiotherapy (Radiotherapy) When treatment is delivered with unshielded FIGURE C.85 Three-field treatment of a prostate in red showing the square or rectangular fields, the dose distribution mimics the fields and the MLC shapes for each beam (VARIAN CadPlan planning beam arrangement, producing brick-shaped high-dose volumes. system). This result makes it difficult to avoid giving high doses of radia- tion to sensitive tissues around the treatment volume. This is because the high-dose volume does not have the same shape, i.e. does not conform to the shape of the intended treatment volume. Conformal radiotherapy can be used to describe any beam modulation that enhances conformation of the dose to the tumour. Three-dimensional (3D) conformal radiotherapy gener- ally refers to multiple field treatments with each portal being reshaped with a multileaf collimator to account for the complex shape of the tumour. In this way, tumour is conformally treated and radiosensitive normal tissue is conformally avoided (see Figure C.84). Figures C.85 and C.86 show a 3D view of a prostate treatment. In Figure C.85a, the patient outline can be seen (green transparent rendering), along with the prostate and seminal vesicles (target region) rendered in red. Three fields can be seen, each shaped with an MLC. In Figure C.86b, the high-dose region (95% of the iso-centre dose) can be seen in white. Also seen are hot spots in the left and right lateral positions, often seen in this type of FIGURE C.86 The white area represents the high dose envelope (95% treatment. of the iso-centre/prescribed dose) (VARIAN CadPlan planning system). Further Reading: Metcalfe, P., T. Kron and P. Hoban. 1997. The Physics of Radiotherapy X-Rays from Linear Accelerators, Medical Physics Publishing, Madison, WI. Conformity index (Radiotherapy) The conformity index (CI) is the ratio of the treated volume to the planning target volume. Related Articles: Treated volume, Planning target volume Consistency (Diagnostic Radiology) The term consistency is used in quality control (QC) assessments. It describes the ability of a system to maintain stable parameters. For example, x-ray tube output consistency refers to the ability of the x-ray system (the generator and the tube) to pro- duce exposures with identical dose output (mGy) when all other parameters (kVp, mA, etc.) are constant. The minimal number of identical exposures (with identical parameters) used for calculating the consistency of an x-ray system is 4 (optimal is 6). The dose output consistency (%) in this case is calculated as the % ratio between the standard deviation of all four mea- sured exposures (mGy) and the average of these (this has to be made for each focal spot) – 100*STDEV/AVERAGE. If the FIGURE C.84 The multileaf collimator shapes the beams to conform resultant figure is less than 5%, the dose output consistency is them to the shape of the tumour. acceptable. Constancy 195 Contained activity A similar method and calculation can be made to calculate, for example the consistency of the kVp (again the lesser the % the better the consistency). The consistency described earlier is a widely used practical term. However, it can be misleading as it actually represents the error of consistency (1% consistency describes an excellent x-ray system, but actually it means that 99% of all exposures will be consistent). The methods to perform various QC measurements are described in detail in the EMERALD materials. Related Articles: Accuracy, Precision C Hyperlink: EMERALD: www .emerald2 .eu Constancy (Nuclear Medicine) A test performed to check the day-to-day reproducibility of a radionuclide dose calibrator. Radionuclide dose calibrator constancy is assessed by per- FIGURE C.87 Interference from two point sources emitting forming a daily measurement of a long-lived radioactive ‘check’ continuously. source on each isotope setting in use. The reading obtained from the check source should be initially established for each isotope setting at commissioning. The daily reading for each isotope should not deviate more than 5% from the decay-corrected baseline reading. Large, random variations in dose calibrator constancy may indicate instability in the electrometer, whereas a rise or fall in chamber response over time may indicate a leak in the pressurised ionisation chamber or electrometer drift. The ideal check source should have a long half-life and an absence of any significant radioactive impurities. Typically, Cs-137 (30-year half-life) or Ra-226 (1600-year half-life) sources are used. Related Article: Radionuclide dose calibrator Further Reading: Measurement Good Practice Guide No. 93 Protocol for Establishing and Maintaining the Calibration of

Medical Radionuclide Calibrators and their Quality Control, National Physical Laboratory, May 2006. Constructive interference (Ultrasound) Two waves that travel together can, dependent on FIGURE C.88 A motion compensating, fully refocused 3D SSFP-fid their respective phase, be observed as a wave that is the sum of image of the internal auditory canal with normal findings. the two waves’ individual amplitudes (constructive interference), or add up to no apparent wave motion at all if the amplitudes are equal and the phases are opposite (destructive interference). In Figure C.87, two point sources emit continuous waves, and in Contact therapy certain directions the waves appear to be in phase, whereas in (Radiotherapy) Contact therapy describes treatment with low- others, to be out of phase. As can be deduced from the figure, energy kilovoltage x-ray beams (accelerating potential 40–50 this depends on the difference in distance from the observation kV). The SSD for these treatment fields is typically a few centi- point to the respective sources. Distance differences that corre- metres or less. Contact therapy is used for superficial treatment spond to an integer number of wavelengths result in constructive depths of around 2 mm. interference. Abbreviation: SSD = Source to skin distance. Related Article: Superficial therapy Constructive interference steady state (CISS) (Magnetic Resonance) The CISS pulse sequence was constructed Contained activity to reduce the banding artefacts that might appear in balanced (Radiotherapy, Brachytherapy) Calibration of source strength SSFP pulse sequences (true-FISP). The idea is to collect two true- is a very important part of a comprehensive brachytherapy FISP pulse sequences where one of them is acquired with a phase quality system. The instruments, ion chambers and electrom- cycling scheme of the RF pulse, i.e. 20x − 20−x – 20x – 20−x – 20x eters, used for source strength determinations, should have …, and the other is not, i.e. 20x – 20x – 20x – 20x – 20x. calibrations that are traceable to national and international In Figure C.88, a motion compensating, fully refocused 3D standards. SSFP-fid image of the internal auditory canal and the inner ear in Specification of Source Strength for Photon-Emitting a patient with normal findings is shown. Sources: Source strength for a photon-emitting source can be Related Articles: Fast imaging with steady-state precession given as a quantity describing the radioactivity contained in the (FISP), SSFP, Steady-state free precession source or as a quantity describing the output of the source: Continuous wave laser 196 Continuous slowing down approximation 1. Specification of contained activity counter may be necessary. Note that if a whole body counter is a. Mass of radium; mg Ra not available, a scintillation camera without the collimator may be b. Contained activity; Ci, Bq used in special cases (depending on the background). 2. Specification of output Further Readings: Council Directive 96/29/EURATOM of a. Equivalent mass of radium; mg Ra eq 13 May 1996, Laying down basic safety standards for the pro- b. Apparent activity tection of the health of workers and the general public against c. Reference exposure rate the dangers arising from ionizing radiation; IAEA Basic Safety d. Reference air kerma rate Standards 1996. e. Air kerma strength C Contamination monitoring Contained activity is a quantity that can be used for all types of (Radiation Protection) Contamination monitoring seeks to deter- brachytherapy sources. mine if radioactive contamination has occurred, and if so, where For brachytherapy dosimetry, the quantity of interest is the and how much. The contamination may be to a person, or to a output of the encapsulated source, not the contained activity. surface (wall, floor, furniture, etc.). Contamination monitoring is (Sources are encapsulated, and it is thus difficult to determine the carried out using an appropriate monitor, sensitive to the radio- contained activity.) The quantity apparent activity, which is an isotopes in use in the area. For contamination involving gamma- output specification, has been used as an alternative to contained emitting radionuclides, a scintillating crystal–photomultiplier activity and it is still used, especially for radiation protection detector is required. For contamination involving medium- to applications. high-energy beta-emitting radionuclides, Geiger–Müller detector In modern brachytherapy dosimetry, reference air kerma rate is required, while for low-energy beta and alpha particles, con- or air kerma strength is the quantity used to calculate absorbed tamination monitoring must be carried out using a wipe test. dose. Contamination monitoring should be carried out after any See Source strength for a full description of specification of work with unsealed sources, and the results recorded. source strength. Related Article: Wipe test Related Articles: Source strength, Mass of radium, Equivalent mass of radium, Apparent activity, Reference air kerma rate, Air kerma strength Contingency plan (Radiation Protection) Whenever ionising radiation is used in applications that can create hazard for the workers and the popu- Continuous wave laser lation, it is essential to prepare contingency plans in order to mini- (Non-Ionising Radiation) See Laser output mode mise the possible risk and provide a prompt response in case of incidents/accidents. Contamination In general, contingency plans mean comprehensive systems (Nuclear Medicine) In nuclear medicine, a radioactive contamina- of emergency response capabilities, which involves the national tion can refer to an unintentional or accidental spill of radioactive authorities and promote overall coordination among the hierar- material on to any material, surface, environment or individual chy of emergency response organisations at regional and local during a working operation, for example labelling, injection, tak- level. Therefore, depending on the specific applications, contin- ing of specimens or blood samples. In the specific case of the gency plans might include radiation protection measurements for human body, this radioactive contamination includes both exter- people and environment, handling of health effects, managing of nal skin contamination and internal contamination, irrespective radioactive material, managing of radioactive waste, cleaning up of route of intake. of sites etc. If a radionuclide used in nuclear medicine is accidentally Although there are unfortunate examples of major accidents spilled, the material could be spread and spoil measurements of related to medical radiation sources, usually, in medical applica- other radioactive samples or worse constitute an undesired source tions, contingency plans should be handled at local and regional of exposure. A surface contamination is expressed in units of level. activity per unit of area, Bq/cm. The national legislation normally Starting from the basic, there must be a clear identification of states the maximum permissible levels of exposure and contami- the major principal causes of incidents/accidents, with clear iden- nation, i.e. the amount of activity per unit area depending on the tification of roles and competences. Contingency plans include radiotoxicity of the radionuclide. the detection of the event, the involvement of qualified staff, the When checking a workbench in a laboratory using radionu- support of laboratory services and eventually other structures. clides, a simple hand-detector giving the count rate may be ade- In medical applications, priority should be given to the safety quate. However, if the radionuclide is unknown, a more accurate of the patients. assay is needed, for example a gamma spectroscopic analysis. Contingency plans shall be studied in detail by the radiation Monitors showing the activity per unit of surface (Bq cm2) are protection specialists and tailored to each equipment/structure, available from various companies. depending on the kind of possible risk. An internal contamination denotes that a radioactive con- Hyperlink: IAEA: http://www .iaea .org tamination has entered the body through ingestion or inhalation. This will contribute to radiation dose of the individual. It is very Continuous slowing down approximation important to use personal protective equipment, such as gloves (Radiotherapy) The various types of interactions, which a and a white coat when working with radiopharmaceuticals. After charged particle experiences during its passage through the a completed operation, the working bench should be monitored medium, produce a gradual loss of energy, bringing the par- with respect to surface contamination. For the measurement of a ticle eventually to stop. The track length distributions and dose whole body content of activity, counting in a special whole body deposition of a charged particle interacting with matter depend Continuous spectrum 197 Contract management on the approximation employed in defining the transport pro- cess. A common approximation in use is the continuous slowing down approximation (CSDA). In CSDA, the charged particles are assumed to lose their energy continuously at a rate that is given by the stopping power S(T), where T is the particle kinetic energy. In CSDA, it is supposed that no knock-down electrons (δ particles) or Bremsstrahlung photons are created. The kinetic energy lost is locally converted into imparted energy. The track length distribu- tion y(T) takes the following form: C y(T ) dT dT = S ( T ) When the fluences ΦT are calculated using the CSDA approxima- tion, the absorbed dose is obtained by ¥ FIGURE C.89 Continuous (white) spectrum – dotted line – inside the S ò (T ) D = FT dT x-ray tube supplied with 100 kV anode voltage. (Courtesy of Sprawls r Foundation.) 0 The CSDA inadequately describes the electron transport. In the healthcare facility (often including clinical engineers that oversee extreme case of a head-on collision between electrons, all of the the whole activities), a team of technicians from an independent incident energy will be given to the knock-on electron. Because service organisation (ISO) (providing specialised services such as of the lack of identity of electrons, it is usual to identify the elec- preventive maintenance and safety checks, or full-risk service on tron with lower energy after the collision as secondary electron. a subset of the whole inventory), a list of service contracts with An electron may therefore lose as much as half of its kinetic the different OEMs for each brand and model of equipment (often energy in a single atomic collision and the statistical fluctuations covering high-end diagnostic imaging equipment, or high-risk in the energy losses are significant. The generation and transport devices like anaesthesia machines). of high-energy secondary electrons (δ electrons) was taken into Whatever the model implemented, when some or all technical account by the Spencer-Fano model (Spencer and Fano, 1954). activities are outsourced to an ISO or to several OEMs there is An energy limit D˙ is chosen in this model. Energy losses above the need to manage the related contracts. Contract management D˙ are considered to result in the production of δ electrons while is therefore a critical activity to be performed in order to make those below represent the imparted energy to matter. The initial sure that the contractual obligations are met, and the technical kinetic energy of the δ electron equals the energy lost by the pri- services delivered are in line with the quality levels defined by the mary electron, so no energy is imparted to matter in this process. healthcare structure. Abbreviation: CSDA = Continuous slowing down Contract management can have slightly different meanings approximation. based on the overall model implemented by the hospital to man- Related Articles: Mass collision stopping power, Mass stop- age their medical equipment: ping power, Electron stopping power Further Reading: Spencer, L. V. and U. Fano. 1954. Energy • In a fully outsourced multi-vendor service, the hospi- spectrum resulting from electron slowing down. Phys. Rev. tal has to manage the complex contract with the single 93:1172–1181. ISO in charge of technical activities. KPIs are usually defined and periodic reports must be delivered to the Continuous spectrum hospital management. (Diagnostic Radiology) A continuous spectrum is one in which • In a mixed-model where the in-house staff does most of there are photons at all energies within the range of the spec- the work and/or an ISO is in charge of a subset of the trum (up to the maximum of the supplied anode-cathode high inventory and other equipment are contracted with the voltage). The Bremsstrahlung process produces a continuous OEMs, each of the OEM contracts must be managed x-ray spectrum (also called white spectrum) inside the x-ray and this activity is sometimes part of the outsourcing tube (Figure C.89). After filtration of the continuous spectrum – contract with the ISO. The technical staff of the hos- absorption by the glass envelope of the x-ray tube and the ele- pital (or the ISO) receives service calls from the user ments of the tube housing – the spectrum loses a significant part departments and have to redirect them to the proper of the low energy photons. OEMs; the arrival

of the OEM technicians has to be Related Articles: Bremsstrahlung, X-ray tube recorded as well as the activities performed; invoicing and payments have to be done according to the contrac- Contract management tual obligations. (General) The management of medical equipment in a healthcare institution can be organised through different models; the most Contract management also includes placing the proper calls to common are usually referred to as in-house service, multi-vendor the OEMs when an equipment that is covered with warranty, service and mixed model. A combination of these models is often or is in the hospital under a service or a leasing/rental contract, implemented, with the presence of in-house staff employed by the is involved; in these cases the related technical activities are Contralateral 198 Contrast agent performed by the OEMs despite the presence of an ISO or an in- tissues – usually the T1 and/or T2 relaxation times. The most com- house structure. monly used relaxation agents are metal-chelate complexes con- A critical component of the overall contract management taining lanthanide ions, predominantly gadolinium. There are activity is the presence of a computerised maintenance manage- also agents based on small particles of paramagnetic iron oxide, ment system (CMMS) that properly records the whole medical which operate primarily as T2-shortening agents. These particles equipment inventory and all related technical activities, includ- may be bound to receptor-specific carbohydrates or to antibody ing the contractual coverage for each equipment (under warranty, fragments for enhanced specificity, for example in hepatic or leased/rented, covered with an ISO contract, managed by the lymph node imaging. ISO, etc.). These early examples of targeted agents are leading to the C development of MRI molecular imaging agents, such as αvβ3- Contralateral targeted paramagnetic nanoparticles to image angiogenesis. Such (General) Directional anatomical terms describe the relationship studies may make use of nuclei that are not present in the body of structures relative to other structures or locations in the body. in significant quantities, such as fluorine, blurring the distinction ‘Contralateral’ means on different sides of the midline. The between contrast agents and tracers. right shoulder and left hip are contralateral to each other. Oral contrast agents are designed to manipulate signal Related Article: Anatomical relationships from the bowel, either to enhance it (gadolinium agents, ferric ammonium citrate or oil emulsions) or, more commonly, to sup- Contrast press it (iron oxide beads, carbon dioxide, barium sulphate or (General) For the human eye, contrast is a measure of the abil- perfluorochemicals). ity to distinguish between two adjacent objects or an object and For vascular imaging, conventional gadolinium agents are the background. In a digital image, the contrast is defined by the frequently used, but are limited by their relatively rapid extrava- signal difference between the two objects. A common goal for all sation. A number of longer lasting blood pool agents have been disciplines in medical imaging is to increase the contrast between developed, and such agents based on gadolinium are now entering an object of interest and the surrounding tissue so that any patho- clinical use. logical tissue or bioprocess becomes evident to the radiologist. In Gadolinium-based agents are normally used in conjunction diagnostic radiology and MR, one uses contrast agents (or con- with T1-weighted imaging, with areas of high uptake appear- trast media) to increase the difference in acquired signal. ing bright due to T1 shortening. Iron oxide–based agents, on the other hand, produce dark regions on T2-weighted images due to Contrast agent T2 shortening. These are known colloquially as ‘hot’ and ‘cold’ (Magnetic Resonance) In medical imaging, a contrast agent is an spots, respectively. exogenous substance introduced into the body by some means Related Articles: Gadolinium, Negative contrast media, (e.g. by injection or orally) in order to modify the signal from cer- Paramagnetic contrast agents, Positive contrast media, USPIO tain tissues or structures, hence making them more conspicuous (ultra small particles of iron oxide) in the acquired image. Contrast agents are therefore chosen for their ability to modify a property of the tissue that has an impact Contrast agent on its appearance in the image, as well as for their suitability for (Ultrasound) use in vivo in terms of safety and practicality. Background: Contrast agents for ultrasound are a fairly new In x-ray imaging, the most commonly used intravenous con- aid to improve the diagnosis, used in cardiology, and for iden- trast agents are iodine based. Barium-based contrast agents are tification and classification of tumours, especially in the liver. usually used orally (e.g. barium meal used for x-ray examination Essentially, the contrast agents are microbubbles with a size of the hollow gastrointestinal tract). The high subject contrast in that is less than the capillary diameter (i.e. a mean size of ∼2 these cases requires use of higher energies (kVp). μm). The microbubbles need to be stabilised by a shell, which In MRI, the ability to modify image contrast by changing usually is composed of lipids or proteins. The gas is also pref- acquisition parameters and by tailoring the pulse sequence to pro- erably one that has low diffusivity, such as perfluorocarbons or duce different weightings arguably reduces the need for exoge- sulphurhexafluoride. nous agents as compared, for example to x-ray imaging. However, Acoustic Properties: A free gas bubble in a fluid acts as a there are a number of situations in which MRI contrast agents are mass-and-spring resonance system, where the gas pressure acts useful, and their use has become widespread. as the spring and the displaced fluid the mass. Consequently, the Applications of contrast agents in MRI include the following: bubble has a resonance frequency given by • Enhancement of signal from selected tissues or areas of 1 3gp wres = R (C. ) 0 r 5 pathology, usually on the basis of vascularity • Enhancement of signal from flowing blood in contrast- enhanced MR angiography where • Elimination of signal from organs that might otherwise ωres is the resonance frequency obscure structures of interest or cause artefacts, such R0 is the equilibrium radius of the bubble as the bowel γ is the ratio of specific heat of a gas at constant pressure and • Dynamic studies of perfusion or function, sometimes at constant volume for tissue characterisation p is the hydrostatic pressure outside the bubble • In emerging molecular imaging techniques ρ is the density of the surrounding fluid (Leighton, 1994) To be useful as a contrast agent for MRI, a material must Consequently, a 2 μm air bubble in blood will resonate at have some effect on one or more of the NMR properties of 3.26 MHz, in perfect range for diagnostic ultrasound. This Contrast degradation factor 199 Contrast detail dramatically increases the detectability of small gas bubbles with ultrasound. The drawback is that small bubbles dissolve in a mat- ter of milliseconds, due to surface tension, which is why contrast bubbles need to be stabilised by a shell. Detection Principles: An interesting property of bubbles is that they behave strongly non-linear to acoustic stimuli, already at relatively small acoustic pressures. The basic underlying mecha- nism is that the bubble tends to expand more easily than being compressed. Thus, harmonics of the incident ultrasound wave are easily produced. One of the first approaches was to simply transmit at one frequency, but have the detection at twice that C frequency (harmonic imaging). This assumes a transducer that is wideband enough, and also that the pulses need to be fairly nar- rowband in order to avoid spectral leakage from the fundamental to the harmonic. In order to increase the bandwidth of the pulse, and thereby the spatial resolution, a detection scheme was devel- oped where two pulses of opposite polarity are transmitted (com- FIGURE C.90 Indicative example of scattered radiation in abdominal monly referred to as pulse inversion). For a linear scatterer, the radiography. (Image courtesy of P. Sprawls.) second received signal will be identical to the first but inverted. On the other hand, a non-linear scattering object will produce sig- nals that are not similar. When adding the two received signals, linear echoes will cancel while non-linear give rise to a detectable signal. Several variants of this scheme are now in use, notably one where the pulses still have the same phase, but different ampli- tude. As a general rule, detections schemes are then divided into single-pulse (such as harmonic imaging) and multi-pulse (such as pulse inversion) methods. Related Articles: Contrast media, Echo-enhancing agent, Microbubbles Further Reading: Leighton, T. G. 1994. The Acoustic Bubble, Academic Press, San Diego, CA. Contrast degradation factor (Diagnostic Radiology) A significant amount of scattered radia- tion in x-ray radiography leads to degradation of image contrast. The contrast degradation factor (CDF) indicates the amount of this decrease of contrast. FIGURE C.91 Indicative example about the influence of scatter to radi- The scatter-to-primary ratio (S/P) for many radiographic ography contrast. (Image courtesy of P. Sprawls.) examinations is above 2 (e.g. for lateral lumbar spine radiography S/P can be above 5). Smaller anatomical regions with less scatter have smaller S/P (e.g. for mammography the CDF is of the order Hyperlinks: www .sprawls .org /ppmi2 /SCATRAD/ of 0.5) (Figure C.90). The radiation reaching the radiography detector includes the Contrast detail useful primary radiation (modulated by the attenuation of vari- (Diagnostic Radiology) The visibility of an object in an x-ray ous anatomical structures) plus the scatter radiation which does image depends on both its physical contrast and its size (detail). not bring diagnostic information and reduces image contrast. The reduction in visibility relating to an object’s physical contrast Although anti-scatter grids significantly reduce the percentage of is caused by limitations of the contrast sensitivity of the imag- scattered radiation, part of it reaches the image receptor and dete- ing process and by the presence of visual noise. The reduction of riorates the quality of the image by decreasing the visible contrast visibility relating to object size (detail) is because of the blurring in the radiograph (Figure C.91). that occurs during the imaging process. The visibility, especially The CDF describes the decrease of useful contrast due to the for small objects that also have relatively low physical contrast, presence of scattered radiation. because they are thin in the direction of the x-ray beam, is affected by a combination of the three factors; contrast sensitivity, Cs = Co ×CDF blurring and noise. Test devices that contain objects that have a combination of different physical contrasts (object thickness) and Cs – Contrast with scatter detail (size) are used to evaluate imaging procedures. An image Co – Original contrast (no scatter) of the device is produced and it is then determined which objects CDF, presented as a function of S/P, is always smaller than 1. are visible. The contrast detail diagram is a graph or curve plotting the CDF = 1 (1+ S / P) ‘just visible’ objects using (physical) contrast versus detail (size) scales. It is a qualitative method combining both the concepts Related Articles: Coherent scattering, Elastic scattering, Grid of contrast resolution and the spatial resolution. The assessment efficiency, Contrast improvement factor Contrast detail (C-D) studies 200 Contrast enhancement (observation) aims to detect the minimal contrast necessary to with 15 cm diameter (zoomed image); a new II with 17 cm diam- visualise an object with certain size. These minimal (limiting) eter. It is obvious that the third curve is closer to the x–y axes and contrasts are then presented against the object size, thus forming represents a system with better imaging characteristics. The dia- a diagram with contrast on the y-axis and detail size on the x-axis. gram shows that using the third II (17 cm), one needs 0.35 (35%) The diagram divides the space into two parts – visible (above the contrast in order to see a detail with 0.5 mm size. Respectively, curve) and invisible (under the curve). If the curve is closer to if the first II (25 cm) is used, such detail can be seen only if its the x–y axes, this indicates better contrast detail for certain sys- contrast is 0.67 (67%). tem (larger visible part). Also, the curve shows that small details Contrast detail and contrast resolution concepts can be require high contrast in order to be visible. described also with the ratio C Figure C.92 shows a test object used to examine contrast and detail of an x-ray system

(usually fluoroscopic system). The test (C * D) object is a Leeds TO10. It includes a number of rows, each with = const N identical detail sizes. Within each row, the contrast gradually decreases (by gradually decreasing the absorption of the objects). where The observer has to identify the object in a certain row, where its C is the contrast resolution image cannot be distinguished anymore from the background (the D is the detail diameter limiting contrast). The value of this contrast (taken from a table N is the noise level accompanying the test object) is then plotted against the diameter of the object. Related Articles: Contrast resolution, Spatial resolution Figure C.93 shows typical contrast detail diagrams of three Hyperlink: EMERALD: www .emerald2 .eu image intensifiers (II) – old II with 25 cm diameter; same old II Further Reading: Oppelt, A. ed. 2005. Imaging Systems for Medical Diagnostics, Siemens, Erlangen, Germany. Contrast detail (C-D) studies (Diagnostic Radiology) See Contrast detail Contrast-enhanced angiography (Magnetic Resonance) Blood vessels can be visualised with MR in several ways (magnetic resonance angiography, MRA). One method uses contrast-enhancing agents (contrast-enhanced MRA, CE-MRA). The contrast agents involved are paramagnetic, and affect the magnetic properties of the tissue they come in con- tact with. When injected into the blood stream, the contrast media reduces the T1 of blood, and CE-MRA exploits this difference in T1 between tissue and blood to visualise the blood vessels while suppressing signal from the background tissue. A CE-MRA sequence is most often a 3D gradient-echo sequence with a short TR, to suppress the signal from background tissue with its relatively long T1. A short TE is used to minimise T2* effects due to the contrast agent. As the high signal in the vessels is primarily due to the presence of contrast media, blood FIGURE C.92 Test object Leeds TO 10 (imaged with a good-quality with low velocity is enhanced, as well as fast flowing blood. This new x-ray fluoroscopic system). is an advantage when investigating veins, or aneurysms and other pathologies. The data are very often post-processed with maxi- mum intensity projection (MIP) to obtain a 3D overview of the 1 vessel tree. As there is a time delay between the injection of contrast agent 0.9 25 cm old II 15 cm old II and the arrival of this agent to the region of interest, a timing 0.8 17 cm new II method has to be used. It can consist of a test bolus, which is 0.7 timed, and the measured delay time entered into the CE-MRA 0.6 sequence. It can also be in the form of a real-time view of the 0.5 area of interest, where the CE-MRA is interactively started as the 0.4 contrast reaches the vessels to be studied. With fast sequences, it is possible to retrieve data for different 0.3 time delays, as is seen in Figure C.94. The second image is taken 0.2 at a later time than the first one. 0.1 CE-MRA is clinically used for a variety of applications, for 0 example to examine vessels in the head–neck, thorax, abdomen 0 0.5 1 1.5 2 2.5 3 3.5 4 or extremities. Detail (mm) Contrast enhancement FIGURE C.93 Typical contrast detail diagrams of three image intensi- (Diagnostic Radiology) Contrast enhancement is a process that fiers (maximal contrast of 1 = 100%). is used to increase the contrast and visibility of specific organs, Contrast Contrast-enhanced mammography 201 Contrast inversion C FIGURE C.94 MIP of a CE-MRA over the thorax. The right image is taken at a later time point than the left one. During this time, the contrast has reached the kidneys, liver and large veins. vessels or tissues in medical imaging. A variety of contrast agents are used to enhance the physical contrast within the body. X-ray spectrum optimisation and techniques such as digital subtraction angiography (DSA) are used to enhance contrast dur- ing the acquisition of x-ray images. Digital processing is used to enhance and optimise contrast, especially in digital radiography. Contrast-enhanced mammography (Diagnostic Radiology) Contrast-enhanced mammography (CEM) is an advanced breast imaging technique that provides addi- tional information beyond a normal mammogram (Figure C.95). A finite number of sequential images are obtained with standard x-ray mammographic equipment, operating at energy above the K-edge of iodine, and with an intravenous iodine-based contrast agent injected between pre- and post-contrast image acquisitions, to show the flow of the contrast agent with time. As the rapidly growing tumours require an increased supply of blood to sus- tain their growth, the contrast agent is selectively accumulated in the regions exhibiting a rapid wash-in and wash-out of iodine in malignant tissue. For benign tissues, they have a slower uptake of the iodine over the study duration. CEM takes approximately 8–10 minutes to perform. Related Articles: Mammography, Contrast agent, Full-field FIGURE C.95 (a) Standard mammogram (b) 2D contrast-enhanced digital mammography mammogram (CEM) of the left breast which shows an abnormal area with iodine enhancement. It was subsequently proven to be an early breast cancer (ductal carcinoma in situ). (Courtesy of Dr Ma. Theresa Buenaflo.) Contrast improvement factor (CIF) (Diagnostic Radiology) The contrast improvement factor (CIF) is the ratio of the contrast when a specific grid is used compared Contrast inversion with the contrast without the grid. It is a function of the grid char- (Diagnostic Radiology) The phenomenon of contrast inversion acteristics and the amount of scattered radiation from the patient’s can be seen during quality control (QC) procedures. It exists in body. anatomical images but is difficult to detect visually. The image As an example, CIF values can be 2.4, 3.3 etc. on Figure C.96 shows a test object (phantom) with a typical res- CIF is one of the measures of efficiency of anti-scatter grids. olution pattern. With the increase of spatial frequency, one can The other methods to measure this are the Bucky factor (the ratio clearly observe contrast inversion. Each spatial frequency is rep- of the radiation entering the grid to the radiation passing through resented by a group of three dark lines and two less dark spaces. the grid) and the primary transmission factor (the percentage of In an image without significant blurring the three dark lines and primary radiation passing through the grid). spaces can be visualised. However, at spatial frequencies where Contrast media 202 Contrast media C FIGURE C.96 The phenomenon of contrast inversion observed in the two highest frequencies patterns/bars of the test object. The right site of the test object image (with low frequencies) is cropped to allow better zoom of the phenomenon (explained on figure: before and after its manifestation – see the arrows). contrast inversion is present the image will show two dark lines spot size of x-ray tubes. It can be related to other ‘imperfections’ between three less-dark lines. The spatial frequencies at which of the imaging system. The contrast inversion can present a false contrast inversion occur are determined by the characteristics of image of small objects with larger magnification (i.e. far away the blurring associated with the specific imaging procedure. from the detector). The phenomenon of contrast inversion is due to the fact that For image quality assessment this contrast inversion is not the modulation transfer function (MTF) is a decaying sinusoidal considered, as the MTF is presented with the modulus of the function (sinc function), and only its initial positive part repre- Fourier transformation of the line spread function. sents the ‘normal contrast’ in an anatomical image, the other Related Articles: MTF parts of the sinc function exist (Figure C.97). Further Reading: Tabakov, S. 2018. A case study of contrast The phenomenon of contrast inversion is most obvious when inversion and modulation transfer related to the finite X-ray tube it is related to the blur arising from the size of the effective focal focal spot size. J. Med. Phys. Int. 6(2):314–317. Contrast media (Nuclear Medicine) In medical settings, a contrast media (or contrast agent) is administered to achieve a higher contrast dif- ference between a structure or blood flow and the surrounding matter. In conventional x-ray imaging, contrast media can be used to enhance the visibility of blood flow, for example in ventilation and perfusion lung studies or in angiographic investigations. A contrast media with high atomic number relative to the sur- rounding matter increases the photon attenuation; hence, regions with high accumulation of contrast media will appear darker compared to an identical situation without the contrast media. As an example, iodine-based contrast media is used for x-ray angio- graphic examinations. Gadolinium (Gd) is used as a contrast media in MR because the presence of Gd molecules induces a quicker relaxation in adja- cent water molecules. Even though the examples are few, the contrast media has induced allergic reactions in some patients and the use of such should therefore be under supervision. FIGURE C.97 A real MTF sinc function with contrast inversion areas, Contrast media visualising positive (true) and negative contrast (contrast inversion). At the (Diagnostic Radiology) A number of anatomical structures, ves- inflection points the contrast of the object bars disappears, i.e. limiting sp. sels and cavities in the body are similar in density and atomic freq. after area A – see Figure C.96. number (Z) to surrounding tissues and do not produce sufficient Contrast media 203 Contrast resolution contrast for imaging. Contrast media is used to enhance the radio- Contrast ratio of monitors logical contrast of these anatomical structures. (Diagnostic Radiology) The contrast ratio of diagnostic display There are two main types of contrast media (or contrast agents) monitors is one of their most important characteristics. The used in x-ray diagnostic radiology: contrast ratio is the ratio between the luminance (cd/m²) of the brightest and the darkest area within an image that the monitor 1. Positive contrast media (most often used) which are can produce. substances with high attenuation. Examples of these The static contrast ratio is the one which is normally used are iodine (e.g. iodine-based solutions injected into for assessment of monitors – this is the ratio at a specific time, the blood to enhance the visibility of blood vessels and measured simultaneously in the brightest and darkest areas of the organs) and barium (e.g. barium meal used to enhance monitor. For a good monitor this contrast ratio is of the order of the visibility of the gastrointestinal tract). 1000:1 or higher. C 2. Negative contrast media are substances with attenu- The dynamic contrast ratio is measured at different times (and ation lower than that of the soft tissues. Examples of sometimes with different settings). Such measurements, made by these are air and other gases. the manufacturers, can reach 1 million to 1, but the user cannot repeat these measurements as their parameters are usually not When using positive contrast media the contrast can be enhanced by known. administering materials, contrast media, that have higher x-ray atten- Related Articles: Brightness, Digital display uation characteristics. Two of the most often used contrast media are based on barium and iodine. Figure C.98 shows the main absorption Contrast resolution parameters of the human tissues and of barium and iodine. (Diagnostic Radiology) Contrast resolution is a characteristic of Barium and iodine produce high contrast with respect to soft an imaging process relating to the ability to see, or resolve, small tissue because of their densities and atomic numbers. The sig- differences in contrast (low contrast) of objects or areas within an nificance of their atomic numbers (Z = 53 for iodine, Z = 56 for image. The minimal (limiting) contrast to be seen in one imaging barium) is that the K-absorption edge is located at very favourable system is highly related to the maximal noise in this system. This energies relative to the typical x-ray energy spectrum. The K-edge way contrast resolution can be quantitatively described through for iodine is at 33 keV and is at 37 keV for barium. the signal to noise ratio (SNR) or noise power spectrum (NPS) Since the typical x-ray beam contains a rather broad spectrum of the system. This SNR measurement is especially suitable for of photon energies, all of the energies do not produce the same digital imaging systems. level of contrast. In practice, maximum contrast is achieved by A simple presentation of contrast resolution is the minimal adjusting the kV so that a major part of the spectrum

falls just (limiting) contrast, which the system can discriminate. For exam- above the K-edge energy. For iodine, this generally occurs when ple, a computed tomography (CT) system discriminates contrast the kV is set in the range of 60–70. (density difference) ∼0.25%, while a classical x-ray radiography Related Articles: Contrast enhancement, Negative con- system discriminates ∼10% contrast. trast media; Digital subtraction angiography; Power injector; Various imaging systems have specific ways to measure their Abdominal imaging, Fluoroscopy intrinsic contrast resolution – for example, CT densitometry can Further Reading: Thomas, A. 2020. The history of contrast be used in CT; Pulse overlap separation can be used in ultrasound; media development in X-ray diagnostic radiology. J. Med. Phys. contrast detail diagram can be used in x-ray fluoroscopy (qualita- Int., Special Issue History of Medical Physics 3, March 2020, tive measurement), etc. www .mpijournal .org /history .aspx. Contrast resolution (and contrast detail) concepts can be described also with the ratio Contrast media (Ultrasound) Microbubbles are used as contrast media in ultra- sound examinations. These are micron-sized bubbles, stabilised (C*D) = const by a shell. Further details are found under Contrast agents. N where C is the contrast resolution D is the detail diameter N is the noise level This constant is maintained for a wide range of diameters and is independent of the spatial resolution of the imaging system. Related Articles: Contrast detail, Signal to noise ratio, Noise power spectrum, Spatial resolution Further Readings: Bushberg, J. T., J. A. Seibert, E. M. Leidholdt and J. M. Boone. 2002. The Essential Physics of Medical Imaging, 2nd edn., Lippincott Williams & Wilkins, Philadelphia, PA; Oppelt, A. ed. 2005. Imaging Systems for Medical Diagnostics, Siemens, Erlangen, Germany. Contrast resolution (Ultrasound) In medical imaging generally, the term contrast res- FIGURE C.98 Physical characteristics of contrast-producing materials. olution describes the ability of a modality to display differences (Courtesy of Sprawls Foundation.) in tissues resulting from the different properties of those tissues. Contrast scale 204 C ontrast threshold The ability to discriminate between different tissues is key to the referred to a film with narrow latitude. Such contrast can also be performance of the imaging system. called high contrast or hard contrast. For ultrasound, changes in acoustic properties at a boundary These qualitative practical terms can often be confusing. In cause reflection from that boundary. The brightness is dependent principle, the contrast scale and the degree of image contrast on the difference in acoustic impedance, and size relative to the depend on the absorption of x-rays by the individual structures/ wavelength and orientation relative to the beam. Within tissue, organs. This way, it is a function of the x-ray spectrum, the organs the characteristic brightness is dependent on the scattering within attenuation, the x-ray film characteristics etc. the tissue. The perception of differences in acoustic properties in In digital imaging, the contrast scale directly depends on the an ultrasound scan is complex and is dependent on, for example parameters of the window width (WW). Large window produces C long-scale contrast or soft contrast, while narrow window pro- The magnitude of differences duces short-scale contrast or hard contrast. The instrument settings used (e.g. power, gain, dynamic For illustrations to the preceding text, see the images in the range) article, High contrast. The size of the target Related Articles: Subject contrast, High contrast, Latitude, The presence of artefacts (e.g. speckle) Window Whether defined edges are present Further Reading: Thompson, M., M. Hattaway, D. Hall and Viewing conditions S. Dowd. 1994. Principles of Imaging Science and Protection, W.B. Saunders Company, Philadelphia, PA. Tests of contrast resolution are difficult for the same reasons. Figure C.99 shows an ultrasound phantom, which contains large Contrast scale targets of different scattering levels. The same phantom has high- (Ultrasound) A contrast scale shows a range of grey levels contrast cyst-like targets, which become difficult to visualise from white to black typically divided into 16 discrete levels when small, illustrating the influence of target size on contrast. (Figure C.100). The level of grey can be described as a percent- On Figure C.99, the +3dB and −3dB targets are difficult to age, for example 100% as white, 0% as black or vice versa for a visualise and the fine speckle pattern contains significantly higher greyscale image printed on white paper. variation, which detracts from the larger target. The high-contrast Contrast scales are used as an aid in setting up monitor bright- cyst-like targets (white box) of 4, 3 and 2 mm diameter are not ness and contrast so that the full range of greys in the processed seen throughout the image; the 2 mm targets are not imaged at image can be displayed for optimum contrast resolution. depth due to poor spatial resolution and speckle. Related Article: Contrast resolution Related Articles: Spatial resolution, Phantom, Speckle, Dynamic range Contrast sensitivity (Diagnostic Radiology) Contrast sensitivity is a characteristic of Contrast scale imaging systems, including the human visual system, which deter- (Diagnostic Radiology) The contrast scale is a term related mines the lowest contrast objects that are visible (Figure C.101). to the subject contrast in x-ray imaging (but also in any other The contrast sensitivity of a specific imaging procedure is imaging). The contrast scale of a black–white x-ray image refers determined by the physical principles of the process (CT and radi- to the number of visibly distinguishable shades of grey in the ography are very different) and the operating factors. The same x-rayed area. contrast sensitivity is not required for all procedures and should Two terms often used in practice are long-scale contrast and be adjusted to optimise visibility for the specific clinical need. short-scale contrast. Usually, long-scale contrast can be referred to a film with wide latitude. Such contrast can also be called low Contrast threshold contrast or soft contrast. Similarly short-scale contrast can be (General) Contrast threshold is the smallest contrast in luminance/ brightness that is perceptible to the human eye. It is dependent on FIGURE C.100 The contrast scale shows 16 levels from white to black. FIGURE C.99 Ultrasound phantom with cylindrical targets of differing By altering screen brightness and contrast, the user can ensure that the scattering levels (box to the right). different grey levels are differentiated in the displayed image. Contrast to noise ratio (CNR) 205 Contrast transfer function between 1) the difference in a structure’s signal and the back- Contrast sensitivity ground signal, and 2) the noise within an image. The precise definition of CNR varies, but is most commonly defined by the equation: Bones Low High High barium bullets S CN = object - Sbackground R sbackground Imaging process where Sobject is the signal arising from a structure of interest within an image and Sbackground the signal within the background area sur- C rounding that structure. s Soft background represents noise within the tissues image and is often taken as the standard deviation of the measured Low signal in the background. Sobject and Sbackground will often be taken Object Image as the mean pixel value of regions of interest (ROIs) placed within physical contrast visual contrast the structure and background regions of the image respectively. Sprawls Additionally, CNR may be measured by a comparison of two separately acquired images with and without a structure present. FIGURE C.101 Sketch illustrating contrast sensitivity. (Courtesy of This allows an ROI to be placed in the location of the structure in Sprawls Foundation, www .sprawls .org) both images to accurately determine CNR (as is commonly the case in routine mammography testing). Here CNR is calculated as: the angular size of the target and varies between observers and S over time in an individual. The contrast threshold also depends on 1 - S CNR = 2 s 2 1 + s 2 the expectation and probability that a target will appear. 2 / 2 Related Articles: Contrast scale, Perception where S1 and σ1 are the average signal and standard deviation of Contrast to noise ratio (CNR) an ROI in the image with no structure present and S2 and σ2 the average signal and standard deviation of the same ROI with a (Diagnostic Radiology) The visibility of a specific structure structure present. within an image is not solely related to its measured signal. By Related Articles: Signal to noise ratio (SNR), Contrast to comparison, its contrast (the difference between the structure’s noise ratio (Nuclear Medicine) signal and the signal from the background in which the structure Further Readings: Baldelli, P. et al. 2009. A novel method for lies) and the noise within an image have a greater effect on vis- contrast-to-noise ratio (CNR) evaluation of digital mammogra- ibility (as displayed in Figure C.102). phy detectors. Eur. Radiol. 19:2275–2285; Huda, W. and R. Brad Contrast to noise ratio (CNR), similar to the signal to noise Abrahams. 2015. Radiographic techniques, contrast, and noise in ratio, acts as a measure of image quality. It is defined as the ratio X-ray imaging. Am. J. Roentgenol. 204:W126–W131. Contrast to noise ratio (Nuclear Medicine) The contrast to noise ratio (abbreviated CNR) is a quantification of a medical imaging modality system’s ability to distinguish between structures and noise in an acquired image. CNR can be quantified with equation C.6: (SA - SB ) CNR = Noise (C.6) where SA and SB are the average signal strength in tissue A and B, respectively. The noise is usually measured as the standard devia- tion in a region of interest (ROI) outside any biological structure. Contrast transfer function (Diagnostic Radiology) The contrast transfer function (CTF) is used in diagnostic radiology quality control (QC). It represents the change of contrast as a function of the spatial resolution (contrast % decrease related to increased lp/mm). CTF presents results comparable with those of the modulation transfer function (MTF). These results lack the precision of the MTF but are more easily obtained. MTF has a theoretical base and shows precise results, but requires special hardware and software. A typical example of CTF used in practice will be from QC of a computed tomography (CT) system. Figure C.103 shows a FIGURE C.102 The effect of differing contrast and noise on the vis- bar pattern test phantom with known contrast values (CT num- ibility of a structure in an image. bers, CT#, or HU) of the materials (in the case Perspex with CT# Control button 206 Control rods of nuclear reactor C FIGURE C.105 Control button of a mobile x-ray equipment. FIGURE C.103 CT test object with density profile across the bar pat- terns with increased spatial frequency. (Note the decrease of the contrast amplitude from most right 3.1 lp/cm to the smallest visible detail with operation of the electrical equipment (e.g. x-ray equipment). The 6.25 lp/cm – the fourth one.) push-to-make switch contains a spring, returning the actuator to a certain position. Contact is made when the button is pressed and is broken when the button is released. The exposure control 100 of x-ray system terminates the x-ray exposure when pressure is released from the exposure control button. 90 An x-ray unit may comprise a two-step control button formed 80 of a standby button (e.g. for starting rotation of the anode) and an execution (exposure) button. 70 An example of a control button of mobile x-ray equipment is 60 shown in Figure C.105. Related Articles: Switch, Dead man’s switch, Exposure 50 switch, Foot switch 40 30 Control of Artificial Optical Radiation 20 at Work Regulations, 2010 (Non-Ionising Radiation) The use of optical radiation in medi- 10 cine is widespread and applies to many different specialist areas. 0 Optical radiation applied for medical treatment and diagnosis is 2 3 4 5 6 7 8 9 termed as ‘artificial optical radiation’ and its use in the workplace Lp/cm in the United Kingdom is controlled by The Control of Artificial Optical Radiation at Work Regulations, 2010. Related Articles: Artificial optical radiation (AOR) FIGURE C.104 Contrast transfer function of the test object from Figure Further Reading: Health and Safety Executive, Control of C.103 made with different scanning parameters. Artificial Optical Radiation at Work Regulations 2010, S.I no. 1140. www .l egisl ation .gov. uk /uk si /20 10 /11 40 /pd fs /uk si _20 10114 +120 HU and water with CT# 0 HU, presenting 100% contrast as 0 _en. pdf. +120 CT numbers). The density profile through the CT scan of the phantom will show the maximal resulting

contrast of each bar Control panel pattern (e.g. 108 CT# for the bar pattern with 3.1 lp cm – the first (General) See Control button on the right). This way, plotting the % of contrast decrease against the spatial resolution of each bar pattern will present the CTF. Figure C.104 presents three CTF functions made with different Control rods of nuclear reactor scanning parameters. (Radiation Protection) Nuclear reactors are based on the fission CTF can present indicative practical information about the reaction (n,f), which occurs for heavy nuclei (mass number A relative change of the image quality parameters of an imaging > 220, e.g. 235U, 239Pu, 233U) induced by neutrons. An example system. of this nuclear reaction induced by the thermal energy neutron Related Article: MTF (0.025 eV) is Control button 235 U 1 236 99 135 1 92 + 0 n ® 92 U* ® 42 Mo + 50Sn + 2 0 n (General) A control button (also called as push-button, push- button or simply button) is a switch mechanism for controlling + ~ 200MeVenergy released Contrast % Controlled area 207 Converging collimator Many other nuclides besides those mentioned earlier are also protection and safety procedures. As an example in a radiology produced. facility, all x-ray rooms shall be controlled areas. The energy produced in nuclear fission by 1 kg of uranium is Supervised Areas: Special areas called supervised areas shall about 1014 J (in burning 1 kg of coal is 3 × 107 J). The daily fuel be identified in connection with the controlled areas (including, requirements for 1 GWe power plant are 3 kg of uranium and 8 × as an example, rooms where x-ray mobile units are used), other 106 kg of coal. than public areas. Each room of the facility should only be used The neutrons produced in fission (usually 2 or 3) may evoke for its specified work. another fission and thus initiate the self-sustaining chain reac- Operational Conditions: As the working conditions might tion if the fissionable fuel amount is equal to or greater than its change, it should be determined whether an area will be main- critical mass. The critical mass of the fissile material may be con- tained as controlled or public area; the evaluation is made on the trolled by absorbing neutrons. Cadmium absorbs neutrons very basis of regular, routine, safety assessment (including the planned C efficiently by the nuclear reaction (n, γ), with its extremely large use of each area) and the evaluation of shielding. cross section. Registrants and licensees are also responsible for (1) the con- The most important components for a controlled nuclear reac- trolled area to be delineated by physical (or other suitable) means, tor are (1) fissionable material (located in fuel rods), (2) moderator (2) the warning symbols (such as those recommended by the to slow down neutrons, (3) control rods and (4) radiation shield. International Organisation for Standardisation) to be displayed, In Figure C.106, the schematic cross section of the reactor core (3) the appropriate instructions to be available, (4) the safety rules is shown. The role of the cadmium control rods is to maintain a and procedures (including local rules) to be established and (5) self-sustained chain reaction or to break it off. It is realised by the access to controlled area to be limited in relation to the mag- removing or positioning cadmium rods in the reactor core. nitude and the likelihood of risk. Further Reading: Thornton, S. T. and A. Rex. 2000. Modern Further Reading: IAEA (International Atomic Energy Physics for Scientists and Engineering, Saunders College Agency). 1996. International basic safety standards for protection Publishing, Philadelphia, PA. against radiation and for the safety of radiation sources, Safety Series No. 155. IAEA, Vienna, Austria. Controlled area (Radiation Protection) Classification of Areas: The responsibil- Controlled area (Laser) ity of designating the controlled areas is attributed to the regis- (Non-Ionising Radiation) The laser-controlled area is defined in trants and licensees, who may appoint qualified experts to deal MHRA guidance as ‘the region around the laser where people with the practical work. They shall designate as a controlled area, may be present and in which specific protective control measures any area in which specific protective measures or safety provi- are required’. The control measures required should be decided sions are or could be required for (1) controlling normal expo- based on a risk assessment. sures during normal working conditions and (2) preventing or The controlled area should include the region around the laser limiting the extent of potential exposures. where the maximum permissible exposure (MPE) level or expo- In determining the boundaries of any controlled area, reg- sure limit value (ELV) is exceeded. Often the NOHD of the laser istrants and licensees shall take account of the magnitudes of is used to establish the size of the controlled area required. The the expected normal exposures, the likelihood and magnitude controlled area is usually physically enclosed. For simplicity, in of potential exposures and the nature and extent of the required medical applications the controlled area is often defined as the whole room in which the laser is used. Abbreviations: MHRA = Medicines and Healthcare products Control rods Regulatory Agency and NOHD = Nominal ocular hazard distance. Related Articles: Nominal ocular hazard distance, Exposure limit value (ELV), Maximum permissible exposure (MPE) Further Reading: Medicines and Healthcare Products Regulatory Agency, Lasers, intense light source systems and LEDs – guidance for safe use in medical, surgical, dental and aesthetic practices, Crown copyright, September 2015. Convection (Nuclear Medicine) Convection refers to the transport process of molecules in gases and liquids and it is one of the major methods for heat and particle transfer. Convective heat and mass transfer in liquids are the sum of two separate processes; diffusion and advection. Diffusion refers to the rearrangement of particles due to their individual Brownian motion and advection to the large-scale currents in the fluid. The converse of convective heat transfer is conductive heat transfer where the energy propagates through the fluid via vibration interactions between individual atoms or molecules. Converging collimator Fuel rods (Nuclear Medicine) The holes in a converging collimator con- verge to a point in front of the collimator, typically at a distance FIGURE C.106 Cross-sectional scheme of the nuclear reactor core. of 40–50 cm. With this type of collimator, it is possible to attain Conversion efficiency of photocathodes 208 C onverter magnified images. The collimator projects a magnified image The geometric efficiency of a converging collimator can be if the object is placed between the front of the collimator and expressed as follows: the point of convergence. An object placed further away will be imaged minified and inverted, but these collimators are rarely 2 æ d ö é d2 ù é ed for such purposes. The magnification factor is the relation- f 2 ù us g » K 2 çç ÷÷ ´ ê ú ´ ê ú ship between the image size I and object size O, as given by l (C.9) è e¢ff ø ëê (d + t )2 ûú ëê ( f - b)2 ûú I ( f + t ) = where K is a constant depending on hole shape. The maximum O f + t - b (C.7) efficiency is obtained at the point of convergence. At typical C imaging distances, 5–10 cm, the converging collimator offers a where good combination of both spatial resolution and efficiency. The f is the collimator front to convergence point distance projected image is magnified, which means that the field of view b is the collimator to source distance must be larger than the object itself and as a result, imaging with t is the collimator thickness converging collimators is best suited for large-area detectors. Related Articles: SPECT, Parallel-hole collimator, Diverging Since the magnification depends on the collimator to source collimator, Collimator, Collimator design, Collimator parameters distance, sources at different depths have a different magnifica- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. tion factor. Differences in object depth, i.e. different magnifica- Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, tions, may result in image distortions and the user must be well Philadelphia, PA, pp. 245–246. aware of this effect or else it may ultimately lead to misdiagnosis and/or ineffective/harmful treatment. Conversion efficiency of photocathodes The converging collimator contribution to the system resolu- (Nuclear Medicine) The conversion efficiency or quantum effi- tion is ciency of a photocathode is the quotient between the number of photoelectrons emitted and the number of incident photons. é d (l Photocathode efficiency is on the order of 20%–30%. The pho- e¢ff + b) ù é 1 ù é (l¢ / oll » ´ eff 2) ù Rc ê ú ´ ê1- ú tocathode thickness must be minimised to allow the excited elec- ëê le¢ff ûú ëê cosq ûú ëê f + le¢ff ûú trons to reach the surface with enough energy to penetrate the (C.8) surface potential barrier. The threshold depth at which electrons (l - 2m-1 ) l¢ originate and reach the surface and penetrate the surface le¢ eff can ff » = cosq cosq potential barrier is called the escape depth. For metals, the escape depth is a few nanometres and for some semiconductors it can extend to about 25 nm. This layer is not thick enough to attenuate where all the incoming light photons. d is the hole diameter Further Reading: Knoll, G. F. 2000. Radiation Detection le¢ff is the effective hole length, accounting for septal and Measurement, 3rd edn., John Wiley & Sons, New York, pp. penetration 266–267. θ is the angle between the central axis of the collimator and an off-centre source as seen in Figure C.107 Converted energy per unit mass (CEMA) (Radiation Protection) CEMA is an acronym for converted From this equation, it is possible to see that the spatial resolu- energy per unit mass. This dosimetric quantity C is defined for tion is best along the central axis. charged particles and can be regarded as equivalent to collision The geometric efficiency is defined as the fraction of incident kerma defined for photons. photons registered by the detector. dE C = el òf (E ) s = el E dE dm r d where dEel is the mean energy lost in electronic interactions in a mass dm of a material by the charged particles except second- ary electrons, incident on dm, ΦE(E) is the distribution of charge particle fluence with respect to energy and sel/ρ is the mass elec- tronic stopping power of a specified material for charge particle of energy E. The special name of the unit of CEMA is Gray (Gy) expressed in JKg−1. Related Articles: Restricted CEMA Object Converter Angle θ (General) In electronics, a converter is a device (e.g. rectifier, inverter) that transforms one or more parameters (e.g. frequency, voltage, number of phases) of electrical power from one value FIGURE C.107 Diagram of a converging collimator. to another. An analogue-to-digital converter (abbreviated ADC, Convex array 209 C onvolution method A/D or A to D) is a device that converts continuous signals to It is used to describe the performance of linear time invariant discrete (digital) values. The reverse operation is performed by a (LTI) systems, which are characterised by the impulse response digital-to-analogue converter (DAC). h(t). For example, if the input is x(t), then the output will be Related Articles: Analogue-to-digital converter (ADC), Digital- to-analogue converter (DAC), Frequency converter, Ramp con- y(t ) = ( x * h)(t ) verter, Scan converter, Digital scan converter, Wilkinson converter Convolution can be carried out in single or multiple dimensions. Convex array In one dimension (i.e. time), the process has the effect of filtering (Ultrasound) See Curvilinear array transducer the data in some way. The convolution of a 2D image with a 2D mask can provide many practical image processing procedures C Convex target volume such as sharpening, edge detection etc. (Radiotherapy) A convex target volume is defined as one whose Convolution Theorem: The Fourier transform of a convolu- interior angles are all less than or equal to 180°. The opposite of a tion equals the multiplication of the individual Fourier transforms: convex target volume is a concave target volume, see Figure C.75 for theoretical examples. FT ( f * g) = FT ( f )FT (g) Convex target volumes

can be treated successfully with mul- tiple beams, for example each beam may conform to the shape Discrete Convolution: The convolution of discrete functions of the tumour presented – the Beam’s eye view. See Conformal is as follows: radiotherapy for more information. However, if critical organs are close to the tumour site, then ¥ the target volume will often be defined as concave, with the aim ( f * g)(n) = å f éëmùû × g éën - mùû to limit the dose distribution within the critical organ. Typical m=-¥ tumour sites and respective critical structures that require con- cave target volumes include Convolution Identities: • Pharynx (spinal cord) Commutativity f * g = g * f • Prostate (rectum) Associativity f * (g * h) = ( f * g) * h • Skull base (optic pathway structures, brainstem) Distributivity f * (g + h) = ( f * g) + ( f * h) Identity f * δ = δ * f = f In order to achieve a concave dose distribution to match the con- cave target volume, techniques such as intensity modulated radio- Related Article: Fourier transform therapy are required. Related Articles: Target volume, Conformal radiotherapy, Convolution kernel Beam’s eye view, Intensity modulated radiotherapy (General) See Convolution method Further Reading: Bortfeld, T., R. Schmidt-Ullrich, W. de Neve and D. E. Wazar. 2005. Image-Guided IMRT: Concepts and Convolution method Clinical Applications, Birkhauser, Basel, Switzerland. (General) The convolution method is a model-based dose calcula- tion model, which directly computes the dose to a patient. The Convolution dose calculation explicitly takes into account the energy and the (Nuclear Medicine) Convolution is a mathematical operator that geometry of the beam, the presence of beam modifiers and also operates on two functions to produce a third function. The final the patient representation, which is usually obtained from image function is typically a modified version of one of the two original CT data. A model-based method computes the dose per unit flu- functions, for example in image processing where one function ence or energy fluence, allowing expression of the beam monitor represents the image and the other function a filter (e.g. smoothing units in a phantom-independent manner such as energy fluence and edge enhancing). The resulting function is then a smoothed per monitor unit. version of the initial function, giving a smoothed image with The convolution method separates the primary photon interac- increased signal to noise ratio (SNR). tions from the transport of any secondary particles, electrons and Related Articles: Signal to noise ratio, SNR, Spatial filtering, scattered photons, produced when the primary photons interact. Kernel The primary interactions are first determined by computing the energy fluence incident on the patient representation and atten- Convolution integral uating the primary intensity as the beam traverses through the (General) Convolution is the basis of many data and image pro- patient representation. cessing techniques. It is a mathematical process similar to cross- The dose distribution is obtained by convolving the convolu- correlation for real data: tion kernels with the distribution of the total energy removed per unit mass (TERMA). The TERMA in the patient representation ¥ is the product of the mass attenuation coefficient and the energy ( f * g)(t ) = ò f (t)g (t - t)dt fluence distribution. The convolution kernel represents the rela- -¥ tive energy deposited per unit volume as a function of position where the primary photon interactions occur. The TERMA dis- As can be seen from the convolution integral, the convolution of tribution accounts for the beam attenuation due to the surface two functions is defined as the integral of the product of f, with a contour and the composition of the irradiated material as well as reversed and shifted g. It is denoted with the star * symbol. inverse-square fall-off in beam intensity. The mass attenuation Coolidge tube 210 Copper coefficient depends on the beam hardening to the point of inter- Cooling curve est and the material at the interaction point. The energy fluence (Diagnostic Radiology) See Anode cooling curve is computed by ray tracing through the phantom along diverg- ing paths with respect to the target. This is adequately accu- Coordinate system rate for photons originating from the target but does not exactly (General) The coordinates in a coordinate system describe the reflect the source positions of the extrafocal radiation. The fail- location in a plane or space. For example, in two-dimensional ing is small if the dose contribution from extrafocal radiation medical imaging, two coordinates (often called x-coordinate and is reduced and the scattered radiation outside the field is well y-coordinate) give the position of an image pixel. The pixel posi- approximated. tion corresponds to a position on a detector and the pixel value C Basically, the convolution method is a blurring operation of is proportional to the amount of energy deposited in the specific the primary TERMA in the patient. The blurring is physically detector element. due to the Compton scatter and the electron transport from the A coordinate system with three physical dimensions of space site of primary photon interactions. A broad beam of an arbi- is called a ‘Cartesian coordinate system’. trary shape is divided into narrow beams or pencils incident on the patient’s surface. The dose at a point of interest is computed Coordinate transformation by summing the contributions from each pencil. The pencil (Nuclear Medicine) Coordinate transformation refers to the pro- beam method has been employed for both photon and electron cess of transforming coordinates from one frame of reference to dose calculations. The dose distributions in a series of parallel another. An example is the transformation between Cartesian planes normal to the central ray are calculated by the convo- coordinates and polar coordinates. lution of the two-dimensional profile of a pencil beam in each plane with the primary photon fluence. The convolution integral Copper can be written as (General) D ( x, y,d ) = òòF(a,b,d )K ( x - a, y - b,d )da db Symbol Cu where Element category Metal D is the computed dose at depth d and at lateral coordinates x Mass number A 63 and y relative to the central axis Atomic number Z 29 Φ represents the relative fluence distribution (TERMA) and Atomic weight 63.546 g/mol includes the effects of all secondary field shaping Electronic configuration 1s2 2s2 2p6 3s2 3p6 3d10 4s1 devices, as well as the phantom Melting point 1357.77 K K is the convolution kernel represented by the two-dimensional Boiling point 2835 K cross-sectional profile of the pencil beam at depth d Density near room temperature 8.96 g/cm3 The pencil beam dose distribution used to compute kernel K can be obtained by Monte Carlo calculations using the energy History: For thousands of years, copper has been used by man spectrum of the specific treatment machine or can be extracted in the making of jewellery, tools and decorative items. The earli- from measurements of broad-beam profiles using a deconvolution est known copper artefact is estimated to originate from nearly method. 9000 BC. Early copper metal was obtained from naturally occur- Abbreviation: TERMA = Total energy removed per unit mass. ring samples, but the process of smelting ores of copper to obtain the pure metal was in global use by around 2000 BC. Coolidge tube Isotopes of Copper: Copper occurs naturally as one of two (Diagnostic Radiology) A Coolidge tube is an x-ray tube with a stable isotopes, 63Cu and 65Cu, which exist with a relative abun- heated cathode, which replaced the gas discharge tubes used for dance of 69.15% and 30.85%, respectively. There are a further x-ray production for the first two decades after Roentgen’s discov- 17 synthetic radioactive isotopes, with atomic masses between 52 ery and investigation of the properties of x-radiation. These early and 80. 62Cu and 64Cu are used in medical diagnostic imaging tubes had several significant limitations. studies and may have potential uses in cancer therapies. The American physicist, William D. Coolidge, working in the Medical Applications: Germicide – Copper is one of a num- research laboratories of the General Electric Company, led the ber of metals that are toxic to living cells such as bacteria when development of a tube with improved characteristics. The tube in direct, prolonged contact. Copper metal is used in hospitals was first introduced in 1913 and is the type used for x-ray produc- to limit the spread of germs, particularly in door handles and air tion today. conditioning systems. A vacuum tube with a heated cathode, the Coolidge tube, pro- PET Imaging Tracer: Two radioactive isotopes of copper are vides several major advantages, including used as tracers in PET studies: • Much greater x-ray output • 64Cu decays by emission of positrons (18%), beta par- • Accuracy of adjustment ticles (39%), electron capture (43%) and a small fraction • Stability of internal conversion and gamma emission. 64Cu can • Flexibility in exposure factors be produced by particle accelerator or nuclear reactor, • Ability to duplicate exposure techniques and decays with a half-life of 12.7 h. When attached • Longer life of x-ray tube to an appropriate pharmaceutical, 64Cu can be used as • Absence of indirect radiation a tracer in PET molecular imaging, and is currently COR, Centre of rotation 211 Cornea (eye) being used to investigate perfusion, angiogenesis and ring-shaped artefacts. It is therefore necessary to apply a COR hypoxia. offset correction to each projection. This is usually done by shift- • 62Cu-ATSM is used in PET imaging, most commonly ing each projection image in the x-direction by the required num- for myocardial perfusion imaging. 62Cu decays by posi- ber of pixels before reconstruction. The required corrections are tron emission (97%) with some electron capture, with a usually determined for each collimator and detector configuration half-life of 9.7 min. A notable advantage of 62Cu is that at installation but should be checked on a regular basis and fol- it can be produced from its parent radionuclide, 62Zn, lowing major mechanical work. by generator. A COR test is usually performed by performing a circular orbit SPECT acquisition of a point source placed a known distance off- 64Cu Therapy: In animal tests, 64Cu-labelled diacetyl-bis(N4- centre from the AOR. The camera manufactures will usually sup- methylthiosemicarbazone (ATSM) has shown promise as a thera- ply software for analysing the COR offset and techniques for data C peutic agent for colorectal cancer. 64Cu-ATSM has been seen to analysis vary, but a common technique for analysis is to plot the target hypoxic cells and increases survival time of the animal x-centroid of each projection as a function of detector angle, and without global toxicity. The short path length of beta emission compare the sine curve obtained with an expected sine curve, to allows significant damage to tumour cells with little collateral determine the mean COR offset for each projection angle. damage to healthy tissue. The ability to image the distribution of Related Articles: Single-photon emission computed tomogra- the agent using PET is likely to be advantageous. phy (SPECT), Axis of rotation (AOR) Related Articles: PET (positron emission tomography), Further Readings: Bushberg, J. T., J. A. Siebert, E. M. Radionuclides in therapy Leidholdt Jr and J. M. Boone. The Essential Physics of Medical Imaging, 2nd edn., Lippincott Williams & Wilkins, Philadelphia, COR, Centre of rotation PA; Rotating Scintillation Camera SPECT Acceptance and (Nuclear Medicine) The intercept between the axis of rotation in Quality Control, AAPM Report 22, Published for the American a SPECT system, and a perpendicular line drawn from the cen- Association of Physicists in Medicine by the American Institute tre of the detectors when the detectors are parallel to the axis of of Physics. rotation. The axis of rotation (AOR) is an imaginary line about which Cornea (eye) the detectors in a SPECT system rotate. When the detectors are (Non-Ionising Radiation) The cornea is the transparent protective parallel to the AOR, a line can be drawn from the centre of the layer of the eye which covers the iris, pupil and anterior chamber detectors to the axis of rotation to define the mechanical centre of of the eye. It is the most important refractive structure in the eye rotation (COR), the central point around which the detectors rotate responsible for the light that is refracted into the eye. (Figure C.108). In a perfect

system, an image of a line source It separates the air with a refractive index of 1.00 from the placed along the AOR would yield a vertical line in the centre of aqueous humour in the eye with a refractive index of 1.33. Having each projection image, indicating that the COR is aligned with a diameter of about 11.7 mm, it is thinnest in its centre (0.5–0.6 the centre of the image. However, in practice, there is likely to be mm) and thicker at the extremities (about 0.7 mm) (Figure C.109). some misalignment between the COR and the centre of the image. Related Articles: AORD, Eye, Lens, UV light hazard The causes of misalignment can be mechanical, due to the detec- Further Readings: Coleman, A., F. Fedele, M. Khazova, P. tors not being perfectly centred in the gantry or may be due to detec- Freeman and R. Sarkany. 2010. A survey of the optical hazards tor sag or wobble. The cause of misalignment can also be electronic. associated with hospital light sources with reference to the Control Parameters such as detector sag, may vary with angle of rotation, of Artificial Optical Radiation at Work Regulations 2010. J. Radiol. meaning that the misalignment will vary across projection images. COR misalignment can cause degradation in the spatial reso- lution of reconstructed SPECT images, and if large, can cause FIGURE C.108 Centre of rotation. FIGURE C.109 Schematic of the eye. Coronal plane 212 Correlation time Prot. 30(3):469; ICNIRP. A closer look at the thresholds of thermal where Xi and Yi are measured values. Another commonly used damage: Workshop report by an ICNIRP task group. Health Phys. metric is r2, the coefficient of determination, which is equivalent 111(3):300–306; ICNIRP. 2013. Guidelines on limits of exposure to to the proportion of variance explained by linear regression; i.e. incoherent visible and infrared radiation. Health Phys. 105(1):74– if r2 = 0.85, then 85% of the variability of Y can be explained by 91; ICNIRP. 2013. Guidelines on limits of exposure to laser radia- changes in X and a linear relationship. tion of wavelengths between 180 nm and 1000 µm. Health Phys. It is important to note that the value of Pearson’s correlation 105(3):271–295; ICNIRP. 2004. Guidelines on limits of exposure coefficient only indicates the strength of a linear relationship to ultraviolet radiation of wavelengths between 180 nm and 400 between the variables; it does not give any indication of any other nm (Incoherent Optical Radiation). Health Phys. 87(2):171–186; relationship (i.e. logarithmic, polynomial). Correlations for other C ICNIRP. 2000. Revision of the guidelines on limits of exposure to relationships can sometimes be calculated by use of transforma- laser radiation of wavelengths between 400 nm and 1.4 µm. Health tions to linear models (e.g. ln(x) vs. y). Figure C.110 shows exam- Phys. 79(4):431–440; Sihota, R. and R. Tandon. 2011. Parsons' ples of non-linear relationships that are not identified with the use Diseases of the Eye, Elsevier, India; Snell, R. S. and M, A. Lemp. of Pearson’s correlation coefficient. 2013. Clinical Anatomy of the Eye, John Wiley & Sons. Pearson’s correlation coefficient also depends on the normal- ity of the sample distributions. If this assumption is not valid, Coronal plane non-parametric methods should be used such as Spearman’s Rank (General) To describe anatomical planes, imagine a person stand- Correlation Coefficient or Chi-Square. Normality can be tested ing in an upright position and dividing this person with imaginary with the use of a test such as the Kolmogorov-Smirnov test. vertical and horizontal planes. Anatomical planes can be used to Spearman’s Rank Correlation Coefficient: This correlation describe a body part or an entire body. coefficient, often denoted by ρ, is a non-parametric measure of Think of a vertical plane that runs through the centre of your correlation, determining how well the relationship between two body from side to side at right angles to the sagittal plane. This random variables can be described by a monotonic function. It is plane divides the body into front (anterior) and back (posterior) equivalent to the aforementioned Pearson’s correlation coefficient regions. This plane runs through the central part of the coronal calculated from the ranks of the data, rather than the raw data suture or through a line parallel to it; such a plane is known as a itself. coronal plane. Note that the correlation reflects the noisiness and direction Related Article: Anatomical body planes of a linear relationship (top row), but not the slope of that rela- tionship (middle), nor many aspects of non-linear relationships (bottom). N.B.: the Figure C.110 in the centre has a slope of 0 but Correction factor in that case the correlation coefficient is undefined because the (Radiotherapy) Correction factors are factors multiplied with a variance of Y is zero. measured value in order to increase the accuracy of that measure- Related Article: Normal distribution ment. An example is the correction factors used in quality control testing of linear accelerators used for external radiotherapy. Correlation time Related Article: Precision (Magnetic Resonance) The term correlation time τc is used to describe the time over which a particular pattern in an experi- Correlation ment is correlated under the influence of a stochastic disturbance. (General) This term can either mean the statistical correlation Thus, it is the characteristic time over which fluctuations of a coefficient between two variables, or the cross-correlation of two parameter occur. In MRI, the ‘correlation time’ concept is pri- functions. For information about the latter, please see articles on marily used in connection with relaxation. In relaxation theory, Cross-correlation and Autocorrelation. the correlation time is the characteristic time between fluctua- The correlation between two random variables indicates the tions in the local magnetic field (the lattice field) experienced by strength and direction of a linear relationship. There are different a spin due to the movements of neighbouring nuclei or molecules. types of coefficients that can be used depending on the situation. The correlation time is often described as being the average time Pearson’s Correlation Coefficient, r: This is the most com- between molecular collisions, and is thus related to the binding mon of correlation coefficients, and ranges from +1 to −1. A state of the water molecules in the system. For example, free correlation of +1 indicates a perfect positive linear relationship water molecules display rapid rotational movements, frequent between variables, −1 indicates a perfect negative linear correla- collisions and short τc. tion. 0 indicates no linear relationship. It is calculated by dividing the covariance between the two variables by the product of the standard deviations (std). Pearson’s correlation coefficient is calculated as follows: 1.0 0.8 0.4 0.0 –0.4 –0.8 –1.0 n 1 åæ X - X öæ Y -Y ö r = i i n -1 ç s ÷ç 1.0 1.0 1.0 0.0 –1.0 –1.0 –1.0 s ÷ i=1 è X øè Y ø X :mean valueof X 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Y :mean valueof Y sX :stdof X s FIGURE C.110 Several sets of (x, y) points, with the correlation coef- Y :stdof Y ficient of x and y for each set. Cost/benefit analysis 213 CPMG (Carr–Purcell–Meibom–Gill) Related Articles: Relaxation, Relaxation rate, Relaxation Related Articles: Apparent source position, Scintillation time, B0, Temperature-dependence, Rotational averaging camera Further Reading: Bloembergen, N., E. M. Purcell and R. V. Pound. 1948. Relaxation effects in nuclear magnetic resonance Counting limitations absorption. Phys. Rev. 73:679–712. (Nuclear Medicine) Counting limitation describes a system’s capability to register and process very high event rates in the Cost/benefit analysis detector or detector system. Very frequent events will saturate the (Radiation Protection) Cost/benefit analysis must be used by detection system and count losses arise due to dead time of the an employer to justify any practice that involves the exposure detector circuitry. The latter can be due to the type of the counting of staff, patients or the public to ionising radiation. The benefits system: paralyzable; non-paralyzable or a combination of both. of the practice must outweigh the costs in terms of detriment or C A scintillation camera can have very severe counting limita- adverse radiation effects to those exposed. tions if the pile-up rejection is not working and mispositioning of For more information, see Justification. count events can occur at very high count rates. It also has to be Related Articles: Adverse radiation effects, Justification remembered that it is the total number of photons impinging on Further Reading: ICRP (International Commission on the crystal across the whole energy spectrum that is of importance. Radiological Protection). 2008. Recommendations of the interna- Related Articles: Count losses, Pile up, Paralyzable counting tional commission on radiological protection, ICRP Publication system, Non-paralyzable counting system 103, Ann. ICRP 37/2-4. Further Readings: Strand, S.-E. and I.-L. Lamm. 1980. Theoretical studies of image artifacts and counting losses for Couch/patient different photon fluence rates and pulse-height distributions (Nuclear Medicine) The patient couch refers to the imaging table in single-crystal NaI(TI) scintillation cameras. J. Nucl. Med. in an image modality system. 21:264–275; Strand, S.-E. and I. Larsson. 1978. Image artifacts In transmission and emission imaging, the couch is typically at high photon fluence rates in single NaI(Tl) crystal scintillation made from a sturdy but low attenuating material so as not to pro- cameras. J. Nucl. Med. 19:407–413. duce any noticeable effect on the resulting images. In MRI, the couch should be made out of non-magnetic mate- Counting systems rial to prevent any image distortion. (Nuclear Medicine) A counting system is a detector that measures Abbreviation: MRI = Magnetic resonance imaging. the activity in radioactive samples. The most common counting system is the well counter. Coulomb Related Article: Well counter (General) The coulomb (named after Charles-Augustin de Coulomb, a French Physicist) is defined as the amount of charge that flows through a given cross section of conductive wire in 1 s Counting times if there is a steady current of 1 A in the wire: (Nuclear Medicine) See Acquisition time q = It Coupling efficiency (Ultrasound) Coupling efficiency describes the efficiency with which the ultrasound wave from the active transducer element is where transmitted into the tissue medium. The design of the transducer q is in coulombs incorporates matching layers of optimal thickness and acoustic I is in amperes impedance to transmit as much energy across the bandwidth as t is in seconds possible. High coupling efficiency is also important for the return- Further Reading: Resnick, R. and D. Halliday. Physics, Wiley ing echoes. International Edition, New York. Related Articles: Backing material, Acoustic impedance, Related Articles: Charge, Force electrostatic Transducer Coulombs/kg Coupling medium (Radiation Protection) Coulombs per kilogram is the SI unit (Ultrasound) Gel is most commonly used as coupling medium in of exposure, given the name ‘Roentgen’. More precisely, it is diagnostic examinations, Figure C.111. The purpose is to avoid quantity of electric charge of one sign (+ve or −ve, measured in air to be trapped between the transducer and the skin. As air has Coulombs) that is liberated by the interaction of ionising radiation much lower acoustic impedance than soft tissue, 400 Rayls com- and collected by the electrodes in an ionisation chamber contain- pared to 1.6 × 106 Rayls, total reflection will occur at the boundary ing a mass of air (in kilograms). 1 R = 1 C/kg. with trapped air, which will reduce the image quality markedly. Related Articles: Exposure, Roentgen (R) Oil or water can also be used as coupling medium. Related Articles: Coupling efficiency, Reflection coefficient, Count rate Rayl (Nuclear Medicine) The counting or count rate is the registered number of photons per unit of time. A high count rate is ben- CP (circularly polarised) eficial in many applications, for example imaging applications (Magnetic Resonance) See Circularly polarised (CP) where high count rates enable fast examinations with adequate statistics. In some applications, high count rates can have nega- CPMG (Carr–Purcell–Meibom–Gill) tive effects, for example in a scintillation camera (see Apparent (Magnetic Resonance) See Carr–Purcell–Meibom–Gill (CPMG) source position). sequence CR (computed radiography) 214 C reatine greyscale values. Usually the CR reader transfers the PSL into 4096 greyscale values (12 bits colour – 212 = 4096), thus forming the contrast resolution of the CR reader. See the articles on Storage phosphor and X-ray film scanner. Related Articles: Storage phosphor, Computed radiography, Matrix size, X-ray film scanner Creatine (Magnetic Resonance) Creatine (Cr) is a chemical compound C that features in in vivo

proton (1H) NMR spectra of a number of organs (Figure C.113). Protons in the methyl group in creatine and phosphocreatine give rise to a resonance at 3.02 ppm (thus, these compounds can- not be distinguished in 1H MRS). There is a further resonance at 3.9 ppm from CH2 protons, but as this is smaller and obscured by overlapping peaks, it is normally considered less important. Creatine plays a role in the body’s energy metabolism. Buffering mechanisms ensure that its concentration in the normal brain is usually stable, and therefore it is often used as an internal reference for quantitative spectroscopy. However, it is reduced in certain types of tumours and slightly increased in reactive gliosis. The importance of phosphocreatine in 31P NMR arises from its role in the body’s energy metabolism. PCr provides a readily FIGURE C.111 Gel is used as coupling medium between the transducer available source of phosphorus to support generation of ATP from and the human skin. ADP, and is depleted to maintain ATP levels during ischaemia and hypoxia. Thus, PCr levels can be used to monitor the effect of fatiguing exercise in skeletal muscle. It also provides a prognostic indicator in birth asphyxia that is well correlated with neurodevel- opmental outcome (Figure C.114). Related Article: Phosphocreatine CH3 O | || NH = C – N – CH2 – C - OH | NH2 FIGURE C.113 Molecular structure of creatine. 0.8 0.6 – CH3 of Cr, 3.02 ppm FIGURE C.112 Introducing a storage phosphor cassette in the CR reader (the door is open to show the scanner). 0.4 CR (computed radiography) (Diagnostic Radiology) See Computed radiography (CR) 0.2 CR reader (Diagnostic Radiology) The CR reader is the component of the 0.0 computed radiography (CR) system that scans the storage phos- phor plate carrying the latent image after the x-ray exposure with 4 3 2 1 a laser beam (Figure C.112). As a result, light, the photostimu- ppm lated luminescence (PSL), is released from the storage phosphor plate. The intensity of the PSL light is recorded as the pixel value FIGURE C.114 1H NMR spectrum of the human brain showing Cr reso- that is later translated into a visual pixel with corresponding nance at 3.02 ppm. Critical structures 215 Cross-hairs Critical structures dN (Radiotherapy) The goal of radiotherapy is to deliver as high a F = da dose as possible to the tumour while sparing surrounding normal tissue. Certain organs within the body are very sensitive to radia- Unit: m−2 tion; hence, the dose delivered to the tumour is limited. These The barn, b, is a special SI unit of cross section: organs are called critical structures or organs at risk. They must be taken into consideration during the treatment planning pro- 1barn = 10-28m2 = 100fm2 cess. Some examples of critical structures are the lungs, the spi- nal cord, the eyes, the rectum and the bladder. Absorbed dose to the critical structures and the percentage of irradiated volume is Related Article: Barn limited. These limits are called constraints. DVHs or NTCPs can Further Reading: ICRU (International Commission on C be used for plan evaluation from the point of view of dose to the Radiation Units and Measurements). 1998. Fundamental quan- critical structures. tities and units for ionizing radiation, ICRU Report 60. ICRU, Therefore, these critical structures near to the tumour limit Bethesda, MD. the amount of radiation that can be administered due to their dose tolerances. Cross-correlation Abbreviations: DVH = Dose volume histogram and NTCP = (Ultrasound) In signal processing, the cross-correlation is a mea- Normal tissue complication probability. sure of how similar two signals are. For instance, it can be used Related Articles: Organ at risk, Conformal radiotherapy, to find features in an unknown signal by comparing it to a signal Custom blocking, Normal tissue complication probability, with desired shape. The cross-correlation is similar to convolution Normal tissue dose, Normal tissue dose response, Normal tissue of two signals, but does not involve a time-reversal, only a shift reaction, Normal tissue toxicity, Tolerance and multiply. The autocorrelation is a cross-correlation, but with the signal itself. Consequently, the autocorrelation will always Crookes tube have a peak at zero lag. (Diagnostic Radiology) The Crookes tube is a partially evacu- Cross-correlation in diagnostic ultrasound is used as an alter- ated vacuum discharge tube with electrodes in a glass envelope. native to the autocorrelation algorithm to estimate blood flow The tube had been developed by Sir William Crookes, an English velocity. The resulting estimate is a time-shift, which is a mea- chemist and physicist, to study several electrical phenomena. sure of how much a scattering target has moved between two Such type of tube has been used by W. C. Roentgen when he dis- transmit pulses. Cross-correlation has a number of advantages, covered the x-rays. foremost the ability to detect higher velocities without aliasing. The same pulse can also be used to form the B-mode image as for Cross section flow velocity estimation. The drawback is however an uncertainty (Radiation Protection) The concept of cross section, symbol, Φ, in the estimation. is used to describe the probability or likelihood of the interac- Each RF line is divided into segments, where each segment tion of charged or uncharged particles, or photons, with nuclei or is correlated to the same segment of the subsequent RF line. The other ‘target entity’. If a beam is interacting with a target, then the estimate can be improved by averaging correlations from sev- cross section is the average area perpendicular to the direction of eral RF lines. Usually, the smallest detectable velocity (corre- the radiation, which has to be assigned to each nucleus in order sponding to one sample point) is too coarse; so one method to to account geometrically for the total number of interactions that improve the resolution is to fit a second-order polynomial to the occur across an incident beam of given fluence. The unit of cross three points at the peak, and then find the peak position from the section is therefore that of area (m2). However, although it can be polynomial. thought of as the ‘area of influence or interaction’, or as the ‘tar- If a full calculation of the cross-correlation is performed, get area’ afforded by each nucleus/atom, it must be considered as this will amount to a very large number of calculations. To completely independent of the actual physical dimensions of the reduce the computational load, the calculation may only need nucleus/atom. to be performed by converting the RF data to one-bit signals, The cross section depends not only on the type of the target i.e. 1 for the positive part of the signal, and −1 for the nega- but on the type and energy of the particles. Cross section is not tive. For an infinite number of samples, this method is exact in limited to beams – it can be expressed in more general terms and finding the time shift, but as the data are limited, there will be the term is defined by the International Commission on Radiation a deviation from the ‘true’ time shift, however normally very Units and Measurements (1998) as follows: small. The cross section, σ, of a target entity, for particular interac- Related Article: Autocorrelation tion produced by incident charged or uncharged particles, is the quotient of P by Φ, where P is the probability of that interaction Cross-hairs for a single target entity when subjected to the particle fluence, (Radiotherapy) Cross-hairs are lines and marks projected onto the Φ, thus image obtained on a simulator in order to visualise the treatment field with respect to the patient’s anatomy and to ensure the cor- P rect treatment geometry. The treatment field can be centred using s = F the central cross of the cross-hairs (see Figure C.115). In addi- tion, cross-hairs can be used to show the extent of the treatment Unit: m2 field and usually have a scale or graticule graduated in centimetre Fluence. The fluence, Φ, is the quotient of dN by da, where intervals in order to allow measurements to be made with respect dN is the number of particles incident on a sphere on a sphere of to the centre or field edges, etc. cross-sectional area da, thus Related Article: Simulation Cross-line curves 216 Cryogen Crusher gradient 11 cm (Magnetic Resonance) A crusher gradient is a gradient that de- phases the signal to eliminate unwanted transverse magnetisation components in order to preserve the desired signal properties. The technique is used with pulse sequences having at least one Cross hairs 8 cm refocusing RF pulse. The coherence of the transverse magnetisation is affected by the crusher gradients altering the phase of the signal. In a spin echo, two lobes having the same polarity are applied C 1 cm before and after the RF pulse ΦL and ΦR. The lobes can have the same or different areas. FIGURE C.115 Cross-hairs with centimetre scale or graticule and superimposed field of 11 × 8 cm. Cross-line curves (Radiotherapy, Brachytherapy) The Paris Dosimetry System: For the Paris System dose cal- culations, dose rates in water at the central plane of a linear source are needed, where oblique filtration and attenuation and scatter- ing in water are taken into account. Besides tabled values and Escargot curves, so-called cross- line curves/graphs are also used. Cross-line curves give the dose rate for 192Ir-wires of unit linear source strength as a function of distance from the centre of the wire, in the central plan and also in planes parallel to the central plane. One graph is used for each 192Ir wire length. Related Articles: Paris system, Escargot curves Illustration of crusher gradient. Further Reading: Venselaar, J. and J. Pérez-Calatayud. eds. 2004. A Practical Guide to Quality Control of Brachytherapy If the two lobes are equal for a spin echo sequence the net Equipment, ESTRO Booklet No 8, European Society for magnetisation will be zero. Therapeutic Radiology and Oncology, Brussels, Belgium. ΦR, if large enough, can completely cancel the FID from the data acquisition by eliminating the signal coherence. Related Articles: Spoiler gradients Crossed grid Further Readings: Goelman, G., G. Pelled, S. Dodd and A. (Diagnostic Radiology) See Grids, crossed Koretsky. Tracking the effects of crusher gradients on gradient- echo BOLD signal in space and time during rat sensory stimula- Crosstalk tion; Wu, G., S. Lee, X. Zhao and Z. Li. Crusher gradient reversal (Magnetic Resonance) Crosstalk is an artefact caused by the to eliminate stimulated echo artifacts in dual spin echo diffusion imperfect shape of RF pulse slice profiles. An ideal RF pulse slice MRI. profile should be square, but in reality they are typically more bell-shaped. In multi-slice mode with contiguous slices, an RF Cryogen pulse can therefore partially excite adjacent slices. Crosstalk (Magnetic Resonance) Cryogen is the cooling agent that is used can reduce the signal intensity and/or modify the image contrast to maintain a superconducting magnet at a sufficiently low tem- through partial saturation. perature. Normally, liquid helium (4.2 K) is used. The cryogen The artefact occurs when slices are too close together and can will tend to boil off and needs to be refilled periodically. Many be eliminated by ensuring the spacing between slices is typically systems now incorporate cryo-compressors to liquify gaseous at least 10% of the slice thickness. 3D imaging can be used if it is helium. Since helium is expensive, extensive boil off is prevented important to view the whole volume. by the cryostat (see related article), minimisation of heat loss Abbreviation: RF = Radio frequency. pathways and for older MR systems by the use of liquid nitro- Related Articles: Slice profile, Slice spacing gen (77 K) as a second cryogen, providing a thermal shield to the Further Reading: Simmons, A., P. S. Tofts, G. J. Barker and more expensive helium. Increasingly, cryorefrigerators are being S. R. Arridge. 1994. Sources of intensity non-uniformity in spin used in smaller superconducting magnets, reducing the need for echo images at 1.5 T. Magn. Reson. Med. 32(1):121–128. cryogen replenishment (Figure C.116). The cryogen (helium) surrounds the superconductive magnet CRT (cathode ray tube) coil. (Diagnostic Radiology) See Cathode ray tube Related Article: Cryostat Cryostat 217 C T fluoroscopy Thermal insulation Magnet bore Cryogen (He) RF body coil Gradient coils Superconductive magnet coil C FIGURE C.116 Schematic cross-sectional drawing of an MRI system. c-plane Cryostat (Radiation

Protection) A cryostat is a device for maintaining a B-mode constant very low temperature, i.e. below 123 K. The use of a cryo- Elevation plane plane stat is indispensable for many semiconductor radiation detectors (germanium or silicon) that are required to operate at the tempera- 2-D scan moved in elevation plane Volume scan ture of liquid nitrogen (77 K), or in magnetic resonance imaging for maintaining the static field superconducting magnet (∼2 K). There are many kinds of cryostats; the simplest cryostat is FIGURE C.117 The light shaded area shows a C-scan acquired either by moving a conventional transducer in the elevation plane or using a a Dewar vessel in which the thermal insulation minimises heat volume acquisition. flows by conduction, convection or radiation from and to the envi- ronment. There are two types of these vessels made from glass or stainless steel double walls, containing a vacuum or a combina- CT (computed tomography) tion of vacuum and super-insulation inside. The Dewar contain- (Diagnostic Radiology) See Computed tomography (CT) ing liquid helium (boiling temperature 4.2 K) is surrounded by a second one containing liquid nitrogen (77 K). Dewars can be CT detectors made in many shapes and sizes from simple cylindrical vessels (Diagnostic Radiology) Often CT scanners use scintillator detec- to large tanks. Related Article: tors, such as cadmium tungstate, to convert x-rays into light. The Semiconductor detector Further Readings: light is then converted into electrical signals using photo diodes. Ekin, J. W. 2006. Reprinted 2007 (with If the efficiency of x-ray detection can be increased and electronic correction). Experimental Techniques for Low-Temperature noise levels reduced in an x-ray detection system patient dose Measurements. Cryostat Design, Material Properties, and reduction is possible. Superconductor Critical-Current Testing, Oxford University One manufacturer uses a garnet gemstone material (Gemstone Press, New York, pp. 18–19; Serway, R. A., R. J. Beichner and Detector; GE Healthcare, Milwaukee, WI) as a scintillator. The J. W. Jewett. 2000. Physics for Scientists and Engineers with advantage of the material includes high x-ray detection efficiency Modern Physics, 5th edn., Saunders College Publishing, Fort and significantly reduces radiation doses. Worth, TX, p. 629; Thornton, S. T. and A. Rex. 2000. Modern Another manufacturer uses praseodymium-activated scin- Physics for Scientists and Engineers, 2nd edn., Saunders Golden tillator in their detector (PUREViSION Detector; Canon Medical Sunburst Series, Fort Worth, TX, pp. 346–347. Systems, Tochigi, Japan). Similar to the Gemstone Detector, its advantage is high x-ray detection efficiency. Cryotherapy A third manufacturer uses integrated CT detectors (Stellar (Radiotherapy) Cryotherapy is a pain treatment that uses and StellarINFINITY Detectors; Siemens Healthineers, Erlangen, localised freezing temperatures to deaden an irritated nerve. Germany) that directly couple the photodiode with the analogue- Cryotherapy is also used as a method of treating localised areas to-digital convertor to reduce the electronic noise levels and arte- of some cancers (also called cryosurgery), such as prostate can- facts, thereby reducing radiation doses. cer and skin cancer. C-scan CT dose index (Ultrasound) The term C-scan describes a scan to produce an (Diagnostic Radiology) See CTDI image in a plane parallel to the transducer face. For a transducer placed on the skin, the resulting image is parallel to the skin sur- CT fluoroscopy face. Compared to a conventional B-mode image, C-scans must (Diagnostic Radiology) CT fluoroscopy (CTF) is a mode of obtain echoes from an axis orthogonal to the B-mode plane. This imaging, which offers a continuously updated image. In CTF, can be done by moving a transducer in the elevation plane or using the couch is not incremented through the gantry in the manner a 3D volume acquisition (e.g. from a 2D array), as illustrated in of sequential or helical scanning, but the same volume of the Figure C.117. patient is viewed over a period of time in a manner analogous to fluoroscopy in conventional radiology. The main use of CTF CSI (chemical shift imaging) is in interventional procedures such as tissue biopsies and fluid (Magnetic Resonance) See Chemical shift imaging (CSI) draining (Figure C.118). CT number 218 CT optimisation represented by the particular pixel. Modern diagnostic scan- ners use Hounsfield units (HU) to express CT numbers. On the Hounsfield scale, attenuation coefficients are normalised to that of water, and the CT number of a material, m, is given by the following formula: 1 2 3 m - m CT number (m) = m water .1000HU (C.10) mwater C where μm and μwater are the linear attenuation coefficients of a material, m and water, respectively. From Equation C.10, the CT number of water is 0 HU, and, as the attenuation of air can be approximated to zero, that of air is −1000 HU. CT images are generally displayed on a greyscale, where more 4 5 6 attenuating materials are represented by lighter shades and the less attenuating by darker ones. The standard Hounsfield scale ranges from approximately −1000 to +3000 HU, so contains 212 FIGURE C.118 Sequence from CTF biopsy showing biopsy needle and lesion. (Reproduced by permission of K. Katada, Fujita Health University, levels of grey (Figure C.119a). The human eye however is only Japan.) capable of distinguishing less than 1000 levels of grey on the monitor. The range of CT numbers displayed on the monitor can be The main requirements for CTF are a continuously rotating varied by adjusting the window level (WL) and window width gantry (as enabled by slip-ring technology), and fast collection (WW), according to the tissues of interest. Better differentia- and reconstruction of data; so there is only a short delay between tion between tissues can be perceived if the WW is reduced, so data collection and image display. The first image in the series is that a smaller range of HU is displayed over the entire greyscale reconstructed from the first rotation. The second image is recon- (Figure C.119b and d). The WL can then be set to display the tis- structed by subtracting the first segment of data, for example the sues of interest (Figure C.120). first 60°, and adding data from first 60° of the second rotation. Typical CT numbers of tissues, on the Hounsfield scale, are Subsequent images are reconstructed by always subtracting the oldest 60° of data and replacing it with the most recently acquired 60°. In this way, six images per gantry rotation (360/60) can be Lung Approx. −550 to −950 HU displayed for each simultaneously acquired z-axis position. In order to provide faster display of images, reconstruction is White matter Approx. 30 HU usually performed initially on a 256 × 256 matrix, rather than the Grey matter Approx. 40 HU conventional 512 × 512. Muscle Approx. 50 HU When used for interventional procedures, additional hardware Fat Approx. −90 HU is usually required so that the CT scanner can be operated from Trabecular bone Approx. 300–500 HU inside the scanner room. This includes a scan room monitor and a Cortical bone Approx. 600–3000 HU console with a foot switch for controlling exposure and a joystick, or similar, for couch-top movement. Radiation dose to the patient should be closely monitored in Related Articles: Window, Hounsfield number, Hounsfield CTF. Although, generally, lower tube currents are employed than scale in conventional CT scanning, the same area of the patient is irra- diated continuously and this can result in high local organ doses. CT optimisation When operating in CTF mode, an alarm is employed to alert the (Diagnostic Radiology) CT is widely used as an essential diag- operator after a preset time limit, for example if 100 s have been nostic imaging tool in clinics. exceeded. For justifying each CT examination, weighing the benefits When performing procedures from within the scan room, against the risks is important. appropriate precautions should be taken by the operator and doses In addition to justifying each CT examination, radiation doses to the hands, eyes and body carefully monitored to ensure limits used should be as low as reasonably achievable (ALARA). To are not exceeded. Operators should wear protective aprons, when implement ALARA principles in CT image acquisitions, radiolo- practical use needle holders to keep the hands out of the direct gists, physicists and technologists should make efforts to produce beam, and stand as far away as possible from the scan plane dur- optimal images with the lowest dose to patients; this process is ing scanning. called optimisation. Further Readings: Keat, N. 2000. ImPACT, Real time CT There are various methods and techniques for optimising radi- and CT fluoroscopy, Evaluation report MDA/00/10, HMSO, ation doses in CT examinations shown as follows: Norwich, UK; Keat, N. 2001. Real time CT and CT fluoroscopy. Br. J. Radiol. 74:1088–1090. • Tube current modulation • Longitudinal (z-axis) CT number • Angular (xy-axis) (Diagnostic Radiology) The CT number is the numerical • Longitudinal and angular (xyz-axis) value of each pixel in a CT image matrix, and is proportional • Organ-based to the linear attenuation coefficient of the material in a voxel • Electrocardiogram (ECG)-gated CT optimisation – angular (xy-axis) 219 CT optimisation – electrocardiogram (ECG)-gated –1000 0 1000 +3000 WL = 1000 WW = 4095 (a) –1000 0 +1000 WL = 0 WW = 2000 C (b) Air Water Bone –200 0 +200 WL = 0 WW = 400 (c) Fat Water Muscle 800 1000 1200 WL = 1000 WW = 400 (d) Bone FIGURE C.119 The CT number scale in Hounsfield units: (a) the full range displayed over the greyscale; (b) WL changed and WW reduced; (c) WW reduced further to display a smaller range of CT numbers over the greyscale; and (d) WL changed to display different range of CT number values. 3000 + HU 3000 + HU 0 HU WL WW WL WW –1000 HU WL –600, WW 500 WL –10, WW 400 –1000 HU (a) (b) FIGURE C.120 Window level and WW set to display (a) lung tissue and (b) pulmonary metastases. • Tube voltage adjustment tube current is adjusted according to the attenuation data from the • Optimising reconstruction kernel localiser radiograph or in near-real time according to the measured • Optimising image slice thickness attenuation from the previous 180° projection (Figure C.121). • Optimising bowtie filter • Optimising number of acquisition phases CT optimisation – electrocardiogram (ECG)-gated • Application of selective organ shielding (Diagnostic Radiology) In one method a standard tube current is • Application of dual-energy CT (virtual non-contrast applied over a limited range of heart phases to ensure low noise [VNC] image) levels while minimising the tube current during the remaining • Application of newly developed x-ray detection system heart phases to reduce the patient dose. Another method of electrocardiogram (ECG)-gated modula- Related Articles: ALARA tion is to employ prospective gating axial acquisitions in which x-rays are turned on only during a limited heart phase and are CT optimisation – angular (xy-axis) completely turned off during other heart phases. However, this (Diagnostic Radiology) X-ray is attenuated more in the lateral method can only be used for patients with low and stable heart direction than in the anteroposterior direction; thus, it is effective rates. Different types of ECG-gated modulation are shown in that the tube current is modulated within one gantry rotation. The Figure C.122. CT optimisation – longitudinal modulation (z-axis) 220 C T reconstruction CT optimisation – longitudinal modulation (z-axis) CT optimisation – organ-based (Diagnostic Radiology) In longitudinal modulation, the tube cur- (Diagnostic Radiology) Organ-based modulation is used to rent is adjusted according to the size and anatomical regions of a reduce the dose in radiosensitive organs such as the breast, thy- patient. This modulation is performed to produce relatively uni- roid, and eye lens. In this technique, the tube current is decreased form noise levels through the entire acquisition range. The tube over radiosensitive organs. current is modulated to provide the desired image quality at the The organ-based modulation techniques can be divided into chosen attenuation on the basis of prior calculations from a local- two types: one is the technique in which tube current is increased iser radiograph. over the anterior surface and is decreased over the remaining sur- Examples of longitudinal dose modulation for a thoracic face to maintain the total tube current over 360 degrees. The other C phantom for different manufacturers are shown in Figure C.123. is the technique that is designed to decrease the tube current over Although the tube current is modulated across various anatomical the anterior surface

without increasing the tube current over the regions, the characteristics of dose modulation vary among man- remaining surface (Figure C.124). ufacturers. The longitudinal modulation that uses data obtained from localiser radiographs cannot appropriately adjust the tube CT reconstruction current if the patient is not positioned at the isocenter of the CT (Diagnostic Radiology) The history of tomographic image recon- gantry. struction goes back to the early twentieth century. Austrian math- ematician Johann Radon (1887–1956) developed the mathematical CT optimisation – longitudinal and angular (xyz-axis) basis of reconstructing tomographic images from sinograms (see (Diagnostic Radiology) The simultaneous combination of lon- images given in the following, in the article on Back-projection gitudinal and angular modulations involves the variation of and also references) containing multiple attenuation profiles col- tube current along both the longitudinal (z-axis) and in-plane lected around the object. Wider use of this method in medical (xy-axis) directions of a patient. This modulation is the most use became applicable with the development of computers and comprehensive approach for reducing CT dose because the dose computed tomography (CT). is adjusted according to patient-specific attenuations in all three Profiles are generated by detecting the thinly collimated planes. x-rays transmitted through the object (patient). The sinogram con- Related Articles: CT optimisation – angular (xy-axis), CT sists of views (number of shots done during one rotation) and rays optimisation – longitudinal (z-axis) (number of elements in one shot). The number of views (∼1000 at modern scanners) affects the circumferential component of resolution, and the number of rays (∼800/view), the radial compo- nent of resolution in the CT image. The task is to compute from the preprocessed sinogram data the linear attenuation coefficient (Equation C.11) for each pixel in the image (2D) or each voxel of the object (3D): æ I ö µt = ln 0 ç (C.11) è I ÷ t ø These values are normalised and scaled to present CT numbers in Hounsfield units (HU), (Equation C.12). The CT numbers are further to be presented as a greyscale image: (m = tissue- m water ) CT(HU) 1000* IGURE C.121 Angular tube current modulation. (m (C.12) F water- m air ) FIGURE C.122 Different types of ECG-gated modulation: (a) retrospective gating and (b) prospective gating. CT scanner 221 C T simulator C FIGURE C.123 Comparison of the tube current when applying longitudinal modulation for different manufacturers. _R eco nstru ction .html ; Epfl .c h: http: / /big www .e pfl .c h /dem o /jto mogra phy /d emo .h tml CT scanner (Diagnostic Radiology) See Computed tomography CT simulator (Radiotherapy) The major steps in the target localisation and treatment field design are as follows: • Acquisition of the patient data set • Localisation of target and adjacent structures • Definition and marking of the patient coordinate system • Design of treatment fields • Transfer of data to the treatment planning system (TPS) FIGURE C.124 An example of organ-based modulation. This type of • Production of an image for treatment verification organ-based modulation decreases tube current to the anterior surface and increases tube current to the other surface. CT simulators are CT scanners equipped with special features that make them useful for certain stages in the radiotherapeutic process. The special features typically are as follows: The methods used for CT reconstruction include back-projection method, iterative reconstruction and analytical methods. Only • A flat tabletop surface to provide a patient position dur- the latter category method, i.e. filtered (or convolution) back- ing simulation that will be identical to the position dur- projection method is nowadays applied. A CT image reconstruc- ing treatment on a megavoltage machine. tion demo with a few choices of different parameters is shown in • A laser marking system to transfer the coordinates of Figure C.125. the tumour isocentre, derived from the contouring of Related Articles: CT, Back-projection, Iterative reconstruc- the CT data set, to the surface of the patient. Two types tion, Filtered back-projection of laser marking systems are used: a gantry-mounted Further Readings: Bushberg, J. et al. 2002. The Essential laser and a system consisting of a wall-mounted move- Physics of Medical Imaging, 2nd edn., Lippincott Williams & able sagittal laser and two stationary lateral lasers. Wilkins, Philadelphia, PA; Brendan F. Hayden, 2005, http: / / • A virtual simulator consisting of software packages hom epage s .inf .ed .a c .uk/ rbf /C Vonli ne /LO CAL _C OPIES /AV04 that allow the user to define and calculate a treatment 05 /HA YDEN/ Slice _Reco nstru ction .html (accessed 31/07/2012); isocentre and then simulate a treatment using digitally Michael Liebling, 2001, http: / /big www .e pfl .c h /dem o /jto mogra reconstructed radiographs (DRRs). phy /d emo .h tml (accessed 31/07/2012) Hyperlinks: Slice reconstruction: http: / /hom epage s .inf .ed .a In CT simulation, the patient data set is collected and target c .uk/ rbf /C Vonli ne /LO CAL _C OPIES /AV04 05 /HA YDEN/ Slice localisation is carried out using CT images with fluoroscopy CT x-ray tube 222 CT x-ray tube C FIGURE C.125 CT reconstruction demo. (Courtesy of Michael Liebling, University of California.) and radiography replaced by DRRs. A laser alignment system CT x-ray tube is used for marking and a virtual simulator software package (Diagnostic Radiology) Basic operation of the CT x-ray tube is is used for field design and production of verification images. similar to that of conventional radiology. Typically, two (‘small’ Transfer of all necessary information to the TPS is achieved and ‘large’) focal spots are available, which are automatically electronically. selected according to the protocol. Total filtration is typically A CT simulator essentially obviates the need for conventional 6–10 mm Al equivalent. The heat capacity and the cooling rate simulation by carrying out two distinct functions: of the tube have been designed for long acquisition times and heavy current loads, typical values are 5–8 MHU and 1500–5000 • Physical simulation, which covers the first three of the kHU/min, accordingly. The tube assembly in the gantry has been six target localisation steps listed earlier designed and has to be balanced carefully for each installation to • Virtual simulation, which covers the last three of the six allow the fast rotation times (<1 s) used in diagnostic imaging. target localisation steps listed earlier Although the new CT x-ray tubes allow for dramatic increase of the power and intensity of the exposure, the focal spot is still Related Article: DRR determined by the size of the cathode filament. A new x-ray tube Further Reading: Podgorsak, E. B. 2003. Review of Radiation (commercial name STRATON) has been developed by Siemens, Oncology Physics: A Handbook for Teachers and Students, which solves this problem (Figure C.126). This tube has a com- International Atomic Energy Agency, Vienna, Austria. pletely different construction. It uses thermal electrons created by CTDIw 223 CTDI Anode The Brilliance iCT and DoseWise strategies, Phili psHea lthca Vacuum re_45 22962 42611 _LR1[ 1].pd f (accessed on 31 July 2012). Cathode Anode Heat CTDI (Diagnostic Radiology) Absorbed dose in CT is usually expressed in terms of the CTDI, computed tomography dose index. The unit Cooling oil of CTDI is gray, but doses are generally at the level of units to tens Cathode Cooling oil of milliGray (mGy). The CTDI represents the absorbed dose from Conventional tube technology Stration x-ray tube a series of scanner rotations, although the dose is traditionally measured with a single axial scan, from which the series dose is C calculated using the following formula: FIGURE C.126 Diagram of x-ray tube STRATON compared with con- ventional x-ray tube – note the system of electrodes to deflect plus focus +¥ the beam of thermal electrons. (Image courtesy of Siemens Healthcare.) 1 CTDI = D (z)dz nT ò -¥ where D(z) is the dose profile along the z-axis (scanner axis of rotation) T is the nominal width of single acquired slice n is the number of slices acquired per rotation therefore nT is the nominal collimated z-axis x-ray beam width The preceding equation represents the general form of CTDI. A number of specific CTDI definitions have been formulated over the years, and currently the most commonly used is the CTDI100. CTDI100 expresses the absorbed dose for a scanned length of 100 mm and is calculated as follows: +50mm 1 CTDI = D z z T ò ( )d n -50mm FIGURE C.127 Image (section) of x-ray tube STRATON – note the con- CTDI100 is a strictly defined term. It is measured in stan- struction, which allows cooling of the anode by rotating of the whole tube dard, cylindrical, polymethyl methacrylate (PMMA) phantoms housing. (Image courtesy of Siemens Healthcare.) (Figure C.128). Both phantoms are a minimum of 14 cm long and have diameters of 16 and 32 cm, to represent a head and a body, respectively. Originally, CTDI measurements were usually made the filament, but further applies electromagnetic deflection and with thermoluminescent (TLD) dosimeters, but currently, special- focusing of the beam of thermal electrons, this way making the ist 100 mm long pencil ionisation chambers are most commonly focal spot size independent of the filament size. In order to dis- used (Figure C.129). The absorbed dose is quoted as dose to air; sipate the high thermal energy imparted to the anode, this tube therefore, for a dosimeter calibrated in air kerma, the dosimeter construction uses a stationary round anode, sealed with the tube reading can be used directly, and the integral of the dose pro- metal envelope (Figure C.127). This way the whole assembly file over 100 mm obtained by taking the product of the dosimeter (envelope plus anode) rotates. This allows for leaving the backside reading and the active chamber length. The standard CT dose of the anode (and the bearings) to be directly open and cooled by the insulating oil of the tube housing. The power of STRATON is very high; the focal spot very small; the cooling time very short (down to 20 s with cooling rates of 4.7 MHU/min). ‘Body’ ‘Head’ A new x-ray tube developed by Philips (iMRC) also uses external focusing of the beam of thermal electrons from the cath- P1 ode, but the large rotating anode is inside the metal envelope. The power of this new tube reaches 120 kW. Most new CT x-ray tubes allow automatic tube current modu- C P4 P lation (ATCM) – adjusting the x-ray tube current according to the 2 variations in patient attenuation (see the eponymous article). 140 mm Related Articles: X-ray tube, CT scanner, Automatic tube cur- 160 mm rent modulation, Heat Unit P3 Further Readings: Advanced Technology Sets New Speed Standards in CT, Siemens Medical Solutions, www .m edica l .sie 320 mm mens. com /s iemen s /en_ US /rg _marc om _FB As /fi les /P ress_ Relea ses /2 004 /P DF /00 3 .04_ speed 4D .pd f; (accessed on 31 July 2012). FIGURE C.128 Diagram of standard PMMA CT dosimetry phantoms. CTDIw 224 Cumulated activity 100 mm C FIGURE C.129 CT pencil ionisation chamber. phantoms have holes at appropriate locations, to allow placement of the dosimeter (pencil ionisation chamber). Measurements are made at the centre of the phantoms, CTDI100,c and at the periph- ery, at a depth of 10 mm, CTDI100,p. The average absorbed dose in the scan plane of the phantom is expressed by taking a weighted average of centre and periphery CTDI100 measurements. This weighted average, CTDIw, is calcu- lated using the following formula: 1 2 CTDIw = CTDI100,c + CTDI100,p 3 3 The preceding equation represents the absorbed dose for a con- tiguous scan, i.e. a sequential (axial) scan where the table incre- ment between rotations is equal to the nominal beam width, or a helical (spiral) scan with a pitch value of 1. For non-contiguous scanning, the CTDIw can be adjusted to give the volume CTDI, CTDIvol, using the following formula: CDTI CTDI w vol = Pitch CTDIvol values are displayed on the scanner console. It should be noted that this quantity does not represent the dose to the patient, but is a dose index that can be used for comparing dose between dif- ferent protocols, different scanner models for the same examination and between patients of a standard size. Another point to note is that the value displayed on the scanner is not a measured value,

but FIGURE C.130 Photo of the standard 16 cm computed tomography dose index (CTDI) phantom with a CT chamber in the central position and the is calculated using data typical for that scanner model. Therefore, peripheral positions marked from 1–4. independent measurements of CTDI should be made at regular intervals to ensure that the scanner is performing to specification. Cumulated activity CTDIw (Nuclear Medicine) When trying to determine radiation doses (Diagnostic Radiology) The weighted computed tomography to patients in therapeutic or diagnostic nuclear medicine appli- dose index (CTDIw) combines values of CTDI100 measured at the cations, one uses the cumulated activity Ã. Ã is a measure of the centre and periphery of the standard 16 cm head or 32 cm body total number of disintegrations in a source organ between initial phantoms. It is given by: uptake and total clearance. The cumulated activity depends on two kinetic properties of the used radio-compound, i.e. (1) how 1 CTDIw = (CTDIPMMA,100,C + 2CTDIPMMA,100,p ) much and how fast is the activation accumulated in each source 3 organ and (2) how long it takes until all activity is cleared. The cumulated activity is the product of these two factors and the Where the CTDIPMMA,100,c quantity is measured at the centre of SI unit is ‘Bq × s’. In other words, the cumulated activity is the the standard CT phantom and CTDIPMMA,100,p is the average of the total number of disintegrations in the source organ between ini- measured at four positions (3 o’clock, 6 o’clock, 9 o’clock and 12 tial accumulation and total radionuclide clearance. o’clock) around the periphery of the phantom (see Figure C.130). The temporal and spatial distribution of a radiotracer in the CTDIw is an approximation to the average absorbed dose for a body depends on a number of radio-compound properties: deliv- single rotation of the scanner. A further value, CTDIvol, takes into ery, uptake (accumulation), metabolism, clearance, excretion and account the helical pitch or axial scan spacing. physical decay. After injecting activity, the concentration of radio- Related Articles: CTDI nuclides (i.e. activity) in a source organ will change over time, Cumulative dose 225 C upping artefact which is described by a time-activity curve A(t). If the time-activ- ity curve is known, then the cumulated activity equals the area Scenario 4: When the uptake is not near-instantaneous, under the curve, i.e. an integration over A(t) from t = 0 to infinity: i.e. when a significant amount of the activity decays before accumulating in the organ, the equations mentioned ear- ¥ lier will overestimate the organ doses. In most cases, the A = òA(t )dt (C.13) uptake can be described by 0 A(t) = A0 (1- e-0.693t /Tu ) (C.17) t = 0 represents the time for activity administration. As previously mentioned, the cumulated activity depends on a number of factors and the time-activity curve is unique for each source organ and where Tu is the biological uptake half-time. In such a sce- C radionuclide. Four scenarios are mentioned in the following text, nario, the cumulated activity is given by where the cumulated activity is estimated. æ T A = . A T ue ö 1 44 0 e ç ÷ (C.18) è Tu ø Scenario 1: The uptake in the source organ is near-instan- taneous and there is no biological excretion (i.e. only a where Tue is the effective uptake half-time derived from physical decay component). Cumulated activity is then described by an exponential radioactive decay function. T T uTp The cumulated activity for such a scenario is ue = (C.19) Tu + Tp A = 1.44Tp A0 (C.14) Related Articles: MIRD formalism, Equilibrium absorbed where dose constant, Absorbed fraction, Mean dose per cumulated T activity p is the physical decay constant A0 is the activity initially present in the organ. Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA, pp. 407–411. Scenario 2: The uptake is near-instantaneous and the bio- Cumulative dose logical half-life Tb is very short compared to the physical (Radiation Protection) The total dose (defined in terms of either decay, i.e. biologcal excursion only. The biological decay the absorbed dose, equivalent dose or effective dose) that would function often consists of several exponential excursion be received by an individual from repeated exposures to ionising components with a fraction f1 of the initial activation radiation over a period of time is called the cumulative dose. This A0 being excreted with halflife Tb1 and fraction f2 being is distinct from the committed dose received over time from a excreted with half-life Tb2 and so on. In such a scenario, the single injection, inhalation or ingestion of a radioactive substance. cumulated activity is given by Related Articles: Absorbed dose, Equivalent dose, Effective dose, Committed dose ¥ ¥ A = A0ò f e-0.693t /Tb1 1 dt + A ò f e-0.693t /Tb2 0 2 dt + 0 0 (C.15) Cumulative dose volume histogram A = 1.44Tb1 f1A0 +1.44Tb2 f2 A0 + (Radiotherapy) See Dose volume histogram Cupping artefact (Diagnostic Radiology) A cupping artefact is a typical beam- Scenario 3: Near-instantaneous uptake with a significant hardening artefact in CT imaging. It will appear in an image as contribution from both biological excursion and physi- increased darkness near the centre. As the x-ray beam passes cal decay. If assumed that the biological half-life Tb is into the body it is filtered by the outer regions shifting the described by a single-component exponential decay, the x-ray spectrum to higher energies or becoming ‘harder’. This physical and biological half-life can be expressed as an alters the attenuation rate in the centre and errors in the CT effective half-life Te given by equation C.16a. For such a numbers compared to the outer regions. The lower CT num- scenario, the cumulated activity is given by equation C.16b. bers will appear darker in an image. In an image of a uniform phantom (water) the centre region will appear darker giving T it a ‘cupped’ appearance. This artefact is most prominent in T pTb e = ( (C.16 Tp + Tb ) a) head scans particularly in the region of posterior fossa because of the filtering produced by the skull. Another beam-harden- ing artefact is the metal artefact that produces bright streaks A » 1.44Te A0 (C.16b) radiating out from metal objects in the body or edges of some bones. (a.k.a. streaks artefact). It is possible to minimise these If the biological half-life is a multi-exponential decay, the artefacts with specific software, but in this case the CT num- effective half-life for each component is given by equation bers’ accuracy is affected. C.16a. The cumulated activity is calculated with the effective Related Articles: Beam-hardening, Metal artefact half-lives replacing the biological half-life in Equation C.15. Hyperlinks: www .r adiol ogyca fe .co m /rad iolog y -tra inees /frcr -phys ics -n otes/ ct -ar tefac ts Curie 226 Curvilinear array transducer Curie (Radiation Protection) Curie (Ci) is an old unit of radioactivity, defined as b 1Ci = 3.7´1010 decays/s a Originally, the curie was chosen as roughly the activity of 1 g of c the radium isotope 226Ra. The SI unit of activity (radioactive decay) is becquerel (Bq), C which is to one decay per second, this way 1Ci = 3.7´1010 Bq The curie is named after Pierre Curie and Marie Sklodowska FIGURE C.131 Electron treatment of the chest wall. Curie. Related Article: Becquerel in region ‘a’. This is an effect of patient curvature and becomes Current consumption greater as the obliquity becomes greater. As the angle of obliquity (General) Current consumption is the current that an electric cir- increases due to curvature, it is often best to arrange the angle of cuit draws from the voltage or current source. incidence of the central axis so that the angles of obliquity at the Related Article: Current source two extreme field borders in this case ‘b’ and ‘c’ are about equal if possible. Current, eddy The problem of oblique incidence caused by patient curva- (Magnetic Resonance) See Eddy currents ture is also important in electron arc therapy. Efforts are made to minimise this by placing the linear accelerator isocentre (the Current intensity rotational axis) as close as possible to the centre of curvature of (General) This term normally refers to the intensity of the anode the region to be treated. current (tube current) in an x-ray tube. Current intensity is directly Related Articles: Electron arc aperture, Electron oblique related to the density of the thermal emission current (see articles incidence on Filament heating and Tube current). The current intensity (the Further Reading: Mayles, P., A. E. Nahum and J-C. number of electrons per unit area) is measured in A/mm2 but, as Rosenwald. 2007. Handbook of Radiotherapy Physics – Theory the source of these thermal electrons (the cathode filament wire) and Practice, Taylor & Francis Group, New York, pp. 701–708. is with standard size for a particular x-ray tube, simply mA is used in practice. Current intensity is linearly related to the intensity of Curvilinear array transducer the x-ray photons in the beam. (Ultrasound) Curvilinear arrays, also referred to as convex arrays, Related Articles: Filament heating, Tube current are array transducers in which the elements, typically 64, 128 or 256 are arranged in an arc (Figure C.132). This format allows for Current source a large field of view from a small contact point (Figure C.133). (General) A current source is an electrical or electronic device The transducer is designed depending on its application; the main that delivers or absorbs electric current where the current is inde- pendent of the voltage across it, in the ideal case. The ideal current source has infinite internal resistance where real current sources can be represented as ideal current sources in parallel with a resis- tance. The simplest current source consists of a voltage source in series with a resistor. Curvature correction (Radiotherapy) Patient curvature can be important when planning A B C D treatments with electron fields. This can be illustrated by consid- ering the set up shown in Figure C.131, where a single electron field is used to treat a chest wall area. Two factors affect the dose E F G just under the skin surface – the increase in SSD at different parts of the beam as it enters the patient, and also the effects of oblique incidence. H J With electron treatments, oblique incidence can give rise to electrons depositing dose non-uniformly across the field area (see Figure C.131). In region ‘a’, the initial direction of travel of the electrons is perpendicular to the skin surface whereas this is not the case in regions ‘b’ and ‘c’, and the obliquity in these regions will result in a relatively higher dose there than at points in region FIGURE C.132 Array transducers. The top row shows linear array ‘a’. This is because the electrons are effectively side scattered transducers (A–D), the middle row curvilinear transducers (E–G) and the to a higher degree in the tissue instead of forward scattered as lower row phased array transducers (H and J). Custom blocking 227 CW Doppler C FIGURE C.133 Curvilinear array image of a kidney. Note the curved contact surface at the skin. FIGURE C.135 The image shows the elements in an endovaginal trans- ducer. An x-ray of the elements shows the arrangement of the elements over 180° to provide a wide field of view. Block transmission factors and tray transmission factors must be known for treatment planning. Other solutions to customised blocks are possible, such as the computerised milling blocks of lead directly from shapes drawn of radiographs. The shape of the shielded area is drawn on a radiograph (a). The mould shape is cut by moving the end of the hot wire along the line drawn on the radiograph (b). Low-melting-point alloy in liquid form is poured into the mould (c) and allowed to cool and set. The Styrofoam is then removed leaving the block (d) and this can then be placed on the linear accelerator (linac) blocking tray (e). Abbreviation: TPS = Treatment planning system. Related Articles: Critical structures, Organ at risk, FIGURE C.134 The width of the elements in a low-frequency array is wider than for a transducer designed for superficial imaging. Conventional radiotherapy, Block transmission factor, Low-

melting-point alloy Further Reading: Podgorsak, E. B. 2003. Review of Radiation factors are length of the array, radius of curvature and frequency. Oncology Physics: A Handbook for Teachers and Students, The width of the elements also dictates its performance; the aper- International Atomic Energy Agency, Vienna, Austria. ture in the elevation plane is designed to focus at particular depth Cut film changer (Figure C.134). (Diagnostic Radiology) Another name for a type of spot-film Tightly curved arrays are used in intracavity probes (link) camera. to provide large fields of view from small contact areas Related Article: Spot-film camera (Figure C.135). Related Article: Convex array Cutoff frequency (Nuclear Medicine) In nuclear medicine, the cutoff frequency Custom blocking typically refers to a frequency in a filter where frequencies on (Radiotherapy) Some treatment units have only collimators that one side of the cutoff frequency are weighted in order to enhance form rectangular fields (not multileaf collimators). Since treatment certain features in a filtered image. Generally, image filtering is volumes are rarely rectangular, high-density shielding blocks performed in Fourier space. In Fourier space, the image data are are used to protect normal tissue and critical structures within represented as a series of sine and cosine functions, where the the irradiated area. The blocks are either individually designed high-frequency components represent high contrast differences, blocks fabricated from a low-melting-point alloy (custom block- for instance at boundaries between bone and tissue, while low- ing) or standard (library) blocks that may be purchased from the frequency components represent the overall image intensity pat- vendor of the treatment machine. All these blocks are placed on tern with low contrast differences. In a low-pass filter, for example a plastic tray to correctly position them within the radiation field. frequency components over a certain frequency threshold are Customised blocks are fabricated as follows. The area to be weighted to decrease the signal to noise ratio. blocked is drawn on a radiograph or exported directly from TPS. Related Articles: Ramp filter, Signal to noise ratio (SNR) A block cutting machine cuts the shape of the block in a piece of Styrofoam. This mould is filled with low-melting-point alloy. CW Doppler Then the block is removed from the Styrofoam. Customised (Ultrasound) CW Doppler is short for continuous wave Doppler. blocks follow the beam divergence (see Figure C.136). The Doppler shift arises when there is a relative motion between Cyberknife 228 Cyberknife Hot wire Low melting point alloy Styrofoam block poured into mould Linac blocking tray C (b) (c) Radiograph (e) Shielding block (d) (a) FIGURE C.136 Customised block cutting using a hot wire cutter and Styrofoam mould. the transmitter and receiver of sound. In diagnostic ultrasound, Transmitter there is no relative motion between these two, but when the emit- Oscillator ted sound is reflected off a surface, or scattered by particles that Audio are moving, there will be an apparent shift in wavelength, and amplifier thereby frequency, which is dependent on the relative velocity Receiver between the surface/particles and the transducers. Lowpass CW Doppler refers to the principle of emitting a continuous filter frequency (or tone) with a single transducer, and receiving on a Mixer To earphones second transducer. As these two cannot be located at the same position, there is bound to be an angle between them, and thereby FIGURE C.137 A block diagram showing a simple CW Doppler system. an overlapping zone between their respective beam patterns. This overlapping zone is referred to as the sample volume from where Doppler shifts can arise. It can be shown that the resulting Doppler shift frequency, target (i.e. the sign of the Doppler shift) is required, a quadrature the difference between the received and transmitted frequencies, demodulation of the received signal is necessary. obeys the following relation when the velocity of the target is Related Article: CW (continuous wave) much smaller than the speed of sound: Cyberknife 2vf (Radiotherapy) This machine manufactured by Accuray deliv- Df = 0 cosq (C.20) c ers multiple highly focused beams of radiation using a robotic 0 arm capable of delivering hundreds of uniquely angled beams in where a single fraction. The system uses a 3D, non-coplanar workspace ∆f is the Doppler shift frequency to provide highly conformal treatments of irregularly shaped v is the velocity of the target tumours, resulting in an enhanced ability to spare surrounding f0 is the transmitted frequency healthy tissue and structures in close proximity to the treated c0 is the speed of sound volume. θ is the angle between the transmitter/receiver combination It is capable of isocentric or non-isocentric treatment deliv- and the velocity of the particle ery to a targeting accuracy of 0.5 mm for targets unaffected by motion and 0.7 mm to those affected by motion, for example in This is known as the Doppler equation and also assumes that the case of tumours affected by respiration. the angle between the transmit and receive beams is small. Treatments are image guided using two kV x-ray beams A practical implementation of a CW Doppler system is shown projected from the treatment room ceiling (see x-ray sources, in Figure C.137, where the received signal is mixed with the Figure C.138), which pass through the patient onto imaging tab- transmitted frequency. Thus, both sum and difference frequen- lets imbedded in the floor (see image detectors, Figure C.138). cies are obtained, where the difference frequency corresponds to These images are compared to gold standard, digitally recon- the Doppler shift as given by the Doppler equation. In the case structed images taken from the treatment planning CT; any move- depicted here, the low-pass filter may actually be omitted since ment in the position of the target or patient can then be measured the frequency response of the audio amplifier is much below and automatically corrected for by repositioning of the beam the sum frequency produced in the mixer. If the direction of the geometry to ensure submillimetre accuracy throughout treatment. Cycles per degree 229 C yclotron X-ray sources Optical C imaging system Linear accelerator Image detectors FIGURE C.138 Cyberknife system. (Courtesy of Accuray.) A typical treatment is usually delivered in between 1 and 5 radionuclides. The basic principle is to accelerate electrically fractions (6–25 Gy); treatment time varies from 30 to 90 min charged particles, typically protons, and direct them towards (depending on the location, shape, size and complexity of the a target. Accelerated particles can cause a nuclear reaction tumour). in the target content and produce radionuclides. The typical Respiratory-induced motion of tumours, for example in the cyclotron design is described in Figure C.139 and it consists of case of lung, can cause significant targeting uncertainty. With two opposite semicircular metal electrodes, often referred to Cyberknife, the motion of fiducial markers placed inside the as ‘dees’ because of their shape. The two dees are positioned tumour prior to treatment is correlated with that of external LED horizontally between the two poles of an electromagnet. In the optical markers placed on the patient’s chest or abdomen. centre between the two dees is an ion source, S, used to gener- During treatment, the surface markers can be tracked with ate charged particles. This entire ‘package’ is kept in vacuum an optical camera (see Optical imaging system, Figure C.138), at ∼10−3 Pa. allowing the machine head to move and follow the motion of During operation, the particles are accelerated in the electric the tumour in real time when the correlation between the two is field in the gap between the two dees. Inside the dees, there is known. The x-ray imaging system captures the position of the no electric field, just a magnetic field that will curve the particle internal fiducial markers at intervals during treatment so that the beam. The proton beam will follow a circular path until extraction correlation can be updated and the motion of the treatment head of the accelerated beam. corrected if necessary. The particle energy in a cyclotron system is given by Cycles per degree 2 ( 4.8´10-3 (H ´ R ´ Z (General) Cycles per degree is a measure for visual acuity (i.e. for E MeV) ) = A spatial resolution), used in optics. One cycle is a pair consisting of one white line and one black where line. The number of pairs (cycles) which can be distinguish at one R is the radius of the particle orbit in cm angular degree of a visual field at specific distance represents the H is magnetic field strength in tesla visual acuity. A and Z are the atomic number and the mass number of the Normally humans can see 30 cycles per degree. The human accelerated particle, respectively eye can distinguish an object of the order of one minute of an arc (1/60 degree) – see more in Visual acuity. Consequently, the principal limitations in particle accelera- Related Articles: Visual acuity, Retina tion are dees radius and the magnetic field strength. When the Further Reading: Hecht, E. 1987. Optics, 2nd edn., Addison particles are accelerated to the maximum energy, they can be Wesley; Koren, N. Understanding image sharpness, www .nor- extracted in two different ways; a target can be placed in the par- mankoren .com /Tutorials /MTF .html. ticle beam path, which is called internal beam irradiation. The other approach is to extract the beam from the cyclotron and Cyclotron direct it towards an external target, i.e. external beam irradia- (Nuclear Medicine) A cyclotron is a charged particle accelera- tion. The stripping (i.e. the extraction of the beam) method differs tor commonly used to produce radionuclides for nuclear medi- from positive and negative ion cyclotrons. Positive ion cyclotrons cine imaging and radiotherapy. Many institutions and hospitals use an electrostatic deflector to divert the beam onto the target. A have their own cyclotron for onsite production of short-lived negative ion cyclotron typically accelerates H− ions (one proton Cyclinac 230 Cyclotron-produced radionuclides and two electrons). When the beam has reached its outermost Cyclotron for proton therapy orbit, it is passed through a carbon foil where the two electrons (Radiotherapy) Cyclotrons are a class of circular accelerators are ‘stripped’ from the atom, leaving only a proton and a posi- which are used in particle therapy due to their efficient size com- tive charge. The proton will bend in the opposite direction onto pared to conventional linear accelerators. How cyclotrons accel- the target. Using this stripping technique, one could extract two erate charged particles is described in detail under Cyclotron. beams simultaneously by using two foils. The first foil is posi- When a cyclotron is part of a particle therapy system, it has a tioned so that only half of the beam is stripped, leaving the rest few additional features (Schippers, 2018). For example, in proton of the beam for the second target. Due to the unstable nature of therapy, an internal proton source, i.e. hydrogen gas, is located at the H− ion, the negative ion cyclotron must be operated at 10−5 Pa the centre of the cyclotron where a block of metal is connected to C (10−3 for positive ion cyclotron). a high potential difference. This high voltage is energetic enough Related Articles: Internal beam irradiation, External beam to ionise the hydrogen gas which flows in a chimney in the centre irradiation of the cyclotron. The protons from the ionised hydrogen gas then Further Reading: Cherry, S. R., J. A. Sorenson and M. E. enter the dee rings of the cyclotron. The cyclotron outputs a fixed Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, beam energy, but for particle therapy we almost always require a Philadelphia, PA, pp. 49–51. range of energies. An energy degrader is therefore added to the exit of the cyclotron. The energy degraders are usually carbon Cyclinac wedges which degrade the beam energy by increasing or decreas- (Radiotherapy) A novel design of accelerator for ion therapy ing the amount of carbon material the beam travels through. For that couples a cyclotron (or a synchrocyclotron) injector with a more efficient acceleration, the cyclotron may have multiple (more linac booster. Cyclinacs have been proposed as a pragmatic way than two) dee rings. For more efficient use of space, the cyclo- to achieve the high energies needed for proton CT and proton tron may use superconducting magnets. The diameter of a proton radiography (up to 350 MeV). Beams extracted directly from the

cyclotron is typically 3.5–5 meters. For carbon ions, the diameter cyclotron could be used for additional purposes such as radio- is about 6 meters. isotope production for diagnostics. In principle, having a linac The advantages of cyclotrons are: high intensity, accurate and booster lends itself to fast energy switching, a desirable feature adjustable intensity and continuous beam. for pencil beam scanning. The disadvantages are that a degrader is needed (which can Related Articles: Linear accelerator, Cyclotron, Cyclotron become radioactive) and it only accommodates one particle type. produced radionuclides, Synchrocyclotron, Pencil beam scanning Related Articles: Cyclotron, Cyclotron electrodes (dees), (PBS), Proton CT (pCT), Proton radiography Positive ion cyclotron, Linear accelerator, Degrader. Further Reading: Schippers, J. M. 2018. Cyclotrons for par- ticle therapy. arXiv preprint arXiv:1804.08541. Cyclotron electrodes (dees) (Nuclear Medicine) Most cyclotrons consist of a pair of hollow Cyclotron target semicircular metal electrodes shaped like a D, hence the name. (Nuclear Medicine) A cyclotron target is a chamber with a gas, The dees produce a magnetic field that is used to bend the parti- fluid or a solid piece of material onto which a particle beam is cle beam. They are housed in a vacuum chamber that is in a uni- directed to produce radionuclides. For 18F production, the ideal form magnetic field provided by an electromagnet. Between the target is constructed in a material with a few key characteristics: dees, there is maintained a potential difference that alternates in time. This potential difference creates an electric field across 1. Chemically inert the gap between the dees. An ion source is placed in-between 2. High thermal conductivity the two dees to produce charged particles. The particles are 3. Low activation of target chamber accelerated by the electrical field in the space between the two 4. Low reactivity of the 18F fluoride produced dees and bent in the magnetic field inside the dees (note that 5. Low contaminants in the 18F fluoride produced the electrons are only accelerated in the gap between the dees) (Figure C.139). The target body should be chemically inert to prevent the pro- Related Article: Cyclotron duced radionuclide from reacting with the target body. The high- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. intensity beam irradiates a small area of the target and produces a Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, lot of thermal energy and it is therefore important to have an ade- Philadelphia, PA, pp. 49–50. quate cooling system and a material with good heat conduction, for example in a water target with insufficient cooling, the water may evaporate due to the temperature rise. Additional desired fea- tures are low activation of the target body to prevent unnecessary radiation dose to cyclotron operators and staff, low reactivity of Magnet the 18F produced and minimal contaminations to the final product. S Dees S Cyclotron-produced radionuclides Magnet (Nuclear Medicine) Radionuclides produced in a cyclotron often Accelerated have similar characteristics. Typically, positive particles are particle beam accelerated and added to the nucleus: Target 1. Therefore, most cyclotron products lie below the line of stability. These isotopes tend to decay via EC and β+ FIGURE C.139 Schematic view of a positive particle cyclotron. Top (left) and side (right) view. S is an ion source. decay. Cylindrical ionisation chamber 231 Cylindrical ionisation chamber 2. The atomic number of the product isotope is often e– C changed in the process, which changes the chemical Signal properties. The product is therefore often considered to Ions+ be carrier-free. 3. Since the cross section for protons is smaller relative to neutron cross section, the quantities of radioactiv- R ity produced in a cyclotron are lower than in a reactor. Therefore, the price of cyclotron products tends to be more expensive than reactor products. + – Regardless of the price, there are a number of good reasons for C cyclotron production. A cyclotron provides onsite production of FIGURE C.140 Schematic diagram of a cylindrical ionisation chamber. short-lived radionuclides like 11C (T½ = 20 min), 13N (T½ = 10 min), 15O (T½ = 2 min) and 18F (T½ = 110 min). 11C, 13N and 15O are com- mon elements in biological substances and they can be labelled b is a radius of an outer electrode (cathode) with a wide variety of biologically relevant tracers. 18F is the most a is a radius of the anode commonly used beta emitter and it is frequently labelled with a glucose analogue fluorodeoxyglucose (FDG) to study the glucose The average energy to produce an ion pair in air is ∼35 eV and metabolism in the body. 18F has a longer half-life, which allows for other typical gases is in range 20–40 eV. If a beta particle of transports from a regional cyclotron to adjacent hospitals. FDG energy about 1 MeV enters the chamber, it will be absorbed in the is used primarily in oncology but also for brain and heart scans. gas and as result the number N of ion pairs produced, for example Related Articles: Cyclotron, Carrier-free specific activity in air, is: (CFSA) Further Reading: Cherry, S. R., J. A. Sorenson and M. E. 1 MeV N = » 2.9´104 ion pairs Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, 35 eV Philadelphia, PA, pp. 51–52. and it corresponds to an electric charge Q = N × e, where e ≈ 1.6 × 10−19 C, then Q ≈ 4.6 × 10−15 C Cylindrical ionisation chamber The amplitude of the voltage pulse V depends on the capaci- (Radiation Protection) Figure C.140 shows a schematic diagram tance C of the chamber as follows: of a cylindrical ionisation chamber filled with a gas. Cylindrical ionisation chambers are used for the determination of the radia- Q tion exposure (dose), for example a pocket dosimeter for x-rays. V = C They are also applied for measuring absorbed dose in beams of heavy particles in depth to which the measured dose refers. The unit of radiation exposure is defined as the amount of electric In the cylindrical ionisation chamber, the electric field lines charge produced in a unit mass of the air by x- or gamma radia- are in radial direction and the strength E of the electric field tion. It is expressed in the SI as C/kg or in roentgen (R) used as depends on the distance r according to the following expression: traditional unit: V 1R = 2.58´10-4 C/kg E = ( r ´ ln (b/a)) Related Article: Ionisation chamber where Further Reading: Knoll, G. F. 2000. Radiation Detection and V is a voltage applied to the electrode Measurement, 3rd edn., John Wiley & Sons, Inc., New York, pp. r is a distance from the anode (inner electrode) 136–139. D DAC (digital-to-analogue converter) (General) See Digital-to-analogue converter (DAC) Damping (Ultrasound) The axial resolution in diagnostic ultrasound imag- ing depends on the length of the pulses used. To produce short D pulse duration it is important to reduce the ringing effect in the piezoelectric transducer element so that the pulse length is only a few cycles long. Therefore the natural resonance of the element must be damped. Ringing describes the internal reverberations in the transducer element caused by the large difference (20 times) in acoustic impedance between PZT and soft tissue. A damping, or backing, layer behind the PZT will reduce the ringing (damp the vibrations) and the properties of this damping material should ideally have an acoustic impedance equal to PZT and with good Dark blood absorption characteristics. Mounting matching layers to ensure (Magnetic Resonance) Dark blood or ‘black blood’ imaging is a efficient transmission of sound into the body in front of the trans- general term for MRI techniques used to null signal from flowing ducer will also help to reduce the ringing effect. blood to provide improved blood/tissue contrast. The pulse duration is one property used to define the damp- Basic spin echo (SE) sequences demonstrate an inherent black ing efficiency of the transducer. The bandwidth and Q-factor are blood characteristic. Signal in SE sequences is generated through others. A short pulse gives rise to a broad bandwidth and a low Q. excitation of a selected slice with a 90° RF pulse followed a time Q is defined as the ratio of the transducer centre frequency to its TE/2 later with a 180° refocusing pulse. Spins associated with blood bandwidth (Figure D.1). flowing into or out of the selected slice in a time <TE/2 will not be Related Articles: Attenuation, Absorption, Matching layer, excited by both RF pulses and do not contribute strongly to signal. Bandwidth, Axial resolution As a result blood flow in SE images will appear dark. This holds only where flow is perpendicular to the slice and is sufficiently fast. Damping block Slow flow or flow in the plane of the slice will be exposed to both (Ultrasound) The damping (backing) block (also known as damp- the 90° and 180° pulses and will contribute to signal. ing layer) is a basic component common to the design of all ultra- In double inversion recovery dark blood imaging two 180° sound transducers. The transducer’s sound-generating element is preparatory pulses are used to null flowing blood (see Figure D.2). the piezoelectric crystal array, which is commonly constructed The first inverting pulse is non selective and inverts all longitu- from lead zirconate titanate (PZT). PZT material’s primary disad- dinal magnetisation within the receptive volume of the RF trans- vantage is an acoustic impedance about twenty times greater than mit coil. The second pulse is selective and restores longitudinal soft tissue. This results in the majority of the ultrasound waves magnetisation in the selected slice. The inverting pulse has no being reflected back at the transducer-tissue interface and rever- net effect on the tissue in the slice. Inflowing blood is however berating back and forth (ringing) within the PZT layer. inverted, with its longitudinal magnetisation recovering towards The damping block has an acoustic impedance slightly lower zero with a T1 time constant. A delay time TI is left prior to start- than that of the PZT material. This provides the best compro- ing the imaging sequence proper or ‘host’ sequence. The delay mise, as it prevents long-lasting internal reverberation through time is chosen so that blood longitudinal magnetisation is zero acoustic absorption (damping) to the sound waves travelling when the imaging sequence starts and cannot contribute to signal. rearward, while not adversely affecting system sensitivity. The Double inversion pulses are typically used in conjunction with a sensitivity is partly preserved by the damping layer’s acoustic fast SE or single shot (i.e. HASTE) SE sequence. impedance being different from that of the PZT material, as this In triple inversion recovery imaging, a third inversion pulse enables some sound waves to be reflected from the damping layer is added to null both fat and blood – essentially an extension of back towards to the PZT layer. If both the damping block and principles of the STIR imaging technique. PZT layer had the same acoustic impedance, all the sound waves Black blood imaging techniques are used extensively in would be absorbed. cardiac imaging and imaging of the great vessels and carotids. The damping block works along with the matching layer which Figure D.3 demonstrates the tissue/blood contrast achievable in a sits in front of the PZT layer. One or more matching layers are double inversion recovery image. similarly used to ‘step down’ the impedance difference between the PZT and tissue. The acoustic properties of both the damping Dark current block and matching layers are selected to maximise sensitivity (Diagnostic Radiology) Dark current is the current (a small while reducing unwanted internal reverberations. constant amount) through a photodetector (such as photodiode, 233 Dark current 234 Dark current Transducer elements Pulse excitation Displacement Long pulse length Undamped transducer element Poor axial resolution Pulse excitation Displacement Backing D Short pulse length Damped transducer element Good axial resolution FIGURE D.3 Cardiac double inversion MRI image showing good con- FIGURE D.1 The backing material reduces the ringing. (Courtesy of trast between myocardium and blood. EMIT project, www .emerald2 .eu) Longitudinal magnetisation Mz in flowing blood Tl set to null blood at start of imaging host 180° 180° RF transmit Tl Imaging host sequence Non-selective Selective inverting inverting pulse pulse Selected slice Blood Blood flow Stationary tissue (i) Arrows (ii) Non-selective (iii) Selective

(iv) At delay represent inverting inverting time Tl in logitundinal pulse pulse restores flowing components of inverts magnetisation blood long. magnetisation magnetisation in slice to be magnetisation imaged is zero FIGURE D.2 Double inversion recovery black blood imaging. Darkroom 235 Daughter radionuclides photomultiplier PMT, TV camera tube, etc.) in the absence of into independent 2D slices and reconstructed using filtered back incident light (radiation). projection. Instead a three-dimensional reconstruction algorithm The dark current consists of the following basic components: is used. These algorithms are more time consuming than the con- ventional 2D reconstruction. But since PET scanners operated in • Photocurrent – generated by background radiation 3D mode demonstrate such high sensitivity, 3D mode is available • Saturation current – at the semiconductor junction in new commercial PET scanners. Due to its high sensitivity PET is well suited for dynamic studies. With dynamic studies it is pos- In cases of precise measurements, dark current should be taken sible to gain knowledge about tissue function by calculating blood into account upon calibration of the photosensitive devices (PMT, flow or studying radiopharmaceutical delivery, accumulation and photodiodes, etc.). It is also considered as a source of noise. clearance. To reduce the influence of dark current over the total signal, Related Articles: PET, Field of view, Beta decay it is compensated either by subtracting it from the final signal or Further Reading: Cherry, S. R., J. A. Sorenson and M. E. reducing to 0. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA, pp. 350–353. Darkroom D (Diagnostic Radiology) A darkroom is a room in which radio- Database graphic or photographic film can be handled and processed safely (General) Database is a collection of data, structured in a specific without being exposed to light. Most types of film are not sensi- way and stored in the computer system. There are various medical tive to all wavelengths or colour of light. Therefore a darkroom imaging databases collecting and organising images of anatomi- can be illuminated with light that is not in the sensitive range of cal structures in norm and pathology. the film to permit human vision in the area. This illumination is provided with ‘safelights’. The colour or wavelength of the light Data mining is determined by the type of light source or filter. This should be (General) Data mining is the process of extracting useful informa- selected to ensure that the light does not extend into the sensitive tion from raw (‘dirty’) dataset. Data mining is a computer-driven spectral region of the films that are being used. process, although there are various techniques and strategies used Quality assurance procedures include inspecting and testing a to perform data mining. The most common way of data mining darkroom for light leaks from the exterior and to ensure the film is software analysis searching for pre-defined patterns or trends is not sensitive to the safelight being used. within big data sets. Related Article: Film processing Data mining stages: • Data collection Data acquisition PET • Data storage and management (Nuclear Medicine) There are two main types of data acquisition • Data structure analysis and pattern definitions in dedicated positron emission tomography (PET) scanners, two- • Data processing and three-dimensional acquisitions. In most dedicated PET scan- • Results ners the detectors are placed in rings along the through-plane. Data mining in healthcare is the area where statistics, computer 2D Data Acquisition Mode: In old scanners the rings were science and artificial intelligence work together to improve separated with axial lead septa between each ring. This acquisi- healthcare. Most common applications of data mining in health- tion mode is called 2D data acquisition, and only photons par- care include, but are not limited to: risk analysis/management; allel to the plane of the detector ring are allowed to reach the early detection/prevention; research; optimising patient flow; detector. This means that a coincidence can only be registered measuring performance, etc. between two detectors in the same ring. This mode is not very sensitive to either random or scattered coincidences. Another Daughter nucleus method for acquiring data is to allow registration of photons in (Nuclear Medicine) See Bateman equations in parent–daughter an adjacent ring. If the septa are shortened, more photons will decay be able to reach the adjacent rings. This method is more sensi- tive, i.e. has a higher count rate for each individual detector. A line of response (LOR) between two adjacent detectors is known Daughter radionuclides as a cross plane and the LOR will cross the direct plane in the (Nuclear Medicine) In a decay chain the initial radioisotope, the centre of the scanner. To further increase the sensitivity more parent, disintegrates to another radioactive isotope. The second cross planes can be included. This acquisition method allows a radioisotope is called the daughter radioisotope or decay product. high sensitivity, but does also degrade the spatial resolution in The daughter radioisotope will eventually decay to a third decay the axial direction. There is also a higher occurrence of random product, or grand-daughter radioisotope. This chain will continue and scatter coincidence in the high sensitive mode than in the until one of the decay products is stable. conventional 2D mode. In nuclear physics, a specific part of a decay chain can be of 3D Data Acquisition Mode: Some of the annihilation photons clinical and/or research interest (the parent radioisotope can be absorbed in the septa in a 2D mode could provide true coinci- chosen subjectively). For example, Mo99 disintegrates to the meta- dences. In a 3D mode PET scanner there is no septum between the stable Tc99m. The daughter radionucleus in this example is Tc99m. detector rings, which means that coincidences can occur between Tc99m has some suitable characteristics for imaging, namely the a detector and a detector in any of the opposite rings. Such a scan- photon energy (140 keV) and a high number of gamma rays per ner has a very high sensitivity compared to one in a 2D mode. disintegration (close to 1). The axial sensitivity peaks at the centre of the middle ring(s). It The Bateman equations describe parent–daughter decay. is therefore of great importance to place relevant structures in the Related Articles: Parent radionucleus, Grand-daughter radio- centre of the axial field of view (FOV). 3D data cannot be divided nucleus, Bateman equations, Molybdenum breakthrough DC offset artefact 236 Deadtime losses DC offset artefact Dead time (Magnetic Resonance) This artefact, also called a central point (Nuclear Medicine) Dead time is the duration after each detection artefact, looks like a high-intensity pixel, either brighter or darker event where the detector is unable to acquire new signals even if than the surrounding pixels, at the exact centre of the image. The they do happen. When several events are recorded near-simulta- DC offset artefact is caused by a DC offset voltage in either the neously, the image detector system is unable to separate all the signal amplifiers or the detector. events which leads to a signal pile up. During a period of time Because DC offset in time corresponds to a peak at zero fre- the electronics cannot discriminate any new pulses in order to quency, the Fourier transform of a time domain signal having a accept new events. This time span is referred to as dead time, i.e. constant DC offset is the transform of the signal, which has a zero- no events are acquired. frequency peak. The image in k-space having a DC offset produces A common approach to deal with high count rate is to use dif- a zero-frequency peak when Fourier transformed. Therefore, DC ferent types of circuitry, like discriminators or buffers. The latter offset artefact is a bright spot exactly in the centre of the image. one ‘holds’ a signal and passes it on to the next circuitry when This artefact can be corrected by recalibration and can be it is ready. Another approach is to limit the read out time of the avoided by keeping a constant temperature for the equipment. PM-tube signal, where the signal is set to zero after a specified D Related Article: Quadrature ghost artefact time. This technique causes a degradation of the intrinsic spatial Further Reading: Han, P. K., H. Park and S. H. Park. DC resolution and the energy resolution but it can still prove useful to artefact correction for arbitrary phase-cycling sequence. Magn avoid dead time losses. Reason Imaging 2017 May, 38:21–26. Related Articles: PET, SPECT, Pile up Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA, pp. 178–182. Dead time (Ultrasound) Dead time, dead-zone, or ring-down distance is defined as the distance from the transducer face to the closest target that can be resolved. The transmit pulse may obscure this first part of the image due to the high amplitude reverbera- tions of the pulse within the lens covering the transducer face. These reverberations take time to ring down to the level of small echoes. Most modern transducers have virtually no dead-zone due to greatly improved matching layers and pulse forming techniques. The dead time can be found using a phantom with wire tar- gets close to the surface. Usually there is a group of targets that are equally spaced axially. If, for instance, the third target can be resolved and the spacing is 1 mm, the dead zone is ‘less than 3 mm’. Deadtime losses (Radiation Protection) The period of time when a detector is insensitive for measuring a second event after the first one is called the ‘dead time’. For Geiger–Müller counters, dead time are on the order of 100–500 μs and for NaI(Tl), as well as for semi- conductors, dead times are in the range from 0.5 to 10 μs. As a consequence of the dead time of a detector the number of measured counts is less than the true one. The counts lost during the dead time are called dead time losses. If the counting rate is low the probability of an event arriving during the dead time is low. The dead time losses are a serious problem at high count rates and particularly for a paralysable counting system in which each event introduces its own dead time. If the dead time is τd and measured count rate Nm then the true count rate Nt as equal to DCT (discrete cosine transform) (General) See Discrete cosine transform (DCT) N N = m t ( Dead man’s switch 1- Nmtd ) (Diagnostic Radiology) Switch which interrupts when released. Typical dead man’s switch is the foot switch used in x-ray fluo- At low counting rates (Nm ≪ 1/τd)Nt ≈ Nm. roscopy, which terminates the exposure when the foot is removed At high rates for paralysable counting system: Nm = Nt from the pedal-switch. exp(−Ntτd). Decay 237 Decay scheme energy state to decrease to one half of their initial number. The EXAMPLE: relation between λ and T½ is If τd = 500 μs for a GM counter then for Nm = 10 cps (10 ≪ ln2 0.693 T½ = = 2 × 105): Nt ≈ Nm but for Nm = 10,000 cps (10,000 < 2 × 105) l l Nt < Nm the correction must be made since Nmτd is about 5. Related Articles: Activity, Half life, Radioactive decay Further Readings: Hendee, W. and E. R. Ritenour. 2002. Related Articles: Non-paralysable counting system, Medical Imaging Physics, Wiley-Liss, New York; International Paralysable counting system Commission on Radiation Units and Measurements. 1998. Report Further Readings: Delaney, C. F. G. and E. C. Finch. 1992. 60. Fundamental quantities and units for ionising radiation. Radiation Detectors, Oxford University Press, Oxford, UK, pp. 252–253; Dendy, P. P. and B. Heaton. 1999. Physics for Diagnostic Decay factor Radiology, 2nd edn., Institute of Physics Publishing, Philadelphia, (General) The term ‘decay factor’ is a general term used in con- nection with the function y = a · bx. When the value of b is less D PA, p. 141; Hobbie, R. K. 1997. Intermediate Physics for Medicine and Biology, 3rd edn., Springer-Verlag, New York, p. 489; Knoll, than 1, b is sometimes called the decay factor as 1 is then deal- G. F. 2000. Radiation Detection and Measurement, 3rd edn., John ing with exponential decay. Where b is positive, b is termed the Wiley &

Sons, Inc., New York, p. 124; Saha, G. 2001. Physics growth factor. and Radiobiology of Nuclear Medicine, 2nd edn., Springer-Verlag, The most common application of the equation y = a · bx in New York, p. 91. medical physics is in radioactive decay: Decay N = N e-lt 0 (Radiation Protection) See Radioactive decay Decay constant where (Nuclear Medicine) The decay constant λ of a radionuclide in a The number of radioactive atoms (N) at time t is equal to the particular energy state is the quotient of dP by dt, where dP is the initial number probability that a given nucleus undergoes a spontaneous nuclear N0 is multiplied by the factor e−λt transformation from an energy state: λ is most often called the decay constant but other terms used are transformation constant and disintegration dP constant = l dt Related Article: Radioactive decay In radioactive decay, the constant λ relates the rate of decay (dN/dt) of radioactive atoms to the number of radioactive atoms (N): Decay scheme (General) A decay scheme, in the context of radioactivity, is a dN = -lN two-dimensional representation of the changes and transitions dt involved in the transformation of a radionuclide into a stable The constant λ will have units of s−1, h−1, day−1 or other inverse nuclide or a different radionuclide. The axes represent energy time unit. Other equivalent names for the constant are transfor- (ordinate) and the proton number (abscissa). Decay schemes can mation constant and disintegration constant. be complicated and care must be taken to check whether the For any given radioactive atom there is a constant probability diagram has been simplified. Usually any neutrinos emitted are that it will decay in a stated period, therefore the number of atoms omitted. remaining will decrease in an exponential fashion: A decay scheme will normally show: • Initial radionuclide: Symbol, with atomic mass and atomic number, and half-life. N = N e-lt 0 • Mode(s) of decay: Particle(s) emitted, with percentage or mean number. This equation is the solution of the preceding equation and states • Nuclear energy levels involved of daughter nucleus and that the number of radioactive atoms (N) at time t, is equal to the gamma rays emitted. initial number, N0, multiplied by the factor e−λt. • Daughter nuclide: Symbol, with atomic mass and The real lifetime of any particular radionuclide ranges from atomic number. 0 to ∞; however, for a large number of parent nuclei the mean However, due to complexities of many decay schemes, some lifetime, τ, of a radioactive parent substance equals the sum of of the information may be placed in an accompanying table. lifetimes of all the individual atoms divided by the initial number Figure D.4 and Table D.1 show some of the information about the of radioactive nuclei in the radioactive substance at time t = 0. The decay of 131I to 131xenon. As beta decay is involved the 131xenon decay constant λ and the mean life τ are related through is to the right compared to the 131I. It should be noted that not all transitions lead to gamma rays and internal conversion occurs. 1 t = Related Articles: Beta decay, Decay series, Gamma ray, l Gamma radiation, Radioactivity, Radionuclide Further Reading: Eckerman, K. F. and A. Endo. 2008. It is more common to use the half-life, T½, of a radionuclide, which Radionuclide Data and Decay Schemes, 2nd edn., Society is the mean time taken for the radionuclides in the particular Nuclear Medicine, Reston, VA. Decay series 238 Deconvolution Decay series Some useful amplitude and intensity ratios expressed in dB are (Radiation Protection) A decay series is when there is a series shown in the table. of nuclear transformations, starting with a radionuclide (‘the parent’) and eventually ending in a stable nuclide. The parent nuclide decays into a different nuclide (‘the daughter’) which is itself radioactive and which decays in turn to yet another daugh- dB p2/p1 I2/I1 ter and the series continues until the final daughter is a stable 0 1 1 nuclide. 3 1.414 2 Most naturally occurring radionuclides are members of one 6 2 4 of three series, named the uranium, actinium and thorium series. 20 10 100 The final stable nuclide of each series is one of the stable isotopes 40 100 10,000 of lead. 60 1,000 1,000,000 Related Article: Radioactive decay D Decibel (dB) Related Articles: Intensity, Attenuation (Ultrasound) In ultrasound contexts the decibel notation is used for comparing pulse pressure amplitudes, intensities and for Decommissioning expressing attenuation. The definitions are (Radiation Protection) Decommissioning is defined as the processes carried out such that an item of radiation equipment • Relative pressure amplitude level (dB) = 20 log (p2/p1) may be taken out of clinical use, or such that a radiation facility • Relative intensity level (dB) = 10 log (I2/I may safely be handed over to an alternative use, whether by the 1) • Note that: 10 log (I2/I1) = 10 log (p2/p1)2 = 20 log (p2/p1) employer, or by another organisation. In areas where unsealed radioactive sources have been used this may involve de-contaminating the area of any residual con- tamination. In areas where sealed radioactive sources are used, this may involve removing the sources from any equipment which 131 8.05 d are to be disposed of. 53I β– 1 Decontamination β– (Radiation Protection) Decontamination is the removal or reduc- 2 723 keV tion of contamination by a physical or chemical process. 637 β– 3 Related Article: Contamination Further Reading: International Atomic Energy Agency. 1996. International Basic Safety Standard for Protection against 364 Ionising Radiation and for the Safety of Radiation Sources, γ1 γ2 γ3 γ4 Safety Series No. 115, IAEA, Vienna, Austria. 80 γ5 Deconvolution 0 (General) Deconvolution is a mathematical process to find the input function, which has previously undergone convolution with FIGURE D.4 Decay of 131I. Transitions and energy levels involving <1% some other function. Formally, the problem can be expressed as disintegrations have been omitted. follows: TABLE D.1 Decay of 131I – Transitions Involving <1% Disintegrations Have Been Omitted Mean Number/Transformation Maximum Energy (MeV) β1 0.02 0.25 β2 0.07 0.33 β3 0.90 0.61 Gamma radiation Transition Mean Number / % Mean Number / Mean Number / Energy (keV) Transformation IC Transformation Emitted Transformation IC γ1 723 0.02 0 0.02 0 γ2 637 0.07 0 0.07 0 γ3 364 0.84 2 0.82 0.02 γ4 284 0.06 5 0.06 <0.01 γ5 80 0.06 61 0.03 0.03 Decoupling 239 Deep learning f * g = h where h is the known output g is the convolution function f is the desired input function that is unknown g may be the impulse response function of a known system, in which case it can be fully deconvolved (in the absence of noise) However in many cases g is not known and it needs to be esti- mated before deconvolution can be attempted. The convolution theorem is a useful tool in deconvolution – to extract the input distribution, the Fourier transform of the output, D FT(h), can be divided by the Fourier transform of the convolution function, FT(g), and then take the inverse Fourier transform: æ -1 FT (h) ö f = FT çç ÷ è FT (g) ÷ ø FIGURE D.5 Left lateral decubitus position. This is straightforward for explicit functions however will not be precise if the functions are unknown or contain noise, since the Fourier transforms may not be fully defined. DECT (Dual energy CT) In imaging, deconvolution is used to remove distortions or (Diagnostic Radiology) See Dual energy CT (DECT) blurring that the input image may have undergone during the imaging process. The response of the imaging system (i.e. g) is known as the point spread function (PSF). Deconvolution is also Decubitus an integral step in the filtered back projection algorithm for tomo- (General) There are a series of terms used to describe the position graphic imaging. of an individual when undertaking different imaging examination. Abbreviations: FT = Fourier transform and PSF = Point Decubitus: Lying on the side. For example, imaging the right spread function. kidney in the left lateral decubitus position (Figure D.5). Related Articles: Convolution integral, Point spread function, Related Article: Patient position Filtered back projection Deep inspiration breath hold (DIBH) technique Decoupling (Radiotherapy) For treatment sites subject to the effects of (Magnetic Resonance) Decoupling refers to the elimination breathing motion, it may be desirable to control the patient’s of the interactions that give rise to spin coupling in systems of breathing in order to freeze this motion and reduce the sever- interacting spins. ity of the breathing. One approach is the use of active breath- The simplest approach to decoupling is intense, continuous ing control (ABC – system supplied by Elekta). Another is to get RF irradiation at the resonance frequency of one of the coupled the patient to hold his breath at full, or close to full, inspiration spins (say A). This has the effect of flipping spins rapidly between during irradiation in deep inspiration breath hold (DIBH). Often orientations (±½ in the case of spin ½ nuclei such as protons), so several breath holds are needed to cover the full treatment dura- that the effect on the other spin species (say X) averages out. tion. DIBH may be particularly useful for tangential irradiation Sometimes, however, decoupling of spins across a broader of the left breast, where it may reduce the volume of heart treated frequency range is required. In such instances, a variety of to high dose, or to exclude the heart from the high dose region broadband decoupling schemes using composite pulse trains completely. of increasing complexity are available: primarily WALTZ and Abbreviations: ABC = Active breathing control, also Active its variants. These sequences rely on the fact that the effects of breathing coordinator and DIBH = Deep inspiration breath hold. coupling within a heteronuclear AX spin system can be eliminated Related Articles: Active breathing control, Gating – by inverting the A spins, and employ composite pulses designed respiratory, Image guided radiotherapy to give good inversion over a wide frequency range. Application of decoupling techniques on clinical MRI equip- Deep learning ment requires special hardware modifications: an additional (General) Deep learning is a specific machine learning technique radiofrequency channel and coil (frequently a surface coil, based on the usage of artificial intelligence. although the body coil may be used). From a safety perspective, Deep learning’s most common applications are related to data the additional power deposited by the decoupling channel must collection and analysis and data pattern production based on big be taken into account when calculating specific absorption rate data sets. Deep learning algorithms have the capability to operate (SAR). without human intervention and produce reliable results with a Further Readings: Freeman, R. 1988. A Handbook of Nuclear quality similar to, or higher than, human-generated results. Magnetic Resonance, Longman, Harlow, UK; Shaka A.J., Keeler Types of common deep learning architectures: J., Freeman R. 1983. Evaluation of a new broadband decoupling • Unsupervised pretrained networks (UPN) sequence WALTZ-16, J. Magn. Reson. 53, 313–340. • Convolutional neural networks (CNN) Deep therapy 240 Deformable image registration (DIR) • Generative adversarial networks (GAN) Horizontal deflection plates • Recurrent neural networks (RNN) Vertical deflection plates Phosphor • Recursive neural networks Electron screen Deep learning techniques with most common application in healthcare provide advanced tools for management and assis- gun Electron tance in the following areas: electronic health records, hospital beam and departmental information systems, diagnostic images analy- Cathode (–1 kV) Plates (~ + 300 V) sis, clinical data analysis, bioinformatics, computer vision, speech Screen potential (~ + 3 kV) recognition, language processing, audio recognition, pharmaceu- tical design. FIGURE D.6 Cathode ray tube. Deep therapy (Radiotherapy) Around the early 1970s the term ‘deep therapy’ oscilloscope) and may still be used to display monochrome or D was still given to treatments using energies of 200–300 kV. This colour images in a video monitor or TV set. was to distinguish it from ‘superficial therapy’ which was the The principle is that a beam of electrons is generated within term given to treatments with energies from around 60 to 140 a vacuum enclosure, usually of glass, formed into a narrow beam kV. However with the coming of high energy beams, this energy and steered and accelerated towards the front phosphor screen range was called

orthovoltage. where it impinges and causes light to be emitted which can be Today’s classifications would be as follows (see IPEMB, 1996): seen from outside through the transparent glass casing. While the relatively large screened monitors and TV sets use the magnetic field from currents in coils attached to the outside Very low energy 8–50 kV 0.035–1.0 mm Al of the vacuum chamber, this does give wide deflection but does Low energy 50–160 kV 1.0–8 mm Al introduce distortion. Medium energy 160–300 kV 0.5–4.0 mm Cu As CRT oscilloscopes are used for measurement of time and Teletherapy/megavoltage Cobalt 60 and 2–25 MV linear accelerators voltage it is important to have the deflection of the beam strictly therapy proportional to the deflecting vertical (input signal) and horizon- tal (time base) signals. This is found to be more practical by using electrostatic deflection and limiting the deflection angle – hence Further Readings: IPEMB. 1996. The IPEMB code of prac- CRT oscilloscopes are quite deep in comparison to the size of tice for the determination of absorbed dose for x-rays below 300 their screens. kV generating potential (0.035 mm Al – 4 mm Cu HVL; 10–300 The electrostatic deflection of the CRT beam is provided by an kV generating potential). Phys. Med. Biol. 41: 2605–2625; orthogonal set of deflection plates placed within the CRT, below Meredith, J. W. and J. B. Massey. 1972. Fundamental Physics and to the sides of the beam as sketched in Figure D.6. of Radiology, John Wright & Sons Ltd, Bristol, London, UK; Each opposing pair of deflection plates is driven by large Wachsmann, F. and G. Drexler. 1976. Graphs and Tables for Use symmetrically varying potentials (∼100 V) which attract/repel in Radiology, Springer-Verlag, New York. the beam as it passes from the cathode ‘gun’ towards the highly positive area of the phosphor screen. Deflection is made sensitive Defective pixel by ensuring the electron beam is deflected early in its travel before (Diagnostic Radiology) Defective pixel is a term used to access it is accelerated to its maximum speed. the status of pixels on LCD screens and/or CCD/CMOS sensors Abbreviation: CRT = Cathode ray tube. in digital cameras. Each pixel that does not perform as expected Related Article: Oscilloscope is considered defective. Defective pixels functional classification: Deformable image registration (DIR) (Radiotherapy) Deformable image registration (DIR) is the 1. Hot pixels – always on process of registering an image data set to a reference image data 2. Dead pixels – always off set by applying an elastic deformation to minimise the difference 3. Stuck pixels – one or more sub-pixels are always on or between the two. This allows comparison or integration of data always off which is obtained from two different measurements. 4. New pixels – newly installed pixels Typically, image registration focusses on rigid shifts (six degrees of freedom) applied to the entire data set such that all The number, type and location of defective pixels for each device/ voxels move or rotate uniformly, maintaining the original voxel- manufacturer are important criteria influencing the image quality to-voxel relationship. In DIR, the voxel-to-voxel relationship and the functional status of the device. changes as not all shifts or rotations are applied to all voxels. Related Article: Bad pixel DIR has become more common in radiotherapy, particularly to monitor the impact of anatomical changes as adaptive radiotherapy becomes more widespread. The deformation field Deflection electrode generated by DIR can also be applied to dose distributions (General) See Deflection plates in cathode ray tubes Quality assurance is required to ensure that DIRs are plausible and that pixels, or voxels, have not been overdistorted. Typical Deflection plates in cathode ray tubes examples of DIR in radiotherapy include: kV CBCT to planning (General) Cathode ray tubes (CRT) form the basis for one form CT deformation and MRI to planning CT deformation. of visual display commonly used in electronic test equipment (the Related Articles: Dose accumulation, Dose warping Degassing of x-ray tube 241 Delta function such as carbon, may reduce this effect, but sometimes apertures are required to compensate for it. A degrader will have a minimum energy loss, due to the minimum overlap needed to ensure the same thickness of material across a beam profile. The maximum energy loss of a degrader is typically restricted by the limitations of beam transport for low energy particles. A cyclotron producing 250 MeV protons, typically has a clinical energy range of around 230–70 MeV. Finally, the degrader is a significant contributor to the total neu- tron production in a cyclotron beam line. This requires the device to be well shielded and away from the patient treatment area. Further Reading: Paganetti H. 2012. Proton Therapy Physics. A rigid registration of two shapes (left). A deformable image registra- CRC Press; Schippers J. M. 2017. Beam-transport systems for par- tion of the same two shapes (right). https://au.mathworks.com/matlab ticle therapy. Proc CAS-CERN Accel Sch Accel Med Appl. 1(CERN- central/fileexchange/58987-non-rigid-registration-between-2d-shapes 2017-004-SP):241-252. doi:10.23730/CYRSP-2017-001.241. D Degassing of x-ray tube Deionised water (Diagnostic Radiology) The high vacuum inside the x-ray tube (Nuclear Medicine) Deionised water refers to water that has assures the undisturbed path of the thermal electrons from the been physically processed to remove mineral ions. Mineral ions cathode filament to the anode target. Internal ionisation of the like sodium and calcium are extracted from the water using ion x-ray tube leads to vacuum reduction and internal discharges (arc- exchange resin beads. Deionised water must not be confused with ing sparks) between the two electrodes or between the tube enve- distilled water where the water has been heated and the steam lope and one electrode (usually cathode). transferred to another clean container where it condenses. During x-ray tube manufacturing, the glass envelope is first vacuumed and then sealed. However, with time the glass, and the Delay relay other parts of the x-ray tube inside the vacuum, emit ions (cold (General) A delay relay, also known as time-delay relay, is a emission). Special measures are taken for reducing this emission relay with a kind of ‘shock absorber’ mechanism attached to – such as polishing and degassing the glass and the metal the armature which prevents immediate, full motion when the electrode assemblies. However, this treatment does not eliminate coil is either energised or de-energised. This addition gives the the problem entirely. The cold emission exists, causing internal relay the property of time-delay actuation. Time-delay relays can ionisation of the x-ray tube volume. This may create artefacts or be constructed to delay armature motion on coil energisation, (rarely) damage the x-ray tube and generator. Due to this reason, de-energisation, or both. new tubes and tubes, which have not been used for a long time (or have been stored for a long period) must be ‘degassed’. The Delta electrons method includes slow warming with several low-power exposures (Radiation Protection) The electrons released in ionising events before regular use. are sufficient to travel a significant distance in the absorber. Delta Related Articles: Warm-up, Arcing of x-ray tube electrons may have energies on the order of 1 keV or greater. They Further Reading: Tabakov, S. 2018. X-ray tube arcing: can be visualised in cloud chambers and similar devices. Manifestation and detection during quality control. J. Med. Phys. Related Article: Delta rays Intern. 6(1):157−161. Hyperlink: www .emerald2 .eu Delta function (General) The delta function has two possible forms. Degrader Kronecker Delta: The first is the discrete form, the Kronecker (Radiotherapy) In particle therapy, a degrader is a device placed delta, δij, which is a single non-zero data value at the time within a cyclotron beamline to reduce the energy of the particle indicated, and zero at other times: beam. Cyclotron energy changes are usually required to occur within ì1 if i = j a fraction of a second and with a step change approximating to a dij = í 5 mm range shift in water. Consequently degraders must move î0 if i ¹ j with a high speed and accuracy, exerting a uniform effect across the beam profile. Often the Kronecker delta is simplified by setting j = 0. This form They are usually constructed from Perspex, graphite or is often called the unit impulse: beryllium and consist of two or more wedges. An even number of wedges and simultaneous movement of the wedges is required to ì1 if i = 0 di = í ensure a uniform effect across the beam. î0 if i ¹ 0 The use of a degrader has several consequences on beam qual- ity. Firstly, the larger the energy loss caused by the degrader, the Dirac Delta Function: Alternatively the continuous form larger the energy spread of the exiting beam. This is caused by is known as the Dirac delta function, and has the following statistical variation of particle path length in the degrader’s mate- properties: rial. An energy selection system (ESS) may be used after the degrader to reduce this energy spread. ì¥ if x = 0 The degrader material also increases the emittance of the d(x) = í beam due to multiple scattering. The use of a low atomic material, î0 if x ¹ 0 Delta rays 242 D ensitometer ¥ Demagnification factor does not affect the fluoroscopic òd(x)dx = 1 image quality, but it influences the image brightness. In fact, the brightness gain of an image intensifier is obtained as the product -¥ of demagnification factor and flux gain. This is not strictly a function but is often treated as such. It can be Related Article: Image intensifier thought of as an infinitely sharp spike of infinite narrowness, with the value 0 everywhere except at x = 0, such that its integral area Demodulation is unit. As with the Kronecker delta, it is often referred to as the (Ultrasound) In general, demodulation refers to the extraction of unit impulse function. an information-carrying signal from a high frequency signal (car- The Dirac delta function has the following characteristic: rier). An example from radio electronics is amplitude modulation, where the amplitude of a high frequency signal is varied accord- ¥ ing to an audio signal. ò f (x)d(x)dx = f (0) In diagnostic ultrasound, there are two principal applications where demodulation is employed. The first is in greyscale imag- D -¥ ing, where the echo from a surface or reflector has the shape of The Dirac comb consists of a pulse train of Dirac deltas, at uni- a pulse that oscillates for a couple of wavelengths. The informa- form intervals, and is an often used sampling function used in tion-carrying signal is actually only that there is a surface, which signal processing. desired to be shown in an image as a thin line. This information is The Fourier transform of the delta function is unity: extracted by finding the envelope of the echodata (often referred to as RF-data, RF for radiofrequency). This can be done by first ¥ rectifying the signal, and then filter the result to obtain a smooth FT éëd(x)ùû (k) = òe2pikxd(x)dx = 1 envelope. The second is in Doppler systems, where the information- -¥ ¥ carrying signal is the Doppler shift, i.e. a frequency modulation FT éd(x x ) (k) e2pikx ë - 0 ùû = ò d(x - x0 )dx = e-2pikx0 of the carrier frequency. Here, the demodulation is performed as a multiplication of the received signal by the transmitted, so -¥ that both sum and difference frequencies are obtained. The sum frequencies are then filtered out, leaving the difference frequency Related Article: Discrete Fourier transform – the Doppler shift. This is also usually done as a quadrature demodulation, where the demodulation frequencies is 90° out Delta rays of phase by the other. This gives the possibility to obtain both (Radiation Protection) When an energetic charged particle or positive and negative Doppler shifts. photon passes through tissue it causes ionisation. Some ionising events may release an electron with sufficient energy to cause Densitometer further ionisations. These secondary electrons may also be (Diagnostic Radiology) A densitometer (Figure D.7) is an instru- referred to as delta rays. ment used to measure the optical density or opacity of a trans- Related Articles: Secondary ionisation, Secondary electrons lucent type film such as used for film/screen radiography. The optical density (OD) is determined by passing a small beam of Demagnification factor light through the film area and measuring the amount of

light (Diagnostic Radiology) The main components of an image passing through. From the ratio of the light value penetrating the intensifier are: input phosphor, photocathode, electron optics and film (I) to the value of the light source without the film in place output phosphor. The incident x-rays are absorbed in the input (I0) (representing a density value of zero) the actual density is cal- phosphor and light photons are emitted which interact with the culated and displayed: photocathode to release photoelectrons. The photoelectrons produced in the photocathode are collected and focused by a series of electrostatic electrodes on the output phosphor, where thousands of light photons are emitted for each photoelectron collected. The number of photoelectrons within the image intensifier does not change but, as the area of the output screen is smaller than the area of the photocathode, the number of electrons/mm2 will increase. The demagnification factor, also called minification gain, can be expressed as Demagnification factor (Diameter of input phosphor)2 = (Diameter of output phosphor)2 FIGURE D.7 A densitometer being used to measure the optical density Considering that the diameters of input phosphors range from 15 of an area in a film. The film is illuminated by the light below it and the to 40 cm and a typical diameter of an output phosphor is 2.5 cm, transmitted light is being measured by a sensor that is in contact with the the demagnification factor is roughly between 36 and 256. top surface. Densitometry 243 Depth dose distribution æ I ö Dental radiography OD = -logç ÷ (Diagnostic Radiology) Radiography is used by the dental profes- è I0 ø sion to evaluate teeth and associated conditions. The small receptors are designed to fit within the mouth. Densitometry Dental x-ray equipment usually uses low power x-ray tubes (Diagnostic Radiology) Densitometry is a method of extracting and generators. The simplest ones use fixed kV and mA and just information from the image, based on assessment of optical control the length of the exposure with a timer. The most complex densities (analogue image) or pixel values (digital image). ones – (ortho pan tomographic equipment, OPG) make a special Optical densitometry assesses the opacity of the exposed film radiograph of the whole jaws. with the use of an optical densitometer (the film is placed between calibrated sources of light and a light detector such as photodi- Dephase ode). The method is often used to estimate the dose which has (Magnetic Resonance) Dephasing is the transverse part of the exposed the film. relaxation process in which the spins lose their coherence and Assessment of pixel values of various digital imaging modali- return to their original state, and become more and more invisible ties (e.g. CT scanner, digital radiography, etc.) provides important to the receive coil. D information about the imaged anatomical structures by assessing Dephasing is the cumulative loss of phase which always occurs the change of their radiation absorption due to some pathological as soon as the RF-pulse is turned off. It is the result of spin-spin processes. interactions following the small differences in precessional fre- Bone densitometry is a routinely used method of densitometry quencies always present in the volume of interest. to assess osteoporosis. As it causes a signal loss it is often a problem solved by shorten- Densitometry is an important part of the methods used in ing the echo time or by using a SE sequence. It may also improve quantitative imaging. contrast however, e.g. in DSC-imaging or diffusion-imaging. The Related Articles: Densitometer, Radiographic film dosimetry, dephasing of spins can be increased by use of paramagnetic con- Grey values, Pixel values, Bone densitometry, Optical density trast agents. Inhomogeneities in the magnetic field can cause unwanted dephasing which can be detracted with shimming gradients. Density correction Related Articles: Relaxation, RF-pulse (Radiotherapy) The electron energy loss is theoretically the same for the same mass of material while in reality the stopping power Depletion layer is smaller for a condensed medium due to the polarisation of (Radiation Protection) In semiconductor physics, the depletion atoms. The polarisation effect influences the soft collision that is layer is an insulating region within a conductive, doped semicon- an energy transfer between the charged particle and the relative ductor material where the mobile charge carriers have been dif- distant atoms. In gases the atoms are spaced widely and therefore fused away or have been forced away by an electric field. The term they undergo interactions independently of each other. On the con- depletion refers to the fact that the region near the pn junction is trary in a condensed medium the density is increased by a factor of depleted of charge carriers (electrons and holes) due to diffusion about 103–104 compared to a shortened distance between atoms of across the junction. The only elements left in the depletion region about 1/10 of the distance between atoms in a gas. The mass col- are ionised donor or acceptor impurities. lision stopping power is therefore decreased in condensed media Related Articles: Diode, Diode detector, Semiconductor because of dipole distortion of the atoms near the track of the pass- detector ing particle that weakens the coulomb force experienced by more distant atoms. Sternheimer et al. (1982) determined a term δ to be subtracted from the mass collision stopping power for electrons Deposition of dose and positrons to correct for the polarisation effect. The term δ is (Radiotherapy) Ionising radiation deposits dose, i.e. absorbed given by energy per unit of mass by ionisations and excitations. The energy deposited by ionising radiation is localised in few atoms. Charged particles produce these effects through direct Coulomb-force d å é (ni + l2 ) ù = filn 2 ê 2 - l (1- b2 ë n ú ) interactions with orbital electrons of the matter being traversed while x-ray and gamma interact by processes in which a second- i û ary charged particle, normally an electron, is generated in the where absorber and the secondary particle in turn deposits most of the β = v/c with v being the velocity of the passing particle energy. Uncharged particles as neutrons deposit energy through f secondary charged particles by various process as a function of i is the oscillator strength of the ith transition, whose fre- quency is ν the neutron energy. i l is a dimensionless frequency pertaining to I which is the Related Article: Stopping power solution of the equation Depth dose curve 1 (Radiotherapy) See Percentage depth dose f 2 -1 = b å i n2 i + l2 i Depth dose distribution (Radiotherapy) The dose distribution is a representation of the Further Reading: Sternheimer, R. M., S. M. Seltzer, and M. variation of dose with position in any region of an irradiated J. Berger. 1982. Density effect for the ionization loss of charged object. The dose distribution is determined from physical mea- particles in various substances. Phys. Rev. 26: 6067. surements in a phantom using an ionisation chamber or other Depth gain compensation 244 Depth gain compensation 100 Electrons 90 X-rays 80 70 2 MeV 60 25 MeV 50 40 0.2 MeV 30 20 1.5 MeV 40 MeV D 10 16.4 MeV 0 0 2 4 6 8 10 12 14 16 18 20 22 24 Depth (cm) FIGURE D.9 Dose distribution along the beam axis (percentage depth dose) for selected energy beams. FIGURE D.8 Colour code isodose lines. radiation detector that can be positioned at various locations in the radiation beams of interest. The resulting data is presented in tabular form for calculating the linac monitor unit or time settings to deliver the prescribed dose to a representative point in the tar- get volume and for calculating the relative dose to other points of interest. The dose distribution in the structures of a patient for a given incoming radiation field is also calculated by treatment planning systems (TPS) by dose deposition algorithm to provide the radiotherapist with a physical description of the treatment. The geometrical distribution of energy deposited by the radiation can be displayed in three dimensions by means of a set of graphs, each of which is a two-dimensional representation of the distribution in a plane. Lines of equal dose (isodose lines) are drawn as a function of two coordinates. Their standard display mode is colour-coded FIGURE D.10 Colour wash display of dose distribution. isodose lines overlaid on the greyscale image (Figure D.8). Often dose distribution planar graphs are used for treatment planning in place of a complete set which of course would show the distribution in all three dimensions. A planar graph may be contracted further to a graphic representation of relative dose vs. depth in water measured from the surface where the beam enters. If this graph describes the dose in the centre of the field it is called a central-axis depth dose distribution (Figure D.9). Several kinds of further techniques are available to display both dose and anatomical information using 3D solid surfaces as well as greyscale images. A ‘colour wash’ display technique assigns colour values corresponding to the dose to each pixel painting the whole dose distribution in transparent bands of colour on the greyscale image. By changing the dose to colour assignment in the display look-up table (LUT) the colour wash can be used to interactively scan a band of colour (dose) over the image interactively (Figure D.10). 3D dose distributions can also be displayed and compared to solid surface display of the target volume and critical normal structures (Figure D.11). Depth gain compensation (Ultrasound) Depth, or time, gain compensation (DGC/TGC) is the equalisation of echoes from different depths to compensate FIGURE D.11 3D dose distribution. Percentage depth dose Depth ionisation curve 245 Depth of interaction for increased attenuation with depth so as to display a uniform level of echoes from similar tissues at different depths. Depth gain compensation is sometimes referred to as time gain compen- sation/swept gain. The transmitted ultrasound pulse and reflected echoes are attenuated by tissue due to absorption, reflection, scattering and beam dispersion and aberration. The attenuation is dependent on the acoustic properties of the tissue and its geometry. Ultrasound scanners are designed and programmed with automatic compensation for this attenuation with increased gain from deeper echoes. This assumes a uniform predictable attenu- ation based on an average for soft tissue in the body, typically around 0.5–0.7 dB/cm/MHz. This may not be appropriate for a particular investigation. Examples where attenuation is low occur when scanning through fluid, e.g. bladder, amniotic fluid, cysts D or ascites. Here deep structures appear too bright if the gain is uncorrected (Figure D.12), a phenomenon known as post cystic FIGURE D.13 DGC – 2. With the DGC modified to reduce gain at depth enhancement. Conversely, uncorrected scanning through more there is a more uniform image of the bladder. highly attenuating tissue leads to dark images at depth because of increased attenuation. Most ultrasound scanners have a depth/time gain control to enable the operator to modify the gain compensation. Typically these are a series of sliders (Figure D.14) corresponding to dif- ferent depths within the image allowing gain to be increased or decreased at specific levels (Figure D.13). Since deeper echoes correspond to echoes with a longer transit to receive time, the terms depth gain control and time gain control are both used to describe this process. A diagrammatic representation of the change in gain may be shown on the image as a line down side of the image showing relative gain settings at different depths (Figures D.12 and D.13). The DGC and overall gain both affect the gain of the image. A common mistake made by those starting ultrasound imaging is to have the gain high, with excessive bright echoes obscuring contrast detail. This is a particular problem if scanning is done in rooms with high ambient light. Related Articles: Attenuation, Absorption FIGURE D.14 DGC controls. DGC slider controls (ringed) on an ultra- Depth ionisation curve sound scanner keyboard. (Radiotherapy) If a Farmer chamber is placed at different depths in water or phantom material and irradiated the electrometer reading at different depths can be plotted. This plot represents the depth ionisation curve since it is a measure of the amount of ionisation that the chamber recorded at different depths. Related Article: Percentage depth dose Depth of interaction (Nuclear Medicine) The depth of

interaction (DOI) is an effect in PET imaging where the spatial resolution is degraded for off- centre radioactive sources. The photons from an off-centre source will in most cases enter the detector at an oblique angle. A sub- stantial thickness of scintillator material is needed to stop pho- tons. A photon with an oblique angle of incidence can penetrate one or several scintillators and be registered in an interaction in an adjacent detector (see Figure D.15). As a result an accurate spatial localisation of the event is impossible and the spatial reso- lution decreases. The magnitude of the problem is determined by the width and length of each detector element and the diameter of the scanner. The effect also increases with the distance from the centre of the FIGURE D.12 DGC – 1. With little change to the DGC (line R) the scanner. weak echoes and artefacts deep in the bladder are amplified and there are Related Articles: Annihilation, Annihilation coincidence high intensity echoes in the bladder wall deep to the fluid. detection, Centre of rotation, PET Depth of interaction effect 246 Depth of penetration Line of response Detector ring Point of annihilation D FIGURE D.15 Depth of interaction effect. A photon that enters a scintillator at an oblique angle can penetrate one or more crystals and interact in an adjacent scintillator instead. As a result the line of response is misplaced. 100 TABLE D.2 15 MV photons 80 A Table of Build Up Depths for Common Photon 10 MV photons Sources 60 6 MV photons Photon Source Build Up Depth (cm) Caesium-137 rays (0.66 MV) 0.12 40 Cobalt-60 rays (mean energy 1.25 MV) 0.5 4 MV x-rays 1.0 20 Cobalt 60 6 MV x-rays 1.5 HVT 3 mm Al HVT 2 mm Cu 8 MV x-rays 2.0 0 10 MV x-rays 2.5 0 5 10 15 20 25 30 Depth (cm) 16 MV x-rays 3.0 dmax for 15 MV photons ~3 cm FIGURE D.16 Central axis dose build up and falloff for different kV The range or depth of penetration of a proton beam is typically and MV photon beam energies. given by d90 or d80, which are the depth in water at the 90% or 80% dose level respectively. The range of a proton follows a power law and can be calcu- Further Reading: Cherry, S. R., J. A. Sorenson and M. E. lated using the Bragg-Kleemann rule, which is given by the fol- Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, lowing expression: (Newhauser and Zhang, 2015) Philadelphia, PA, pp. 334–336. R(E) = aE p Depth of interaction effect (Nuclear Medicine) See Depth of interaction where α is a constant that depends on the material, E is the ini- tial energy of the proton beam and p is a constant that accounts Depth of maximum dose for the dependence of energy of the proton or velocity. α and p (Radiotherapy) The depth at which dmax occurs is called the depth can be derived by fitting to ranges or stopping power data from of maximum dose and is energy dependent. For photons higher measurements. beam energies produce more forward scattering of electron and In a clinical scenario, there are many factors that contrib- longer scatter path lengths. The higher the beam energy, the ute to the uncertainty in proton range, such as the conversion of deeper the in tissue or phantom that dmax occurs. See Figure D.16 Hounsfield units in the planning CT to relative stopping power, and Table D.2. CT image reconstruction artefacts caused by hip prostheses or Related Articles: Percentage depth dose, Build up, Build up dental fillings, motion artefacts, partial volume effects, variation dose in human tissue, etc. It is therefore advisable to avoid placing the Further Reading: Podgorsak, E. B. 2003. Review of Radiation distal edge of the Bragg peak within critical organs at risk such Oncology Physics: A Handbook for Teachers and Students, as the spinal cord or the brain stem during treatment planning International Atomic Energy Agency, Vienna, Austria. (eLfH, accessed April 2019). Related Article: Pristine Bragg peak Depth of penetration Further Readings: eLfH, e-Learning for Healthcare, (Radiotherapy) The depth of penetration or range of a charged e-Protons, https://portal .e -lfh .org .uk/, accessed April 2019; particle Bragg peak depends on the particle energy. The range is International Commission on Radiation Units and Measurements usually expressed in terms of water-equivalent path length, since (1993). Stopping Powers for Protons and Alpha Particles. ICRU water closely mimics the properties of human tissues. Report 49. Bethesda, MD; Khan, F. M. and J. P. Gibbons. 2014. Percentage depth dose Dermis 247 Detection efficiency Khan’s the Physics of Radiation Therapy, 5th edn., Wolters Kluwer Health; Newhauser, W. D. and R. Zhang. 2015. The physics of proton therapy. Phys. Med. Biol. 60: R155−R209; NIST‐Pstar, http: / /phy sics. nist. gov /P hysRe fData /Star /Text /PST A R .htm l, accessed April 2019. Dermis (Non-Ionising Radiation) The dermis is the thickest layer of the skin, immediately below the epidermis. The dermis contains blood vessels, hair follicles and most of the living structures of the skin. Related Articles: AORD, Basal cell carcinoma, Melanoma, UV light hazard, UV dosimetry Further Readings: Blumenberg, M. 2018. Human Skin Cancers: Pathways, Mechanisms, Targets and Treatments. D London; Montagna, W. and P. F. Parakkal. 2012. The Structure and Function of Skin. 3rd edn., Elsevier, New York and London. FIGURE D.17 Interference from two point sources emitting continuously. Design dose per week (Diagnostic Radiology) A barrier between the source and an indi- vidual to be protected must attenuate the radiation level to the Visibility of detail design effective dose limit or dose constraint. This means that the Low blur maximum dose attenuated through the barrier should not exceed this limit or constraint (Benjamin R. Archer PhD. Department of Radiology Balor College of Medicine, 2019). Essentially, it is the maximum dose limit set per week for a Medium blur person to receive after being protected by a barrier from a source of radiation. A design limit of 0.3 mSv per year is used for public areas in the UK and corresponds to a design dose per week of 6μGy High blur (International Atomic Energy Agency, 2006) Barrier calculations are based on these limits. Depending on local regulations, other limits may be applied and different barrier requirements will be calculated. FIGURE D.18 Increased blurring reduces the visibility of small objects Related Articles: Controlled area, Radiation shielding and detail. (Courtesy of Sprawls Foundation, www .sprawls .org) Further Readings: Archer, B. R. PhD. Department of Radiology Balor College of Medicine, H., T. 2019. IRPA-10 Course EO-6 Shielding of diagnostic X-ray facilities for cost- Detail resolution effective and beneficial use and protection. [Online] Available (Diagnostic Radiology) Detail resolution, also known as visibility at: www .irpa .net /irpa10 /pdf /E06 .pdf [Accessed 23 July 2019]; of detail, is the ability to form a visible image of small objects or International Atomic Energy Agency, 2006. Radiation Protection structures, i.e. detail within an image. The major factor that limits in the Design of Radiotherapy Facilities, Safety Report Series 47, detail resolution and visibility of detail is the blurring that occurs Vienna: IAEA. during the imaging process as illustrated in Figure D.18. Hyperlinks: IRPA: www .irpa .net /irpa10 /pdf /E06 .pdf; IAEA: Blurring has the same effect on the visibility of lines in a reso- www -p ub .ia ea .or g /MTC D /pub licat ions/ PDF /P u b122 3 _web .pdf lution test pattern (bar phantom) as illustrated in Figure D.19. Blurring reduces the visibility of small objects and anatomi- cal detail. This is the clinical significance of blurring and why it DESS (dual echo steady state) must be taken into consideration for each type of imaging proce- (Magnetic Resonance) See Dual echo steady state (DESS) dure. Blurring also reduces spatial resolution when imaging test devices. This is often used to measure and evaluate the effects of Destructive interference blurring in imaging procedures. (Ultrasound) Two waves that travel together can, dependent on The detail resolution is measured through the limiting spatial their respective phase, be observed as a wave that is the sum of resolution in line pairs per mm (lp/mm). In Figure D.19 the system the two waves’ individual amplitudes (constructive interference), with high blur has limiting spatial resolution of 4.5 lp/mm. or add up to no apparent wave motion at all if the amplitudes are Thus the smallest detail to be seen = 1/[2 * (lp/mm)]. equal and the phases are opposite (destructive interference). In the For example, if the limiting spatial resolution of one system is figure two point sources emit continuous waves, and in certain 4 lp/mm, the smallest detail to be seen with it is 1/[2 * 4] = 0.125 directions the waves appear to be in phase, whereas in others, to mm. be out of phase. As can be deduced from the figure, this depends Related Articles: Line pair, Spatial resolution on the difference in distance from the observation point to the respective sources. Distance differences that correspond to an Detection efficiency integer number of wave lengths plus one half wavelength result in (Radiation Protection) The efficiency of a radiation detector can destructive interference (Figure D.17). be seen from various angles. Detective quantum efficiency (DQE) 248 D etective quantum efficiency (DQE) 100 with energy of ~18 keV) with an absolute efficiency of about 100%. The intrinsic efficiency of a detector, used for a specific radia- tion, is related to detector absorption and also to the effective- ness with which the detector converts this radiation energy into another measurable energy. The ability of a detector to absorb the incident radiation pho- tons is called quantum efficiency (EQ). EQ depends on the linear 1 2 3 4 5 6 7 8 absorption coefficient of the detector material (µ) and its thick- Spatial frequency (LP/MM) ness (x). FIGURE D.19 Increased blurring reduces the ability to resolve the lines EQ=1−e−µx in a test pattern. (Courtesy of Sprawls Foundation, www .sprawls .org) D Obviously, the more absorbent the detector, the higher its EQ. For example, a very absorbent semiconductor detector will have The absolute efficiency εabs is equal to the ratio of the num- an EQ of almost 1, while a gas detector will have an EQ of the order ber of events registered by the counting system to the number of of 0.01. A scintillation detector will have an EQ depending on the events (e.g. gamma radiation quanta, beta particles etc.) emitted specific absorption of photons by the phosphor (e.g. NaI:Tl will by the source: have an EQ of the order of 0.8). The parameter related to conversion of radiation energy into Number of events registered e another measurable energy is called conversion efficiency (EC), abs = Number of eventsemitted by thesource sometimes also called extrinsic efficiency. A gas detector will have an EC of the order of 1, as almost all ion pairs created by the The absolute efficiency depends on the geometry of measurement ionising radiation will be detected by the electrometer. A similar and on the detector response to the radiation. example of EC is with the semiconductor detector (all electrons The efficiency related to the geometry of measurement (geo- and holes created by the radiation in the depleted zone create a metric efficiency) is related to the interception between the radia- measurable current between the two electrodes). However, the tion beam and the detector area/volume. For example, when a conversion of radiation into light in the scintillation detector narrow radiation beam passes through a substance, the exit radia- depends on the specific phosphor (e.g. NaI:Tl will have an EC of tion measurement will include mainly the primary radiation. On the order of 0.5 – only 50% of the radiation photons will create the contrary, when a broad radiation beam passes through a sub- light photons). stance, the exit radiation measurement will include the primary The total efficiency of a radiation detector ET for specific radi- radiation plus a significant percentage of secondary (scattered) ation is presented as the product of EQ and EC: radiation. The geometrical efficiency of a point source of radiation (iso- ET = EQ × EC tropic) will be related to the detector area and its distance to the source. For example, a detector with spherical surface (4π steradi-

Having the examples of EQ and EC from above, the total effi- ans) will have a 100% geometric efficiency. ciency of a gas detector will be 0.01 (1%), the same for a scintil- The geometrical efficiency εg is equal to the ratio of the num- lation detector will be about 0.4 (40%), while for a semiconductor ber of particles (the term particle can include photons in the case detector ET will be almost 1 (100%). of electromagnetic radiation), striking the detector to the total Related Articles: Geiger-Muller counters, Liquid scintillation number of particles emitted by the source: (LS) counting, Broad beam geometry Further Readings: Bethge, K. et al. 2004. Medical Number of particlesstriking thedetector Applications of Nuclear Physics. Springer-Verlag, Berlin, eg = Number of particlesemitted by thesource Heidelberg, p. 62; Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd edn., John Wiley & Sons, Inc., New York, The intrinsic efficiency of a detector, εint, is defined as the ratio Chichester, Weinheim, Brisbane, Toronto, Singapore, p.116; Saha, of the number of registered events to the number of events (e.g. G. P. 2001. Physics and Radiobiology of Nuclear Medicine, 2nd gamma radiation quanta, beta particles etc.) incident on detector: edn., Springer Verlag, New York, Berlin, Heidelberg, pp. 90−91. Number of registered events Detective quantum efficiency (DQE) eint = Number of events incident ondetector (Diagnostic Radiology) The detective quantum efficiency (DQE) of an imaging system is a measure of both the efficiency with which the imaging system detects photons and of the noise cre- The absolute detection efficiency can be expressed as the product ated in the image. It is used in diagnostic radiology as a standard of the intrinsic efficiency and the geometrical one: image metric to quantify the quality of an imaging system. The DQE is always dimensionless, and can have a value no greater eabs = eint ´ eg than unity. This can be used to infer, e.g. the minimum x-ray dose to a patient which will provide images of satisfactory quality in an Usually the detection efficiencies are expressed in percents. x-ray radiograph. The DQE is defined (Equation D.1) as The Geiger-Muller counter intrinsic efficiency for beta radia- tion is 100% and a few percent for gamma radiation. Liquid scin- NEQ DQE = (D.1) tillation counting system can be used for tritium (beta particles N Contrast Med ium bl ur High blu r Detective quantum efficiency (DQE) 249 Detector where NEQ is the noise equivalent quanta over the image area, Related Articles: Modulation transfer function (MTF), Noise being a measure of image quality, and N is the number of inci- power spectrum (NPS), NEQ (noise equivalent quanta), Signal to dent photons over this area. The NEQ of the image is derived noise ratio (SNR) from the square of the signal to noise ratio (SNR) measured in Further Readings: Beutel, J., H. L. Kundel and R. L. Van the image. It describes the number of quanta which would result Metter, eds. 2000. Handbook of Medical Imaging. Physics and in the measured SNR if the measurement system was perfect Psychophysics, Vol. 1, SPIE, Bellingham, WA; International (i.e. only Poisson noise was present). For example, an imaging Electrotechnical Commission (IEC). 2003. Medical Electrical device may measure 100 photons over a unit area (N = 100). If Equipment – Characteristics of Digital X-ray Imaging Devices. this device was ideal, then the noise in this signal is defined Part 1: Determination of the Detective Quantum Efficiency, IEC as 100 = 10 and the SNR is also equal to 10. If we are now 62220-1:2003, IEC, Geneva, Switzerland. to consider the same system but with a realistic detection effi- Hyperlink: www .i magin gecon omics .com/ issue s /art icles /MI ciency of less than 1 (or one which introduces extra noise to the _2 001 -0 5 _10. asp image), then the actual SNR which is measured could be say 5. The NEQ derived from this SNR would be 52 = 25 and the DQE Detective quantum efficiency (DQE) would therefore be 0.25. This can be interpreted as the detector (Nuclear Medicine) The DQE is defined as the relative number of D only being able to utilise 25% of the incident photons to create quanta of incident radiation detected by a detector. DQE is a mea- the image. sure of the information conversion quality. Later DQE has been This discussion considers the image metrics DQE and NEQ generalised for all detecting devices and the new definition is to have no dependence on spatial frequency. For a real imaging detector however this is not the case and the aforementioned 2 æ SNR ö quantities will vary depending on the spatial frequency compo- DQE = out ç nent in the image. The terms in Equation D.1 should therefore è SNR ÷ in ø be replaced by their frequency-dependent counterparts DQE( f) and NEQ( f). Furthermore NEQ( f) must be described using the where SNRout and SNRin are the SNRs of the detector out and in modulation transfer function MTF( f) and the noise power spec- signals, respectively. The DQE can be considered as the product trum (NPS( f)) (for further details see the article Noise equiva- of all the signal or data conversion steps in an imaging system. lent quanta (NEQ)). The full form of the frequency-dependent For example, in a conventional scintillation camera the DQE DQE( f) can be written (Equation D.2) as includes the photocathode quantum efficiency. Further Reading: Morgun, O. N., K. E. Nemchenko and Y. V. NEQ( Rogov. 2003. Detective quantum efficiency as a quality parameter f ) qMTF(f )2 DQE(f ) = = (D.2) of imaging equipment. Biomed. Eng. 37(5):258–261. q NPS(f ) Related Articles: Quantum efficiency, Conversion efficiency of photocathodes where q is the number of incident photons per unit area (the aver- age uniform input). This has replaced the term N in Equation D.1 due to NPS being a measure of the noise power per unit area. Detector MTF( f) and NPS( f) are both linearised functions with respect to (Diagnostic Radiology) A general term used either to describe input intensity. an instrument used to measure/monitor radiation (e.g. germanium The interpretation of the frequency-dependent detector quan- detector), or to observe radiation (e.g. image detector). The latter tum efficiency (DQE( f)) is very similar to that for standard also measures the radiation, but transforms it into a visual image. DQE. DQE( f) describes the efficiency of the detector but at Image detectors are also known as image receptors. particular spatial frequency component. The DQE is sometimes In x-ray diagnostic radiology incident radiation is recorded expressed by the squared output SNR and the squared input SNR and used to create a radiographic image of a patient’s anatomy. as (Equation D.3) The major types of image detectors used in x-ray diagnostic radi- ology and their classification are shown in Figure D.20. Traditionally, analogue film/screen systems were used for SNR2 DQE = out ( . ) medical imaging. In recent years, analogue detectors are now SNR2 D 3 in being replaced by modern digital ones. Digital detectors include computed radiography (CR) which uses an imaging phosphor but this form must be used with caution, e.g. it is only correct, if plate that fits into the original film/screen bucky and direct digital SNRin corresponds to the ideal photon-counting detector rather radiography (DDR) which usually takes the form of an integrated than any other type of detector such as energy-integrating detec- flat panel detector. Within the DDR class of detectors the image is tor. In 2003, the International Electrotechnical Commission formed by either indirect or direct x-ray conversion, which is illus- (IEC) published an international standard for the measurement of trated in Figure D.21. In indirect conversion the incident x-rays the DQE. The methodology states that the measurement must be photons are converted to visible light photons before detection, under certain standardised conditions for the x-ray energy spec- an example of which is the silicon diode detector. In direct con- trum and the imaging system geometry. Under such conditions version systems the incident x-ray photons are measured directly the MTF can then be determined through the edge response tech- without the need for conversion, an example of this technology is nique and the NPS is measured through Fourier analysis of a flat the amorphous selenium flat panel detectors. field image (see Modulations transfer function (MTF) and Noise The classification of detectors used in diagnostic radiology power spectrum (NPS) articles for further details). The incident can sometimes be misleading. As CR technologies were devel- photon signal q is determined through measurements of the inci- oped before DDR detectors some sources use the term DDR when dent air kerma at the detector face using an appropriate radiation referring to CR and vice versa. Also, the term direct detectors is meter (ionisation chamber). used to refer to all direct digital detectors and just direct x-ray Detector 250 Detector X-ray detector technologies Analogue Digital Computed radiography Direct digital (CR) radiography (DDR) Non-integrated detectors Integrated detectors Indirect x-ray Direct x-ray D conversion conversion Film/screen Photostimulable phosphor Phosphor + CCD Phosphor + a-Si + Photoconductor+ TFT TFT (Flat panel detector) (Flat panel detector) FIGURE D.20 The classification of x-ray detector modalities. Computed Charged-coupled Flat panel: Flat panel: Film-screen radiography devices (CCD) indirect-digital direct-digital Image formation stages PSL: Scintillator Photostimulable Scintillator Scintillator X-ray luminescent interaction phosphor Visible light Visible light e-trapped Visible light X-ray photo- conductor Measurement Laser stimulated (a-selenium) or Photographic luminescence Photodiode conversion to film (a-silicon) electrical CCD charge Visible light Light guide Readout Film processing TFT active matrix TFT active matrix array array PM tube Obtaining digital image ADC ADC Digital Digital Digital Analogue to Analogue to signal signal signal digital digital conversion conversion FIGURE D.21 Image formation steps used in x-ray detectors. Detector absolute efficiency 251 Detector intrinsic efficiency conversion detectors. Direct digital detectors may also be called Detector extrinsic efficiency integrated detectors. Likewise, CR may be called non-integrated (Radiation Protection) See Detector conversion efficiency. detectors. Detector fill factor Detector absolute efficiency (Diagnostic Radiology) The detector fill factor, fill-in factor or (Radiation Protection) The absolute efficiency εabs is equal to the fill factor is the ratio of the active area of a detector (area where ratio of number of events registered by the counting system to signal can be detected) over the total detector area (active area + the number of events (e.g. gamma radiation quanta, beta particles TFT switch) – see diagram in the article Flat panel detector. etc.) emitted by the source: Active detector area Number of events registered Fill factor = eabs = Total detector area Number of events emitted by the source It is often used in diagnostic radiology to describe flat panel x-ray See Detection efficiency. imaging systems as the detector fill factor is usually less than Related Article: Detection efficiency D unity as a proportion of the active surface is covered with thin Further Readings: Dowsett, D., P. Kenny, and R. E. Johnston. film transistors and other circuitry. A low fill factor can require 2006. The Physics of Diagnostic Imaging, 2nd edn., CRC Press; a higher patient dose to produce the same signal strength as a Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd similar detector with a higher fill factor as any incident radia- edn., John Wiley & Sons, Inc., New York, Chichester, Weinheim, tion falling upon an inactive component does not contribute to the Brisbane, Toronto, Singapore. final image signal. This can also be described as the pixel fill factor which is the Detector air kerma ratio of the active area of one pixel over the total pixel area. (Radiation Protection) The air kerma incident upon an imaging Related Article: Flat panel detector detector. By relating the measured detector air kerma at stan- dardised exposures to pre-processing pixel values, a conversion Detector geometric efficiency factor can be determined to provide an indicator of the detector (Radiation Protection) The geometrical efficiency of a point air kerma for other exposures. This indicated detector air kerma source of radiation (isotropic) will be related to the detector area can then be compared with target detector air kerma values in and its distance to the source. For example, a detector with spheri- order to monitor the adequacy of exposures. cal surface (4π steradians) will have a 100% geometric efficiency. Related Articles: Kerma,

Air kerma The geometrical efficiency εg (acceptance) is equal to the ratio Further Readings: International Electrotechnical of the number of particles (the term particle can include photons Commission. Medical electrical equipment – Exposure index in the case of electromagnetic radiation), striking the detector to of digital X-ray imaging systems – Part 1: Definitions and the total number of particles emitted by the source: requirements for general radiography. International Standard. 2008:62494-1; Shepard, S. J., J. Wang, M. Flynn, E. Gingold, L. Number of particles striking the detector Goldman, K. Krugh, D. L. Leong, E. Mah, K. Ogden, D. Peck, eg = Number of particles emitted by the source and E. Samei. 2009. An exposure indicator for digital radiogra- phy: AAPM Task Group 116 (executive summary). Med. Phys. Jul 1, 36(7):2898–2914. Related Article: Detection efficiency Further Readings: Dowsett, D., P. Kenny, and R. E. Johnston. Detector conversion efficiency 2006. The Physics of Diagnostic Imaging, 2nd edn., CRC Press; (Radiation Protection) The parameter related to conversion of Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd radiation energy into another measurable form of energy is called edn., John Wiley & Sons, Inc., New York, Chichester, Weinheim, ‘conversion efficiency’ (EC), sometimes also called ‘extrinsic Brisbane, Toronto, Singapore. efficiency’. A gas detector will have EC of the order of 1, as almost all Detector intrinsic efficiency ion pairs created by the ionising radiation will be detected by the (Radiation Protection) The intrinsic efficiency of a detector, εint, is electrometer. defined as the ratio of the number of registered events to the num- Another example of EC of the order of 1 can be seen with the ber of events (e.g. gamma radiation quanta, beta particles etc.) semiconductor detector. Here all electrons and holes created by incident on the detector: the radiation in the depleted zone will create a measurable current between the two electrodes. Number of registered events e However, in the case of a Scintillation detector, the conver- int = Number of events incident on the detector sion of radiation into light depends on the specific phosphor (e.g. NaI:Tl will have EC of the order of 0.5 – only 50% of the radiation The intrinsic efficiency of a detector, used for a specific radia- photons will create light photons). tion, is related to detector absorption and also to the effective- Related Article: Detection efficiency ness with which the detector converts this radiation energy into Further Readings: Dowsett, D., P. Kenny, and R. E. Johnston. another measurable form of energy. 2006. The Physics of Diagnostic Imaging, 2nd edn., CRC Press; Related Article: Detection efficiency Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd Further Readings: Dowsett, D., P. Kenny, and R. E. Johnston. edn., John Wiley & Sons, Inc., New York, Chichester, Weinheim, 2006. The Physics of Diagnostic Imaging, 2nd edn., CRC Press; Brisbane, Toronto, Singapore. Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd Detector PET 252 Detector scatter event edn., John Wiley & Sons, Inc., New York, Chichester, Weinheim, detector segments. Another modification is to use two differ- Brisbane, Toronto, Singapore. ent types of scintillator materials arranged in two separate lay- ers (known as a phoswich) with one upper and one lower. Each Detector PET layer has a different decay time, so by analysing the decay time (Nuclear Medicine) In an early stage of PET imaging NaI(Tl) in the pulse the event can be localised to either the upper or was used as a detector. NaI(Tl) has some disadvantages, namely lower level. This method can reduce the DOI effects by a factor a low density and low atomic number of the scintillator making of two. These detectors, often a combination of LSO and GSO, it less efficient in stopping annihilation photons (511 keV). Today, are hard to manufacture and the general detector performance is because of this reason, most dedicated PET scanners use denser slightly degraded compared to a pure LSO detector of the same high-Z detector materials. These detectors are often arranged in dimensions. rings around the patient. By using a ring shaped detector it is pos- Abbreviations: BGO = Bismuth germinate, GSO = Gadolin sible to simultaneously collect all projections. ium oxyorthosilicate and LSO = Lutetium oxyorthosilicate:Ce. Block Detector: The block detector was designed by Casey Related Articles: PET, Beta decay Further Reading: Cherry, S. R., J. A. Sorenson and M. E. D and Nutt in the mid 1980s. The construction of the block detec- tor allows smaller detectors (higher resolution) and a reduction Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, of photomultiplier tubes used (less expensive). A scintillator Philadelphia, PA, pp. 342–346. segment or block (BGO or LSO) is cut into an array of smaller elements. Each segment is read out using four PM-tubes as illus- Detector quantum efficiency trated in Figure D.22. (Radiation Protection) The ability of a detector to absorb the An opaque reflective material is placed in between the cuts incident radiation photons is called quantum efficiency (EQ). EQ and is used to avoid scintillation light escaping from one detec- depends on the linear absorption coefficient of the detector mate- tor element to another. The depth of the cuts varies depending rial (µ) and its thickness (x). on the position of the element. The spatial position is deter- EQ = 1 − e−µx mined by the signal in the four PM tubes assigned to the detec- Obviously, the more absorbent the detector, the higher its EQ. tor segment: For example, a very absorbent semiconductor detector will have an EQ of almost 1, while a gas detector will have an EQ of the (PMT PMT PMT PMT = A + B ) - ( C + D ) order of 0.01. X PMT A scintillation detector will have an EQ depending on the A + PMTB + PMTC + PMT D (D.4) ( specific absorption of photons by the phosphor (e.g. NaI:Tl will PMTA + PMTC ) - (PMTB + PMTD ) Y = have EQ of the order of 0.8). PMTA + PMTB + PMTC + PMTD Related Article: Detection efficiency Further Readings: Dowsett, D., P. Kenny, and R. E. Johnston. PMTA, PMTB, etc. are the signal strength of the individual PM 2006. The Physics of Diagnostic Imaging, 2nd edn., CRC Press; tubes. X and Y are then used to determine in which sub-element Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd the event occurred. A few modifications have been suggested edn., John Wiley & Sons, Inc.: New York, Chichester, Weinheim, to the original block detector design. One is to use bigger PM Brisbane, Toronto, Singapore. tubes so that each detector segment is monitored by a quarter of four different PM tubes. This method is called quadrant sharing Detector scatter event and it reduces the number of PM tubes used by a factor four. (Nuclear Medicine) Detector scatter events refer to events This method has its disadvantages; one of them is higher dead in which a photon undergoes Compton scattering inside the time losses since each PM tube handles signals from several detector. If the scattering occurs in the crystal the scattered photon can either interact in a different location in the crys- tal or escape. In the latter case it is possible to discriminate events by analysing the pulse height, which is proportional to the amount of energy deposited. If the energy deposited is less Top than the lower threshold in the energy window, the event is discarded. But if the scattered photon interacts in a different part of the crystal the total amount of deposited energy will be within the energy window. The detector will not be able to separate the two events and the registered event will there- Front R side fore be placed somewhere in-between the two interactions. The same effect occurs at high count rates if two events are simul- taneously detected. Misplacing detected events causes a loss of image contrast. Another type of detector scatter event is a collimator scat- tered event, i.e. photons that scatter in the collimator before being PMT detected. Such events can be discriminated using an energy win- A back left dow. If however the system energy resolution is low, some of these events might be registered as true events. Related Article: Collimators PMTC PMTD PMTD PMTB Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, FIGURE D.22 Diagram of block detector. Philadelphia, PA, p. 222. Detector total efficiency 253 D FT (discrete Fourier transform) Detector total efficiency Although there are some differences in the chemistry of devel- (Radiation Protection) The total efficiency of a radiation detector oper solutions supplied by various manufacturers, most contain (ET) for specific radiation is the product of quantum efficiency the same basic chemicals. Each chemical has a specific function (EQ) and conversion efficiency (EC): in the development process. Reducer: Chemical reduction of the exposed silver bromide E grains is the process that converts them into visible metallic sil- T = EQ × EC ver. This action is typically provided by two chemicals in the Using the examples of EQ and EC from above, the total efficiency solution: phenidone and hydroquinone. Phenidone is the more of a gas detector will be 0.01 (1%), and in the same way for a scin- active and primarily produces the mid to lower portion of the tillation detector will be about 0.4 (40%), while for a semiconduc- greyscale. Hydroquinone produces the very dense, or dark, areas tor detector ET it will be almost 1 (100%). in an image. See Detector quantum efficiency and Detector conversion Activator: The primary function of the activator, typically efficiency. sodium carbonate, is to soften and swell the emulsion so that the Related Articles: Detection efficiency, Detector quantum reducers can reach the exposed grains. efficiency, Detector conversion efficiency Restrainer: Potassium bromide is generally used as a D Further Readings: Dowsett, D., P. Kenny, and R. E. Johnston. restrainer. Its function is to moderate the rate of development. 2006. The Physics of Diagnostic Imaging, 2nd edn., CRC Press; Preservative: Sodium sulphite, a typical preservative, helps Knoll, G. F. 2000. Radiation Detection and Measurement, 3rd protect the reducing agents from oxidation because of their edn., John Wiley & Sons, Inc., New York, Chichester, Weinheim, contact with air. It also reacts with oxidation products to reduce Brisbane, Toronto, Singapore. their activity. Hardener: Glutaraldehyde is used as a hardener to retard the swelling of the emulsion. This is necessary in automatic proces- Deterministic effects sors in which the film is transported by a system of rollers. (Radiation Protection) The detrimental biological effects of expo- Related Article: Film processing sure to ionising radiation seen at higher doses/dose rates. These effects occur above a threshold in all persons exposed and the severity increases with the dose received. Development time Related Articles: Non-stochastic effects, Stochastic effects (Diagnostic Radiology) The development time is the duration that a film is actually in the developer chemistry. Since devel- Detriment opment is a progressive process, the development time is a fac- (Radiation Protection) Detriment is a mathematical expectation tor that determines the amount or degree of development that of harm caused to a population by exposure to ionising radiation, is produced. If the development time is too short, some of the taking into account not only the probabilities of each type of del- exposed silver halide crystals will not be converted and the eterious effect but also the severity of the effect. Detriment also density or darkness of the film will be reduced. If the devel- includes deleterious effects not associated with health, such as opment time is too long, some of the unexposed crystals will the restriction of the use of some areas or products. Assuming a be converted resulting in excessive film density or darkness linear dose–response relationship, the detriment associated with (fogging). health may be directly related to the collective dose equivalent The optimum development time is determined by the design of commitment. the film, composition of the developer chemistry, and the devel- oper temperature. In most automatic film processors it is set to a Deuterium fixed value. (General) Deuterium is an isotope of hydrogen with one proton Related Article: Film processing and one neutron in the nucleus. It is stable and is neither radioac- tive nor toxic. Its natural abundance is 0.0115%. It

is represented Dewar by either the chemical symbol D or 2H. Approximately one part (General) A Dewar or Dewar flask is a container designed to ther- in 5000 of the hydrogen in seawater is deuterium. The nucleus of mally insulate the contents and used to store or transport liquids deuterium is referred to as the deuteron. at temperatures different to the surroundings. Applications and Uses: Deuterium is used as a solvent in pro- Named after its inventor Sir Thomas Dewar, a Scottish scien- ton nuclear magnetic resonance spectroscopy and as a non-radio- tist, and first patented in the United States under the trade name active tracer to study chemical and biochemical processes, where ‘Thermos’, the Dewar or Thermos consists of two thin walled its presence is measured by mass spectroscopy. It is also used in vessels separated by a vacuum to prevent the conduction and the form of heavy water (D2O) as a moderator in nuclear reactors. convection of heat, and with a reflective layer to reduce thermal Related Articles: Hydrogen, Isotope, Neutron, Nuclear radiation. magnetic resonance, Proton Originally made from glass with cork stoppers, some are now made from plastics and come inside integral plastic or metal con- Developer tainer to provide a more practical and rugged device (Figures D.23 (Diagnostic Radiology) The developer is a mixture of chemicals and D.24). that converts the invisible latent image in an exposed film to a Large Dewar flasks are used for storing materials for long visible image. The exposure activates some of the silver halide periods at the temperature of liquid nitrogen and are sometimes crystals which make them sensitive to the developer chemistry. referred to as ‘cryostats’. During the development process the developer converts the acti- vated silver halide crystals into small specks of black metallic DFT (discrete Fourier transform) silver. (General) See Discrete Fourier transform (DFT) Diagnostic radiology 254 Diagnostic x-ray Diagnostic radiology is also the medical specialty for physi- cians with the education and training to make diagnosis from the images and to conduct related medical procedures (radiologists). Currently many countries use the term medical imaging as a broader professional descriptor than diagnostic radiology. Diagnostic reference level (DRL) (Diagnostic Radiology) The ICRP Publication 135 (2017) intro- duces four different terms: • DRL (a form of investigations level) • DRL quantity (a commonly and easily measured metric) • DLR value (a value of a DRL quantity, obtained from D surveys or other means) • DRL process (the cyclical process of establishing, using and updating DRL values) Radiation protection of the patient is based on the basic principles of justification and optimisation. DRLs are intended as investigations at levels adopted to sup- port the optimisation process in diagnostic and interventional procedure. DRL is not the suggested or ideal dose for a particular procedure or an absolute upper limit for dose. In conjunction with FIGURE D.23 A small domestic Dewar flask. an image quality assessment, reference levels act as ‘trigger lev- els’ to initiate quality improvement. Many professional and regulatory organisations, including the ICRP, American College of Radiology (ACR), American Double layer glass Association of Physicists in Medicine (AAPM), United Kingdom container (U.K.) Health Protection Agency, International Atomic Energy Vacuum Agency (IAEA) and European Commission (EC) endorse the establishment, regular review and use of diagnostic reference lev- els for radiodiagnostic examinations. DRL quantities should be easily assessed, if possible from a direct measurement for the examination. Table D.3 reports the recommended quantities for different equipment. DRL values are typically set at the 75th percentile value of the distribution of median values of a DRL quantity obtained from a survey conducted across a broad user base (i.e. large and small facilities, public and private, hospital and out-patient). They are established both regionally and nationally, and considerable vari- ations have been seen across both regions and countries. Median values of distributions of DRL quantities at a facility should be compared with DRL values to support the optimisation process. Dose surveys should be repeated periodically to establish new reference levels to ensure that the optimisation process continues to evolve. Silvered inner surface Diagnostic x-ray (Diagnostic Radiology) Diagnostic x-ray is the use of x-radia- tion to produce images for diagnostic purposes. This includes FIGURE D.24 Dewar section. radiography, mammography, fluoroscopy, and computed tomog- raphy (CT). Different x-ray spectra are used for the various procedures to Diagnostic radiology optimise the contrast characteristics and exposure to the patient. (Diagnostic Radiology) Diagnostic radiology is the use of radia- The KV (in the units of kVp) is the primary factor used to adjust tion for diagnosing, evaluating, and managing diseases and the spectrum. Typical values are shown on Figure D.25. injuries. Most diagnostic radiology methods produce images. Relatively low values (24–32 kV) are used for mammography X-radiation was the first and is still the most common radia- to produce the required high contrast sensitivity to visualise soft tion used in diagnostic radiology but radiation from radioactive tissue structures and small calcifications. At the other extreme, sources are also used along with RF signals (in magnetic reso- high KV values are used for chest imaging to provide good pen- nance imaging) and ultrasound. etration through bones. Diamagnetic materials 255 Diamond detector Related Articles: Paramagnetic, Susceptibility, Magnetic flux TABLE D.3 density Recommended DRL Quantities for Different Diamond detector Equipment (Radiation Protection) Diamond is a material with a very large Modality Quantity [unit] band gap (∼5.6 eV), and can be operated as a simple conduction Radiography Entrance-surface air kerma [mGy] counter by applying ohmic contacts to opposite faces of the crys- KAP [mGy·cm2] tal. The dosimeter is based on a natural diamond crystal sealed Mammography, Entrance-surface air kerma [mGy] in a polystyrene housing with a bias applied through thin golden breast Incident air kerma [mGy] contacts. The interactions of ionising radiation induce a tempo- tomosynthesis Mean glandular dose [mGy] rary change in the electrical conductivity of the diamond through the production of electrons and positive holes that have sufficient Dental intra-oral Incident air kerma [mGy] energy to be free to move through the crystal. Dental panoramic KAP [mGy·cm2] It may be used for detection of ionising radiation in an active Diagnostic KAP [mGy·cm2] mode as well as in a passive one. The atomic number of carbon (Z = fluoroscopy, D 6) is close to soft tissue (Z = 7.5). The interactions of ionising radia- interventional KAP [mGy·cm2] tion induce a temporary change in the electrical conductivity of the fluoroscopy Air kerma at the patient entrance reference point diamond through the production of electrons and positive holes that [Gy] have sufficient energy to be free to move through the crystal. Fluoroscopy time [s] The x-ray or gamma radiation undergoing a photoelectric Number of images in cine [Number] interaction in diamond disappears and liberated electrons are free digital subtraction angiography runs [Number] to move from the valence to the conduction band. Diamond detec- CT CTDIvol [mGy] tor working in active mode consists of natural diamond crystal DLP [mGy·cm] sealed in a polystyrene housing (or a thin diamond polycrystalic Cone-beam CT Air kerma at the patient entrance reference point membrane, about 100 μm) placed between two electrodes, e.g. Cr/ [Gy] Au. This detector connected to an electrometer may be used to KAP [mGy·cm2] x-rays and alpha particles dose measurement in the range up to 106 CTDIvol [mGy] Gy. Besides, the dose detection can be performed using diamond DLP [mGy·cm] resistor. The diamond electrical resistivity is inverse proportional depending on the available quantities to the dose rate from x-rays, gamma radiation and electrons beam. Nuclear medicine Administered activity [MBq] Diamond detector (Figure D.26) works also in a passive ther- or Activity per body weighty [MBq/kg] moluminescence mode. The thermoluminescence is the emission of optical radia- tion in the form of prompt fluorescence when the diamond is heated. Diamond contains deep traps in the bandgap between the Barium valence and conduction bands. Some of the electrons fall into the Head traps which are too deep to release them. If the energy is deliv- spine ered to the diamond by heating it, the electrons can move up to shoulder conduction band and hence produce photon of visible radiation. Hip The radiosensitivity of a synthetic diamond crystal depends on Iodine Abdomen the conditions in which these crystals are grown. The diamond dosimeter must have sufficient signal reproducibility, low fading Breast Extremity Chest 0 20 40 60 80 120 KV FIGURE D.25 Typical kVp for some radiographic examinations. (Courtesy of Sprawls Foundation, www .sprawls .org) Diamagnetic materials (Magnetic Resonance) Substances without permanent atomic electronic magnetic moments are called diamagnetic. The mol- ecules have filled electron shells and therefore no net magnetic moment. When placed in an external magnetic field, a weak mag- netic field is induced in the opposite direction to Bo according to Lenz’s law. As a result, the effective magnetic flow density is reduced. Thus diamagnetic substances have a negative magnetic susceptibility χ on the order of 10−6. Most tissues in the body are FIGURE D.26 (Photo courtesy of Ms Barbara Marczewska, PhD, diamagnetic. Differences in susceptibility lead to B0 distortions, Institute of Nuclear Physics, Polish Academy of Sciences, Warsaw, e.g. at the air-tissue interface above the paranasal sinuses. Poland.) Diaphragm (collimator) 256 DICOM (loss of trapped electrons at the room temperature) and linear pro- portionality between signal and dose over a relatively wide range of doses as well as mechanical and chemical stability. The dose response can be represented either by the measured light intensity as a function of detector temperature in a glow curve or by the linearity factor f(D) defined as the ratio of the measured intensity I per delivered dose D, normalised to this ratio over the linear range of dose-response: f (D) (I / D) = (I / D) lin The diamond detectors are applied for the doses exceeding 1 D Gy. Diamond detectors are attractive in megavoltage photon and electron beam measurements either for the high spatial resolution because of their small size or for their near tissue equivalence. (Z = 6 compared to Zeff = 7.42 for soft tissue). Other advantages FIGURE D.27 Collimator (light beam diaphragm [LBD]) with square include high sensitivity, low leakage current and high resistance shutter. to radiation damage. Diamond dosimeters are designed to mea- sure relative dose distributions in high energy photon and electron beams, especially in high dose gradient regions. Due to the high density of carbon in comparison to air, the recombination of charge carriers is much more important for diamond detectors than it is for ionisation chambers. The recom- bination rate in a crystal, when an equilibrium number of free electrons is established, is proportional to square root of the rate of ion-pair production and hence to the dose rate. This is due to the increase in the probability of recombination with the number of vacant hole. The charge collection efficiency therefore decreases with the dose rate. If impurities are present, metastable states are introduced which trap many electrons which would otherwise recombine with holes. If the number of electron in traps is large, the proportional increase in the number of vacant holes can be almost independent of dose rate. This means that the recombina- tion rate, and hence the efficiency of charge collection, is almost independent of the rate of ion-pair production, giving an almost linear increase in detector signal with dose rate. The response of the diamond detector has been shown to be FIGURE D.28 Collimator with circular shutter. (Courtesy of Ing. A nearly independent of the incident photon energy, since the mass Litchev, Medical University Plovdiv, Plovdiv, Bulgaria.) attenuation coefficient ratio of water to carbon is almost nearly constant. The company General Electric introduced a similar detector light field and the x-ray field has to be checked during quality using garnet (gemstone detector). -control of x-ray equipment. This usually uses a simple tool with copper marks, according to which the light field is adjusted and Diaphragm (collimator) further exposed (Figure D.30). Displacement of more than 1 cm (Diagnostic Radiology) Diaphragm is usually only a metal piece usually requires re-adjustment of the LBD mechanism. (absorber) with an opening, mounted beneath the x-ray tube. It Related Articles: Beam restrictor, Light localiser simply restricts the x-ray beam filed. Complex diaphragms are called collimators. All these tools are beam restrictors. They have DICOM metal shutters (usually lead) which allow changing the size

of (General) DICOM (Digital Imaging and Communications in the irradiating field. Most often these devices include also a light Medicine) is the international standard to transmit, store, retrieve, localiser which projects a light field with the same shape and size print, process and display medical imaging information. as the x-ray beam field. Such devices are also known as light beam DICOM’s main characteristics: diaphragm (LBD). A typical LBD for x-ray radiographic devices • It makes medical imaging information interoperable is shown on Figure D.27. • It integrates image-acquisition devices, PACS, worksta- The movable shutters of the collimator (LBD) produce most tions, VNAs and printers from different manufacturers often a rectangular field (to match the size of the x-ray film), also • It is actively developed and maintained to meet the known as jaws. However shutters used in fluoroscopy produce cir- evolving technologies and needs of medical imaging cular field (to match the image field of the Image Intensifier) – see • It is free to download and use Figure D.28. DICOM was officially introduced in 1993 by the American College The light beam inside the LBD uses a system of mirrors to of Radiology (ACR) and the National Electrical Manufacturers mimic the x-ray beam size (Figure D.29). The coincidence of the Association (NEMA) as the third version of the ARC-NEMA DICOM-RT 257 DICOM-RT A DICOM data object consists of a number of attributes, gen- erally referred to as meta info/header and data set. The meta info portion of the DICOM file contains the file preamble and prefix, while the dataset contains a number of data elements. A single DICOM object can have only one attribute containing pixel data. DICOM uses three different data element encoding schemes. The same basic format is used for all applications, including network 1 and file usage, but when written to a file, usually a true ‘header’ (containing copies of a few key attributes and details of the appli- 2 cation which wrote it) is added. DICOM is among the most widely used standards in medicine. 3 It ensures the interoperability of healthcare systems used to: pro- 4 duce, store, display, send, query, process, retrieve, print medical images and derived structured documents, as well as to manage the related workflow. D Benefits for medical staff: better management of image data; remote access; faster diagnosis. Benefits for patients: faster and more effective care; easier access to health information (Figure D.31). Hyperlink: www .dicomstandard .org DICOM-RT (Radiotherapy) DICOM-RT is a subset of the DICOM 3 standard FIGURE D.29 Construction of typical light beam diaphragm with applicable in the radiotherapy domain. It usually contains the square shutter: 1,2 – Shielded light source (imitation of point source); 3,4 images (e.g. CT slices) along with associated data such as patient – mirrors. details and scan acquisition parameters and any structures, etc. which have been outlined. DICOM RT specifies a series of additional data objects spe- cific to radiotherapy: • RT image (e.g. DRRs, portal images, simulator images) • RT structure set (e.g. contours, regions of interest, refer- ence points, points of interest, isocenters) • RT plan (e.g. geometric linac parameters, fractionation, patient setup details) • RT dose (e.g. dose matrix, isodoses, DVHs) • RT record (e.g. number of fractions, doses received) DICOM also specifies a series of communication actions: • Storage class user (i.e. sends data) • Storage class provider (i.e. receives data) FIGURE D.30 Radiograph of test object used to check the light beam dia- phragm alignment. (Courtesy of EMERALD project, www .emerald2 .eu) standard developed by the two organisations in the 1980s. Since then, DICOM has turned into the most widely used tool for digital data exchange in healthcare. The DICOM standard is managed by the Medical Imaging & Technology Alliance – a division of the National Electrical Manufacturers Association. DICOM is updated and republished several times per year. The goal of the DICOM standard is to achieve compatibility and improve workflow efficiency between imaging systems and other information systems in healthcare environments worldwide. DICOM’s innovative approach is to keep the data in datasets rather than groups of individual files. This approach guarantees data integrity and traceability due to the fact that the data header is an integral part of the complete dataset. FIGURE D.31 DICOM data organisation. Dielectric 258 Differentiation • Query-Retrieve service class (i.e. asks for data and The integral cross section is the integral of the differential receives data requested) cross section on the whole sphere of observation: • PrintService class (i.e. prints data) s s = òæ d ö ç ÷dW Abbreviations: ACR = American College of Radiology, è dW ø DICOM = Digital imaging and communications in medicine, DRR = Digitally reconstructed radiograph, DVH = Dose vol- Related Articles: Radiation scattering, Cross section ume histogram and NEMA = National Electrical Manufacturers Association. Differential scattering cross section Hyperlink: http://medical .nema .org/ (Ultrasound) The differential scattering cross section is defined as the time-averaged scattered power in the direction (θ, ϕ) per Dielectric unit solid angle (the solid angle being related to the surface of a (General) A dielectric material is a substance where charges are sphere in the same way as an ordinary angle is related to the cir- D bound so that its conductivity is ideally zero. Even if a field applied cumference of a circle). In other words this describes how sound to the dielectric produces no migration of charge, it can produce is scattered with angular direction. Integrating this over the sur- a polarisation of the dielectric which consists in a displacement face of a sphere with a radius larger than the particle yields the of the electrons with respect to their equilibrium positions. In a scattering cross section (hence the name differential scattering dielectric charges are not free to move far enough to completely cross section). Of special importance is the differential scattering cancel the effect of any external electric field but they can move in the opposite direction of the incoming sound, called the dif- far enough to cause a partial cancellation. ferential back-scattering cross section. Related Article: Conductivity Related Articles: Scattering cross section, Absorption cross section, Extinction cross section Dielectric constant (General) The electric field within a dielectric is reduced by the Differentiation factor 1/ε from that which would exist without the dielectric. ε is (General) Differentiation is a mathematical method to obtain the called dielectric constant of the dielectric. rate of change or slope of a curve at a given point. Consider the Related Article: Dielectric function y = f(x). The slope s of the curve between the points x and x + δx is given by the equation: Differential absorption f ( x + dx) - f (x) (Radiation Protection) The concept of differential absorption s = (D.5) describes any other situation where the component parts of an dx object have unequal radiation absorption (and scattering) charac- As the value of δx tends to zero, the value s becomes the differen- teristics. It is often associated with radiological image formation. tial (the derivative) dy/dx of the function at the point x: Radiological images are produced when the x-ray beams pass through an inhomogeneous object and reach the image receptor system. When x-rays penetrate such inhomogeneous tissue, they dy æ f ( x + dx) - f (x) ö = limç ( . dx x 0 ç ÷ D 6 ® è dx ÷ ) are not homogeneously absorbed; some tissues absorb x-rays ø more efficiently than others. The image receptor system displays these differences in If the function represents a straight line such as y = 2x + 3, then absorption as variations in brightness across the image that the slope of the line is 2 and it is the same at every point. This can describes the relative radio-opacity of each part of the body being be demonstrated by putting these data into Equation D.6: imaged: e.g. bone is more radio-opaque than soft tissues, and therefore appears as a different brightness. dy æ 2x + 2dx + 3 - 2x - 3 = ö lim D dx x 0 ç ÷ = 2 ( .7) ® è dx ø Differential cross section (Radiation Protection) The differential cross section is a descrip- In nuclear medicine, differentiation is very evident in the law of tion of the probability of scattering in a particular direction. In radioactive decay. This states that in a sample containing N radio- other words, the differential cross section can be defined by the active atoms of a particular radionuclide, the average decay rate fraction of particles that can be found within a given cone of ΔN/Δt is given by the equation observation of solid unit angle when a target is irradiated: DN = -lN (D.8) ds æ F ö = s Su Dt dW ç ÷ è Fi ø Wu where λ is the decay constant and t is the time. In the limit as t where tends to zero this equation becomes σ is the scattering cross section Ω is the solid angle dN = -lN (D.9) Φi is the incident flux dt Φs is the scattered flux Su is the surface unit Equation D.9 represents the decay rate or activity of the radioac- Ωu is the solid angle unit tive sample. Diffraction 259 Diffraction-enhanced imaging Ultrasound field from a single transducer-continued X X Near zone Far zone (fresnel zone) (fraunhofer zone) Ultrasound beam pressure amplitude profile at X-X FIGURE D.33 Pressure distribution from a circular disc ultrasound transducer. Near field close to the transducer surface to the left. (Courtesy D Near field-complex of EMIT project, www .emerald2 .eu) Far field-more distribution of pressure uniform distribution with interface of pressure FIGURE D.32 The ultrasound field, Huygen’s wavelets. The near field and far field. (Courtesy of EMIT project, www .emerald2 .eu) Diffraction (Ultrasound) Diffraction is a phenomenon describing the effect on a wave that travels throw some kind of an aperture. How the sound wave spreads out as it moves away from the aperture depends on the shape and size of the aperture and the wavelength. FIGURE D.34 Pressure distribution from a circular disc ultrasound In ultrasound applications the referred aperture is often the ultra- transducer. Near field close to the transducer surface to the left. Side lobes sound transducer, Figure D.32, and in many cases the wanted are clearly shown. (Courtesy of EMIT project, www .emerald2 .eu) beam should be as collimated as possible. The Huygen’s principle is frequently used to describe diffraction, side lobes and other wave phenomena. For example, the ultrasound distribution from a a map of the sample-induced refraction angle in the direction transducer surface can be calculated by dividing the surface area orthogonal to the plane defined by the laminar x-ray beam. in a large number of point sources and add up the contributions Given the popularity of this algorithm, the term DEI is some- from all of these. The waves from the fictive point sources inter- times used interchangeably with ABI to indicate the imaging fere when in phase and cancel each other when out of phase. This method itself. theory opens up for the solution with many small transducer ele- The rocking curve is obtained by measuring the intensity ments in a matrix to create a collimated beam, which is the case reflected (or transmitted) by the analyser crystal toward the imag- for modern B-mode imaging. ing detector as a function of the angular displacement between the The diffraction pattern of a plane disc source can be divided analyser and monochromator diffraction planes. This bell-shaped into a complex near field (near zone or Fresnel zone) that has a curve describes the angular band-pass filter due to the analyser cylindrical shape with approximately the same radius as the system, and it is key in transforming refraction angles into detect- source, and the far field (far zone or Frauenhofer zone) that able intensity modulations. diverges with the angel Θ = arcsin (0.61λ /a) where a is the radius When the sample is present, the intensity reaching the image of the disc and λ is the wavelength, Figure D.33. The far field plane can be written as: starts at the distance a2/λ from the source. This implies that a large disc compared to the used wavelength produce a long near I(q0; x, y) = IR(x, y)R(q0 + DqR(x, y))

field with less divergence angle. The pressure distribution from a circular disc transducer is shown in Figure D.34. where R is the rocking curve, θ0 is the angular displacement or Related Articles: Transducer, Ultrasonic field, Side lobes working point, ΔθR is the refraction angle due to the sample, IR is the apparent absorption intensity, and the coordinates x,y identify Diffraction-enhanced imaging the object plane. (Diagnostic Radiology) Diffraction-enhanced imaging (DEI) For sufficiently small refraction angles, the flanks of the is an algorithm applied to images acquired with the analyser- rocking curve (see Figure D.35) can be well described by the based imaging (ABI) technique. DEI consists in the combina- first-order Taylor expansion around the angles identifying its full- tion of two images of a given sample acquired at the flanks of width-half-maximum (FWHM): the rocking curve, that allows for the separation of absorption and phase effects (sometimes referred to as ‘phase retrieval’). I(q0; x, y) = IR(x, y) ×[R(q0 ) + R¢(q0 )DqR(x, y)] The application of DEI results in two parametric images: the apparent absorption image accounts for sample’s absorption where R’(θ0) is the first derivative of the rocking curve at the and scattering to large angles, while the refraction image is position θ0. Diffraction imaging 260 Diffraction imaging From synchrotrons to conventional sources. Rivista del nuovo cimento 37(9):467–508; Rigon, L., F. Arfelli, and R. H. Menk. 2007. Generalized diffraction enhanced imaging to retrieve absorption, refraction and scattering effects. J. Phys. D: Appl. Phys. 40(10):3077; Rigon, L. et al. 2008. Generalized diffraction enhanced imaging: Application to tomography. Eur. J. Radiol. 68(3):S3–S7. Diffraction imaging (Diagnostic Radiology) Diffraction imaging is a technique used to characterise materials with a degree of short-range or long- range order. It is based on the different x-ray diffraction (XRD) signatures of different materials, i.e. the peaks in the intensity of D the coherently scattered radiation measured as a function of the momentum transfer, q, defined as FIGURE D.35 Rocking curve with the two working points θ1 and θ2, corresponding to the flanks of the rocking curve (at FWHM) used in the 1 J ö q = æ sin DEI algorithm. The FWHM of the rocking curve, which is usually in the l ç 2 ÷ è ø order of few tens of microradians, is indicated with ∆θD. where λ is the wavelength of the x-rays and θ is the scattering DEI consists in combining the two images acquired at work- angle. ing points θ1 = –∆θD/2, θ2 = ∆θD/2, ∆θD being the FWHM of A range of momentum transfer values can therefore be the rocking curve. In this way, by using the previous formula, achieved by using a monoenergetic beam and varying the scat- we obtain a system of two independent equations, which can be ter angle (angle-dispersive diffraction) or by keeping the scatter solved pixel-by-pixel to yield the unknowns images IR and Δθ angle constant and using a polychromatic x-ray beam and a spec- R: troscopic detector (energy-dispersive diffraction). In highly-ordered structures, the constructive interference I(q I x 1; x, y)R¢(q2 ) - I(q2; x, y)R¢(q ) R( , y) = 1 between different crystalline planes causes the diffraction pattern R(q1)R¢(q2 ) - R(q2 )R¢(q1) of the material to feature sharp peaks; in structures with a lower degree of order, such as some biological tissue, the peaks still I exist but are broader. Dq (q2; x, y)R(q ) - I(q ; x, y)R(q ) R(x, y) = 1 1 2 I(q1; x, y)R¢(q2 ) - I(q2; x, y)R¢ Although x-ray diffraction was originally used to analyse (q1) homogeneous materials as it does not intrinsically provide spatial information, various approaches to diffraction imaging have been In addition to refraction and apparent absorption maps, DEI suggested to provide spatial information, involving a pencil beam approach has been extended to include also ultra-small-angle- and sample scanning, or a laminar beam, collimation and scanning scattering (USAXS), yielding three parametric images. This gen- of the sample. As the yield of scattered photons within a narrow eralised DEI (GDEI) algorithm requires three images as input, angle is small, the dose needed to achieve good statistics is very typically acquired at the flanks and on the top of the rocking high; therefore, the technique is not applicable in vivo, but it can curve. In its original version, GDEI makes use of second-order be used for in vitro tissue characterisation, either in planar or in Taylor expansion of the rocking curve, where the second-order tomographic geometry. In particular, promising results have been term contains the angular dispersion due to USAXS. obtained for characterisation of breast tissue (Pani et al., 2010; The rocking curve description through a Taylor expansion can Moss et al., 2017; Griffiths et al., 2008; Castro et al., 2005) due to be seen as a limitation of these algorithms since it requires the the significant differences in the scatter signatures of normal and refraction and/or scattering angles to be small when compared neoplastic breast tissue, shown in Figure D.36 (Pani et al., 2010). with the rocking curve width. To overcome this issue, alterna- tive retrieval algorithms, as extended DEI and G2DEI, have been implemented: these approaches model the rocking curve, e.g. a Gaussian function, thus not requiring any series expansion and extending the range of validity of the retrieval method to larger refraction and scattering angles. Of note, the formalism introduced to describe DEI algorithms can be translated with minor modifications to the edge illumina- tion imaging technique, where the rocking curve is substituted by the illumination curve. Related Articles: Analyser-based imaging, Phase constrast imaging, Edge illumination Further Readings: Chapman, D. et al. 1997. Diffraction enhanced x-ray imaging. Phys. Med. Biol. 42(11):2015; Diemoz, P.C. et al. 2017. Non-interferometric techniques for x-ray phase- contrast biomedical imaging. In P. Russo (ed.), Handbook of X-ray Imaging: Physics and Technology. CRC Press, pp. 999–1023; FIGURE D.36 XRD patterns from healthy and neoplastic breast tissue. Olivo, A., and E. Castelli. 2014. X-ray phase contrast imaging: (Pani et al., 2010.) Diffusion 261 Diffusion kurtosis imaging (DKI) ár2ñ = 2nDt where r is the displacement n is the dimension of the diffusion (n = 1 in diffusion MRI) D is the diffusion coefficient t is the time during which the diffusion is measured In MRI, the self-diffusion of water is studied, where spin dis- placements produce signal decay due to two subsequent magnetic field gradients of opposite polarity (see Diffusion imaging). Diffusion kurtosis imaging (DKI) (Magnetic Resonance) Diffusion kurtosis imaging (DKI) is an D extension of conventional diffusion-weighted imaging where the apparent diffusion coefficient (ADC) map is calculated on the Gaussian diffusion model. In addition to the apparent diffu- sion coefficient (Dapp), the method provides the excess kurtosis (Kapp) of the diffusion displacement probability distribution, which is a metric of the departure from a Gaussian model (see Figure D.38). DKI requires the acquisition of diffusion- weighted images, some of them obtained at b-values higher than in ADC studies, and a modified image postprocessing pro- cedure. Parametric maps of Dapp and Kapp are created by fitting the image signal intensities to the formula in Figure D.38, on a voxel-by-voxel basis. Tissue structure (cellular compartments and membranes) is responsible for the deviation of water diffusion from the Gaussian behaviour observed in homogeneous solutions, so far Kapp maps have been used in tissue and lesion characterisation studies. In clinical studies, 5 b-values, or more, are applied in the range 200– 2500 s/mm2. The DKI investigates slow components of the diffusion pro- FIGURE D.37 transmission CT (a) and diffraction CT (b) of a breast cess, detectable using high b-values. Conversely, intravoxel inco- tissue sample, highlighting healthy (H) and pathological (P) tissue. herent motion (IVIM) studies compare perfusion and diffusion effects in the diffusion-weighted images applying very low b values. Figure D.37 (Castro et al., 2005) shows comparison of an XRD CT image of a breast tissue sample with the corresponding transmission CT, showing discrimination between healthy and pathological tissue. Abbreviations: XRD = X-ray diffraction, CT = computed tomography. Further Readings: Castro, C. R. F. et al. 2005. Coherent scattering X-ray imaging at the Brazilian National Synchrotron Laboratory: Preliminary breast images. Nucl. Inst. Meth. A 548:116–122; Griffiths, J. A. et al. 2007. Correlation of energy dis- persive diffraction signatures and microCT of small breast tissue samples with pathological analysis. Phys. Med. Biol. 52(20):6151– 6164; Moss, R. M. et al. 2017. Correlation of X-ray diffraction sig- natures of breast tissues and their histopathological classification. Sci. Rep. 7:12998; Pani, S. et al. 2010. Characterization of breast tissue using energy-dispersive X-ray diffraction computed tomog- raphy. Appl. Radiat. Isotopes 68(10):1980–1987. Diffusion (Magnetic Resonance) Diffusion is the phenomenon by which mol- ecules or particles randomly migrate due to their thermal energy. This random motion is called a Brownian motion and is not only FIGURE D.38 Diffusion signal intensity decay (ln S/S0) versus b-value dependent on the temperature, but also on the size and surround- (s/mm2). Comparison of monoexponential model, reflecting solely ings of the molecules. Free isotropic diffusion, which occurs in, Gaussian diffusion, and kurtosis model, reflecting both Gaussian and e.g. a glass of water, is described by a Gaussian distribution with non-Gaussian diffusion components. ADC is calculated in the b-value mean zero and a variance given by Einstein’s equation as range 0–1000 s/mm2. Diffusion encoding 262 Diffusion spectrum imaging (DSI) Related Articles: Apparent diffusion coefficient (ADC), y Diffusion weighting, Diffusion encoding, b-values, Intravoxel incoherent motion (IVIM) Further Reading: Jens, H. Jensen et al. 2005. Diffusional kur- tosis imaging: The quantification of non Gaussian water diffusion by means of magnetic resonance. Imaging Magn. Resonan. Med. 53:1432–1440. x Diffusion encoding (Magnetic Resonance) A spin group exposed to an RF pulse will exhibit a transverse magnetisation component. When exposed to a magnetic field gradient, this component will acquire an additional phase angle in the rotating frame of reference. A ‘particle’ moving D in a magnetic field gradient will subsequently be affected by dif- FIGURE D.40 In one voxel there are many spins groups and the net ferent magnitudes of the gradient field, depending on its trajectory. effect after two diffusion encoding gradients will be a phase dispersion By submitting a particle in motion to two gradient pulses with the reducing the amplitude of the net magnetisation vector. same magnitude but with opposite polarity (obtained by the 180° pulse), a net phase shift is therefore obtained (Figure D.39). The particle in black, which is seen to the left in Figure D.39a, is moving The net magnetisation of a system exposed to the two gra- to a new position, which is indicated in the right image. The par- dient pulses will be lower if the particles are diffusing and this ticle will be affected by gradients of different polarity and different will lead to phase dispersion and a subsequent loss in signal, see magnitude (Figure D.39b). The first gradient will impose a phase Figure D.40. The sensitivity of the diffusion encoding is deter- shift, ϕ mined by the b-value (see b-value). With a higher b-value the 1 (Figure D.39c, left) and the second gradient will impose a phase shift in the opposite direction, ϕ2 (Figure D.39c, right). When phase dispersion will be more enhanced. the particle is moving the net effect will be a phase shift different Related Articles: b-value, Diffusion from zero (see Figure D.39d). A particle that is stationary during both gradient one and two, represented by a dotted arrow in Figure Diffusion imaging D.39b, will have the net phase shift, Φ, equal to zero. (Magnetic Resonance) Diffusion imaging is by now a well-estab- The net phase shift can be determined from lished technique within parametric (or, in a broad sense, func- tional) MRI. Several different types of images can be derived, t such as apparent diffusion coefficient (ADC) maps and fractional F(t) = gòG¢(t)x(t)dt anisotropy (FA) maps. The technique can also be used for diffu- sion tensor imaging (DTI) as well as for tractography. 0 The diffusion imaging method is based on an observed signal where decay occurring due to the self-diffusion of water in the presence G′(τ) is the time-dependent gradient strength of two subsequent pulsed gradient of opposite polarity. The gradi- x(τ) is the displacement along the direction of the diffusion ent pulses impart a location-dependent phase angle on the spins. encoding gradient Since the sign is reversed for the two gradient pulses, spins are γ

is the gyromagnetic ratio refocused in the case of no motion, but introduce a phase disper- sion in the case of moving spins. The phase of each spin is pro- portional to the distance traversed between the two pulses as well 180° as the strength and duration of the gradient pulses (see b-value) 90 ! and consequently, the signal decay is proportional to the average t displacement of the spins in combination with the b-value. (a) In diffusion imaging, the basis is the T2-weighted image together with a diffusion weighted (DW) image obtained with a specified diffusion sensitivity (b-value). Depending on the pur- pose of the examination, a protocol can consist of one T2-weighted (b = 0 s/mm2) image and several DW images obtained for several B Gradient 1 B Gradient 2 diffusion encoding directions. For clinical practice, the diffusion sensitivity is commonly set to b = 1000 s/mm2. x x In the case of stroke diagnosis, it can be sufficient to acquire DW images in three orthogonal directions, but if the purpose is (b) DTI, diffusion encoding must be performed in at least six non- y y y collinear directions. To obtain FA maps and tractographies of φ1 Φ higher quality a larger number of directions is preferred. Related Articles: Apparent diffusion coefficient (ADC), x + x = x b-values, Diffusion tensor imaging (DTI), Fractional anisotropy (c) φ2 (FA), Tractography FIGURE D.39 A particle subjected to diffusion will acquire a phase Diffusion spectrum imaging (DSI) shift different from zero when it is affected by two gradient pulses of (Magnetic Resonance) Diffusion spectrum imaging (DSI) was equal amplitude but with different effective polarity. introduced as an attempt to solve the problems of crossing and Diffusion tensor 263 Diffusion time kissing fibres in diffusion tensor tractography (DTT). The method for highly anisotropic diffusion, while the displacement distribu- is based on a 3D probability density distribution, which can be tion giving rise to this anisotropic diffusion will be characterised determined if diffusion encoding is performed in a large number by an ellipsoid. of directions. A parallel to this technique is the so-called q-space The diffusion tensor assumes Gaussian diffusion and it has imaging technique where a large number of differently diffusion long been recognised that this model fails to resolve fibres of dif- sensitised images are acquired, but whereas the normal q-space ferent directions when present in the same voxel, e.g. fibre cross- imaging technique provides information regarding diffusion ings in cerebral white matter. This can be a problem in, e.g. fibre distances and possible restrictions, the DSI results in a 3D prob- tracking. ability density distribution showing the likelihood for different Related Articles: Apparent diffusion coefficient (ADC), diffusion directions. Fractional anisotropy (FA), Relative anisotropy (RA) Related Articles: q-space (used in diffusion MRI and NMR), Tractograph Diffusion tensor imaging (DTI) (Magnetic Resonance) Diffusion tensor imaging is a technique Diffusion tensor based on diffusion weighted imaging (DWI) that enables a three- (Magnetic Resonance) In a media where the self-diffusion is free dimensional representation of the water self-diffusion. The self- D and shows no directional dependence, the diffusion is called iso- diffusion of water in cerebral white matter (WM) is faster along tropic. In this case a single ADC (Apparent Diffusion Coefficient) than across the fibre bundles. By measuring the ADC in at least is sufficient to characterise the diffusion. However, in more com- six non-collinear diffusion encoding directions, the diffusion plex media, e.g. cerebral white matter, where the parallel bun- tensor can be estimated. Three different parametric maps are dles of myelin sheeted axons facilitates diffusion along a certain often calculated from a DTI measurement: (1) the mean diffusiv- direction, the diffusion becomes anisotropic. To characterise the ity map (MD) defined as the mean value of the eigenvalues of diffusion in such a case and assuming that the molecular displace- the diagonalised tensor (i.e. one third of the trace); (2) the degree ment distribution is Gaussian, a diffusion tensor model can be of diffusion anisotropy, determined from the variance of three employed. The diffusion tensor describes diffusion that may be eigenvalues, is often described by the FA; and (3) the preferred different in three arbitrary and orthogonal directions. The tensor diffusional direction, determined from the direction of the prin- is often represented by an ellipsoidal isosurface of the Gaussian cipal eigenvector, commonly represented by a colour coded map, probability distribution. This ellipsoidal is described by six inde- where blue colour corresponds to a superior-inferior direction, pendent parameters, of which three describes its height, width red to a left-right direction and green to anterior-posterior direc- and length and three its orientation. tion (Figure D.41). The diffusion tensor D describes the ADC in the direction n as Related Article: Fractional anisotropy (FA) Diffusion time éDxx Dxy Dxz ù énx ù (Magnetic Resonance) In diffusion-weighted (DW) magnetic- ADC(n) = nTDn = [nx y nz ]ê ú n êDxy Dyy D ê ú yz ú êny ú resonance imaging (MRI), when using a conventional Stejskal– ëêD D D ûúxz yz zz ëênz ûú Tanner pulsed gradient sequence, the diffusion time (TD) is defined as or written differently as d TD = D - 3 ADC(nn) = nn × D = é where ënxx nyy nzz 2nxy 2nyz 2nxz ùû×éë Dxx Dyy Dzz Dxy Dyz Dxz ùû Δ is the time between the leading edges of the pulsed gradients δ is the duration of the pulsed gradient where nij = ni … nj. Given that the ADC is measured in at least six non-collinear directions the diffusion tensor can be determined as In this case, the ramp-up and ramp-down times of the diffu- sion encoding gradients are approximated to zero. The diffusion time relates to the mean square distance <r2> traversed by the é nn1 ù é ADC (nn1 ) ù ê ú ê ú diffusing molecules as ê ú × D = ê ú Û N × D = ADC Þ D ëênnm ûú ê ú ëADC (nnm )û ár2ñ = 2DTD = (NTN)-1NTADC where D is the diffusion coefficient. This equation implies that if the diffusing molecules are restricted within a confinement, the where m is the number of directions in which the ADC is observed value of D (i.e. the apparent diffusion coefficient, ADC) measured. would be dependent on TD. This has indeed been observed on The diffusion tensor contains much information. For example, excised tissue (Assaf, 1999), but not in healthy volunteers (Clark one third of the trace of the tensor, i.e. the mean of the eigenvalues, et al. 2000). gives the mean ADC, while the variance of the eigenvalues cor- Related Articles: Apparent diffusion coefficient (ADC), responds to the anisotropy of the diffusion. Primarily, two metrics Diffusion weighting have been employed to characterise the diffusional anisotropy: Further Readings: Assaf, Y. and Y. Cohen. 1999. Structural the fractional anisotropy (FA) and the relative anisotropy (RA). Information in neuronal tissue as revealed by q-space diffusion The 3D distribution of the ADC-values will be shaped as a peanut NMR spectroscopy of metabolites in bovine optic nerve. NMR Diffusion weighting 264 Digital breast tomosynthesis D FIGURE D.41 A fractional anisotropy map, obtained from a healthy volunteer and the corresponding colour coded directional map. Biomed. 12: 335–344; Clark, C. A. and D. Le Bihan. 2000. Water may also move, depending upon the system design. Typically, a diffusion compartmentation and anisotropy at high b value in the small number (9–25) of low dose projection images are obtained human brain. Magn. Res. Med. 44: 852–859. over a limited range of angles (±7° to ±30°), from about the nor- mal to the desired image plane. Diffusion weighting (Magnetic Resonance) Diffusion weighting (DW) is a technique used in both chemistry and medicine for probing diffusion to understand microstructural characteristics with the use of mag- netic fields. By applying two additional magnetic field gradients to a SE pulse sequence, the measurement becomes sensitive to translational molecular motion, i.e. to self-diffusion. The first applied gradient gives the molecules a ‘phase label’, while the sec- ond gradient partly refocuses the phase of the molecules, depend- ing on the net-movements of the molecules. The strength of the diffusion encoding is normally defined by the b-value, defined as b = (γδG)2TD, where γ is the gyromag- netic ratio, δ and G the duration and amplitude of the diffusion encoding gradients and TD is the diffusion time. A higher b-value gives higher diffusion sensitivity for molecular motions. In clini- cal routine, a b-value of approximately 1000 s/mm2 is commonly used to generate DW images. DW imaging is most of all used for differential diagnosis and assessment of ischemic stroke. Related Articles: b-value, Diffusion time, Self diffusion, Spin echo Digital breast tomosynthesis (Diagnostic Radiology) Although mammography represents the golden standard for breast imaging, it suffers from the inherent limitations of planar 2D imaging, the main being the masking effect of tissue superposition. To overcome this limitation, tech- nology for the 3D imaging of the breast has been developed, through the application of digital breast tomosynthesis (DBT). The DBT method uses the idea of linear (classical) tomogra- phy; however, in DBT the detector does move. Instead, different elements of the digital detector become sensitive with tube move- Block diagram illustrating the principle of DBT image formation. ment, what creates the tomographic effect. Tomosynthesis images are obtained using a system that looks The x-ray spectra used in tomosynthesis are usually of higher very similar to the conventional mammography unit. However, energy, based for example on W/Al anode filter combinations. the arm supporting the x-ray tube pivots about a point, while at the The movement of the tube can be continued or in small steps same time the compressed breast remains stationary. The detector (also called ‘step and shoot’). As with mammography, the total Digital detector array 265 Digital detector with direct conversion acquisition time must be minimised to avoid image degradation 1. The upper level is responsible for the capture of the due to patient motion. x-ray photon, and its conversion to electrical charge, Once the acquisition of the images on different angles are either complete, planar cross-sectional images are reconstructed using i. Directly, with photoconductors, in the case of filtered back projection or an iterative reconstruction algorithm. direct conversion, or In terms of geometrical characteristics, it has to be highlighted ii. Indirectly with a phosphor and photodiodes, in the that the resolution during tomosynthesis is anisotropic, with lower case of indirect conversion. resolution between planes and higher in plane. 2. The next level houses the charge collection electrode Additional software can be utilised to incorporate the infor- and the TFT array. mation of all cross-sectional images into one 2D image (‘synthetic 3. The last level is the thin (~0.7 mm) glass substrate. image’), similar to the FFDM image, to minimise the need for acquiring the 2D image separately. They are self-scanned readout systems, meaning that the plane Related Articles: Full field digital mammography, in which the image is created, is the same one that the readout is Mammography, Linear (classical) tomography, Tomosynthesis made. The advantage of such systems is that they are thin in the Further Reading: International Atomic Energy Agency. 2014. third dimension. D Diagnostic Radiology Physics: A Handbook for Teachers and Simplifying the way flat panel detectors operate, the TFT of Students, Vienna, Austria. each pixels has a switch that allows the readout of the pixels in a line-by-line scanning manner, after the signal from each pixel Digital detector array has been amplified using an amplifier and digitised, utilising an (Diagnostic Radiology) One of the most important technologies analogue-to-digital converter (ADC) – Figure D.42 that has allowed the development of digital medical x-ray appli- Related Articles: Computed radiography, Flat panel detector, cations is the flat panel active matrix array, which was originally Digital detectors, Matrix array, TFT developed for laptop computer displays. The term ‘digital detec- Further Readings: Cowen, A. R., Davies, A. G., Sivananthan, tor array’ is often used instead of ‘flat panel detector’. The under- M. U. 2008. The design and imaging characteristics of dynamic, lying technology utilised is a large area integrated circuit called solid-state, flat-panel x-ray image detectors for digital fluoroscopy an active matrix array (because they include an active switching and fluorography. Clin. Radiol. 63(10):1073–1085; International device, the thin film transistor [TFT]), made of millions of identi- Atomic Energy Agency. 2014.

Diagnostic Radiology Physics: A cal semiconductor elements deposited on a substrate material. Handbook for Teachers and Students, Vienna, Austria. All flat panel detectors have different layers, responsible all steps from the capture of the initial x-ray photons to the genera- Digital detector with direct conversion tion of electrical charges and subsequent storage and readout. (Diagnostic Radiology) See Detector FIGURE D.42 Flat panel detector. Digital detector with indirect conversion 266 D igital mammography Digital detector with indirect conversion Digital image (Diagnostic Radiology) See Detector (Diagnostic Radiology) Digital image is the display of a two- dimensional array of binary data stored in a computer. Each Digital display single data in the array is named picture element, or pixel, and (Diagnostic Radiology) In diagnostic radiology digital displays its value is correlated to the image representation in the output are used to observe radiological images. Within a hospital digi- device (e.g. monitor, printer). Usually the digital image derives tal image information is stored via the picture archiving and from a sampling process in which a continuous analogue signal is communication system (PACS) and the digital display is the converted into a digital one. users’ interface to these stored medical images. Although older There are two main parameters of the sampling process: PACS systems may use analogue cathode ray tube displays (CRTs), at present newer systems use digital liquid crystal dis- 1. The sampling frequency (or sampling rate), which rep- plays (LCDs). resents the interval between a sample and the next one. Medical displays are classified as either primary displays, used In the case of digital image this is defined as number of for initial diagnosis of medical images, or secondary displays samples per distance unit. D which are used for reviewing images alongside a radiologists 2. The sampling bit depth, which represent the number report. Display devices are categorised by Spatial resolution, con- of bits reserved for storing each sample, so that it is trast resolution, screen size, greyscale bit depth and pixel defects related to the maximum amplitude of the sampling data classification. Many different bodies have published recommen- (dynamic range). dations for medical displays, as an example the UK Royal College of Radiologists (2008) guidelines: screen resolution of ≥1280 × In digital image the sampling frequency determines the pixel 1024 native pixel array; a screen size of between 42 and 50 cm; a dimension (spatial resolution) while the sampling depth influ- maximum luminance of between 170 and 500 cd/m2; a luminance ences the range of the pixel colour levels displayed (contrast). contrast ratio 250:1 to 500:1; and a greyscale bit depth of between In medical digital images the pixel level can represent different 8 and 10 bit. parameters (depending on the techniques). For example, greyscale The American Association of Physicists in Medicine (AAPM) levels (as in CT and MRI), false colour scale (as in ultrasound and task group 18 has published guidelines for the assessment of dis- nuclear medicine images), etc. play performance for medical imaging systems. These guidelines recommend that the responsibility of display quality control falls Digital imaging and communication in medicine (DICOM) upon the medical physics staff within the hospital, and gives (Nuclear Medicine) DICOM is an international standard for han- descriptions of the suggested tests performed. The areas tested dling, storing, printing and transferring medical image data and include reflection, geometric distortion, luminance, the spa- patient information. The standard defines which file format and net- tial and angular dependencies of luminance (or viewing angle), work communication protocol to use. Ideally with DICOM; work- resolution, signal to noise ratio, glare, chromaticity and display stations, printers, servers, etc. will be able to communicate with artefacts. each other and medical imaging hardware using a joint archiving Related Articles: Radiology information system (RIS), LCD and communication system. DICOM data files are divided into (liquid crystal display), Viewing angle, Artefact, Image artefact, data sets. Each data set in DICOM data has the patient information Picture archiving and communication system (PACS) included as a part of the data set and not as a separate header file. Further Readings: Samei, E. et al. 2005. Assessment of dis- Each dataset usually contains some attribute with pixel informa- play performance for medical imaging systems: Executive sum- tion, i.e. an image or a series of images and the patient information mary of AAPM TG18 report. Med. Phys. 32(4):1205–1225; The can therefore never be separated from the picture. This is to ensure Royal College of Radiologists. 2008. Picture archiving and com- that the images are always attributed to the correct patient. munication systems (PACS) and guidelines on diagnostic display Related Article: Picture archiving and communication sys- devices, http: / /www .rcr. ac .uk /docs /radi ology /pdf/ IT _gu id anc e tems (PACS) _PAC SA Apr08 .p df (accessed 31 July). Digital industrial radiology (Diagnostic Radiology) Some of the newly developed digital Digital fluoroscopy imaging methods have been introduced in non-destructive test- (Diagnostic Radiology) Digital fluoroscopy is the method used ing (NDT) alongside film-based industrial radiography. These are in almost all contemporary x-ray fluoroscopic systems. It digi- different methods of digital industrial radiology. Although the tised the image from the output of the Image Intensifier, either methods are somewhat similar to those in medical imaging, the directly (through a CCD camera), or indirectly (through TV cam- requirements and standards are very different for digital indus- era and ADC). The digitised image is then stored in the system trial radiology. RAM memory. The final mage resolution depends on the image Related Article: Industrial radiography intensifier active input screen and the matrix size of the memory (e.g. 4096 × 4096 pixels). This memory can perform last image Digital mammography hold, and sometimes forms of digital subtraction (see article on (Diagnostic Radiology) Digital mammography is a specialised Digital subtraction angiography (DSA). Recording of the digital imaging technique for imaging breast tissue that uses a digital image (digital fluorography, recorded on hard disk or other digital detector instead of the traditional screen-film systems. As noted memory) can be made with rates from 6 to 60 frames per second in the article on Mammography the technique is mostly used for and more. Systems with such fast memory replace conventional imaging breast cancer but is also used for other breast diseases. cine fluorographic systems. Digital fluoroscopy applies various Mammography requires specialised imaging techniques because methods of image processing. it requires high contrast to separate normal breast tissue from car- Related Article: Digital subtraction angiography (DSA) cinoma. These two tissues have very similar linear attenuation Digital radiography 267 Digital reconstructed radiographs (DRR) coefficients. In addition it requires high resolution to visualise radiographic receptor with a digital receptor. It has advantages microcalcifications, which are small calcium specks that are an in both a practical and engineering level. Digital receptors can early diagnostic sign of breast cancer. be used with both fixed and mobile radiographic units. From a Scientific studies have shown that digital systems outperform practical point of view, the digital radiographic receptor improves screen film in women who are pre-menopausal or have dense both the efficiency and safety of a radiographic examination. With breasts. conventional screen film and computed radiographic techniques Several types of digital detectors have been adapted to mam- the technologist (radiographer) must leave the patient to process mographic units. These include indirect detectors and selenium the images. This increases the time required for the examination detectors. In addition specialised computed radiographic systems and because the patient sometimes may be left alone, this can have been developed for mammography lead to unsafe practices. In the digital radiography environment High Contrast: Two achieve high contrast the imaging is done the images are immediately available to the technologist within at low kVp. However since the image processing that is available the imaging room. This improves both the efficiency and safety in digital systems allows an increase in contrast somewhat higher of the examination. kVs are used. Most of the contrast is generated using the photo- Two general classes of digital radiographic receptors are avail- electric effect. While the traditional molybdenum anode is still able – direct and indirect. In the direct type receptor the radia- D widely used in digital systems, tungsten anodes with filters such tion interacts directly with the detector. Selenium detectors are an as silver with a higher k-edge are becoming popular. This type example of the direct type. In the indirect type the radiation inter- of x-ray tube has low output, so a 65 cm source-to-skin distance acts with a scintillator and the light from the scintillator is con- (SSD) is usually used. verted to an electrical signal in a TFT array. Digital radiographic High Resolution: In order to achieve high resolution a small detectors based on caesium iodide are an example of the indi- focal spot is used. This is about 0.3 mm. Digital systems are not rect type. In most cases, the use of a digital radiographic detector able to achieve the spatial resolution that a screen-film system can. requires replacement of the entire radiographic unit. A resolution of less than 8 lp/mm is typical. However because of Digital radiographic units usually have a better signal-to-noise image processing and improved contrast, the micro-calcifications ratio than screen film systems or CR. The resolution lies between are well visualised. screen-film and CR. The dynamic range of digital radiography is The Mammography Unit: The unit has a short source to much better than screen film and similar to CR. This improved image distance (SID) because of the lower output. The x-ray dynamic range decreases the probability of an incorrect exposure. tube is tilted to improve resolution. The tube is mounted so that However, it also leads to the possibility of high doses being used the central ray passes through the chest wall edge of the breast (dose creep). Consequently doses should be monitored on a regu- instead of the centre of the receptor. This improves visualisation lar basis. Because digital radiography can increase the number of of all of the breast tissue. The unit has a grid of about 5/1. The unit patients that can be done each day in a radiographic room, addi- is equipped with a breast compression system. Compression has tional lead shielding is sometimes required. The absence of the a number of benefits: many quality control issues that are associated with screen film processing is a major advantage for digital radiography. However, • It immobilises the breast. This is important since digital radiography units have substantially increased cost, as exposures of greater than 2 s are not unusual. compared to a conventional unit. Digital radiographic units are • It spreads the tissues to improve diagnostic accuracy. slowly replacing film screening and computed radiographic units • It reduces breast thickness which reduces Compton for both fixed and mobile applications. scatter. Related Articles: Computed radiography, Flat panel detector • It makes the breast thickness more uniform. This reduces the dynamic range of the x-ray signal and there- Digital reconstructed radiographs (DRR) fore allows for higher contrast film. (Radiotherapy) The planar simulation x-ray film provides a beam’s eye view (BEV) of the treatment portal but does not pro- The units also have a system for magnification imaging which is vide 3D information about anatomical structures. CT, on the other done without a grid and a 0.15 mm focal spot. hand, provides anatomical information and target definition but Mammography units are so specialised that they have a vari- does not allow a direct correlation with the treatment portals. ety of specialised quality assurance techniques which are neces- A digitally reconstructed radiograph DRR is the digital equiv- sary to achieve constancy of image quality. One of the earliest alent of a planar simulation x-ray film and can be reconstructed and most well-known was designed by the American College from a CT data set using virtual simulation software available of Radiology. This has been adapted to the digital environment. on a CT simulator or a TPS. It represents a computed radiograph Most countries have a mandatory quality assurance programme. of a virtual patient generated from a CT data set representing In addition to the image quality improvements of digital the actual patient. Just like a conventional radiograph, the DRR mammography another important aspect is the elimination of accounts for the divergence of the beam. the many film processing problems that arise in

the screen film The basic approach to producing a DRR involves several steps: environment. choice of virtual source position; definition of image plane; ray Most experts believe that screening mammography is an tracing from virtual source to image plane; determination of the important factor in reducing the mortality from breast cancer. CT value for each volume element traversed by the ray line to gen- Related Article: Mammography erate an effective transmission value at each pixel on the image plane; summation of CT values along the ray line (line integra- Digital radiography tion); and greyscale mapping. An extension of the DRR approach (Diagnostic Radiology) Digital radiography is a technique that is the digitally composited radiograph (DCR), which provides replaces the conventional screen–film receptor or computed an enhanced visualisation of bony landmarks and soft tissue Digital subtraction angiography (DSA) 268 Digital subtraction angiography (DSA) structures. This is achieved by differentially weighting ranges of format. The mask image is stored in one of the matrix RAM CT numbers that correspond to different tissues to be enhanced or memories (M1), and the image with contrast is stored in the suppressed in the resulting DCR images. other matrix RAM memory (M2). Each pixel from each mem- Abbreviation: DRR = Digital reconstructed radiograph. ory should correspond to exactly the same anatomical region Related Article: CT simulator (that is why it is imperative that the patient be absolutely still Further Reading: Podgorsak, E. B. 2003. Review of Radiation during the examination). After this both images go to the sub- Oncology Physics: A Handbook for Teachers and Students, tractor, where the content of M2 is inverted and one image is International Atomic Energy Agency, Vienna, Austria. subtracted from the other pixel by pixel. The resultant sub- tracted image can be processed by several methods to increase Digital subtraction angiography (DSA) the contrast and improve the SNR. Finally the subtracted (Diagnostic Radiology) Digital subtraction angiography is an image passes through a digital to analogue converter (DAC) x-ray imaging procedure used to increase the visible contrast of and is displayed on a monitor. blood vessels. The digital method uses the idea of the classical D subtraction angiography, where two images of the same anatomi- cal region are used – one without contrast media (called the mask), the other one with contrast media (called the contrast image). The Imlnt subtraction is achieved when the second image is inverted (black- TV ADC white) and is superimposed exactly over the mask. The resultant image contains only the image of the difference between the two images (the vessels filled with contrast), as all other anatomical structures remain almost invisible (the result of the superimposi- tion of identical, but inversed structures – i.e. image subtraction). The fact that the anatomical structures (e.g. ribs or skull) have M1 been removed from the image increases the visibility of the blood vessels (which would otherwise be superimposed with the image ∫ Subtractor of the ribs or skull). Figure D.43 shows two subtracted images – brain blood vessels on right and kidney vessels on left. The image on the left is more complex (called functional subtraction image), M2 as it shows with colour the time when the contrast media reaches specific anatomical region. Classical subtraction (superimposition of the two x-ray films – mask film and inverted contrast film) is not used anymore. However digital fluoroscopic and radiographic x-ray systems produce excellent digital subtraction images. As these images are Image DAC predominantly used in angiography, the method is known as DSA. proc. Figure D.44 shows a block diagram of a typical DSA sys- tem. The images from the Image receptor (image Intensifier with TV tube or flat panel detector) are stored in memory. FIGURE D.44 Typical block diagram of a DSA system with image The analogue image from the II tube must be digitised but intensifier and TV tube (the resultant image is DSA of the arteries of both the image from the flat panel detector is acquired in a digital legs). FIGURE D.43 Two subtracted angiographic x-ray images – brain blood vessels on right and kidney vessels on left. Digtial-to-analogue converter (DAC) 269 Diode The subtraction process can use various algorithms – linear, quadratic, logarithmic, etc. The logarithmic subtraction produces images with better contrast, as it amplifies more structures with less contrast, than images with high contrast. These algorithms influence directly the SNR. However they change the pixel values and these can only bear relative information. This information can be used as an indicator of the contrast media bolus in different time frames (functional imaging). Additionally DSA allows other digital measurements as percentage of stenosis (narrowing of the blood vessels), cardiac ejection fraction, etc. The most common artefact in DSA is movement artefact, as it leads to displacement of both images prior subtraction and con- secutive misregistration. Related Articles: Digital fluoroscopy, Mask mode fluoroscopy FIGURE D.45 Male DVI-D connector. Further Reading: Dowsett, D. J., P. A. Kenny and R. E. D Johnston. 1998. The Physics of Diagnostic Imaging, Chapmann & Hall Medical, London, UK. from the dilution of a scintillation solution with sample. The scin- tillation mixture is prepared by dissolving a primary fluor, such as Digtial-to-analogue converter (DAC) PPO (2.5-diphenyloxazole), with a secondary fluor as solvent, e.g. (General) A digital-to-analogue device converts a digital signal toluene, xylene or phenylxylylethane. The radioactive sample is to an analogue signal. Digital-to-analogue is abbreviated as DAC. added to it for counting. Liquid scintillation spectrometry is used The digital signal (usually binary) with discrete signal levels is mostly for beta radiation emitters, e.g. H-3 or C-14, with maxi- converted to a continuous analogue signal that represents a physi- mum energy of beta-particles, respectively about 18.6 and 159 cal quantity, e.g. a current, voltage or electric charge. The reverse keV. If the samples contain a single beta-isotope only then count- conversion is performed using an analogue-to-digital converter. ing (total number of counts or count rate) and not spectromet- DACs can be found in modern music players where the signal ric technique is applied for measuring radioactivity. In order to must be converted from a digital form, e.g. mp3 to an analogue determine the absolute activity of a sample we need to know the signal in order to be heard through a speaker. counting efficiency (Cef) defined as the number of recorded counts Related Articles: Analogue-to-digital converter, Analogue (Nc) to the number of disintegrations (Nd) in the sample in time t: signal æ N c ö Cef = ç t % è N ÷ ´100 d ø Digital video disc (DVD) (Nuclear Medicine) A digital versatile disk (DVD) is an optical The dilution changes the energy transfer from beta-particles to disc storage media format. Compared to a regular compact disc phosphor molecules, resulting in reduced light emission. This (CD), a DVD can store more than six times the data. A DVD- reduction is greater for lower energy beta-particles such as for ROM is a read only DVD, DVD-R and DVD+R can be written H-3. To minimise this effect it is necessary to prepare an adequate once and function like a DVD-ROM. On a DVD-RAM, DVD-RW dispersion of sample and scintillation liquid or to determine the and DVD+RW information can be erased and written multiple counting efficiency using either an internal or external standard. times. DVDs are one of the most common storage media formats Further Readings: Knoll, G. F. 2000. Radiation Detection used today. Future formats include blu-ray discs which can store and Measurement, John Wiley & Sons, Inc, New York; Thornton, even more information than a DVD. The wavelength used to read S. T. and A. Rex. 2000. Modern Physics for Scientists and a DVD is typically 650 nm (red light) whereas blu-ray uses 405 Engineering, Saunders College Publishing, Philadelphia, PA. nm (blue light). Dimensional metrology Digital visual interface (Diagnostic Radiology) Dimensional metrology is a method of (General) The digital visual interface (DVI) is a video interface digital industrial radiology – specifically industrial x-ray com- standard for the transmission of video signals (e.g. pixel intensity puted tomography. The method includes measurement of dimen- values from a medical imaging device) to a display device (e.g. a sions and other material characteristics. The theory behind some display monitor). DVIs can carry a digital video signal (termed of these could be used in medical physics. DVI-D), an analogue signal (termed DVI-A) for a video graphics Related Article: Digital industrial radiology array (VGA) or a mixture of both (termed DVI-I). Figure D.45 The successor to DVI is high-definition multimedia interface Diode (HDMI), which is similar in function and more popular in mod- (General) Electronic component allowing electric current to flow ern consumer electronics but lacks analogue VGA compatibility. in one direction only. Diodes could be either electron-vacuum Related Articles: Video signal, Video recorder tubes or semiconductor devices (most often used today). Hyperlinks: Digital visual interface: www .drhdmi .eu /diction- Semiconductor diodes are made of silicon or germanium. ary /dvi .html; HDMI vs. DVI: www .i tpro. co .uk /moni tors/ 24842 / When these materials are doped with specific impurities they pro- hdmi -vs -d vi -wh ats -t he -be st -av -inpu t-3 duce n-type (with electrons as majority charge carriers) or p-type (with positive holes as majority charge carriers). The combination Dilution quenching of n-type and p-type semiconductors creates p-n junction. When (Radiation Protection) Dilution quenching in liquid scintillation negative voltage is applied to the n-type and positive to the p-type, spectrometry is caused by the reduction of light emission resulting the diode is forward biased and passes DC current true. If the Diode detectors 270 Dipolar coupling voltage across the p-n junction is reversed (positive voltage to the area, causing the charge carriers created by the radiation to be n-type and negative voltage to the p-type), a depletion layer is cre- sweep away into the body of the crystal. The sensitivity of the ated between both semiconductors which stops the current flow detector depends on the lifetime of the charge carriers and con- (reversed biased). Diodes are the main components used in recti- sequently on the amount of recombination centres in the crystal, fiers. The x-ray high voltage generators use special power diodes which is determined by the diode type, the doping level and the which rectify the kV anode voltage (Figure D.46). accumulated dose. The diode sensitivity decreases with the accu- There are many types of diodes. For example, photodiodes are mulated dose as the radiation induce recombination centres into made from special light-sensitive semiconductors. These are used the lattice. mainly in reversed bias and the light generates current through The effect of radiation damage represents the main limitation the depleted p–n junction. The current passing through this diode of silicon diodes. Other effects related to the detector material is proportional to the light intensity. Similarly, radiation can cre- have also to be considered. The output signal of the diode depends ate electrical current through the depleted zone of a reverse biased on the photon energy because of the higher atomic number of diode (a type of photodiode) – an effect used in semiconductor silicon (Z = 14) compared to soft tissue (Z ∼ 7) and the resultant dosimeters (Figure D.47). higher contribution to the signal from the photo-electric effect. D The diode signal is also dose rate dependent because at high dose Diode detectors rates the recombination centres will be occupied resulting in a (Radiotherapy) The irradiation of a semiconductor material can relatively lower rate of recombination. Moreover the radiation result in the creation of a hole-electron pair, provided enough damage may change the dose rate dependence of diode. energy is given to the electron to raise it into the conduction band. An increase of the diode response with the temperature has Therefore a semiconductor detector is the solid state analogue of been reported, as increasing the detector temperature causes the an ionisation chamber. Semiconductors can be used in the form of energy of the minority carriers to increase and their probability a silicon diode which is a p–n junction diode. The dosimeters are for escaping recombination also increases. The thickness and the produced by taking n-type or p-type silicon and counter-doping shape of the build-up cap will influence the angular response of the surface to produce the opposite type material. These diodes the diode. are referred to as n-Si

or p-Si dosimeters, depending upon the Diodes can be used both for in vivo measurements and those base material. N-type or p-type diodes behave differently because inside phantoms. Diodes are particularly useful for measurements their minority carriers are holes or electrons, respectively. in high dose gradient areas such as the penumbra region and in Due to its high density the sensitivity of a diode detector small field dosimetry. They are also often used for measurements exceeds that of similar gaseous detectors by factor of several of depth doses in electron beams measuring the dose distribu- tens of thousands, which implies that a point-like detector can tion directly in contrast to the ionisation distribution measured be designed with the sensitive volume of less than 1 mm3. In by -ionisation -chambers. Diodes are also widely used in routine the boundary between two regions one of p-type and another of in vivo dosimetry. To determine the diode calibration factor for n-type silicon there is a depletion layer which is free of charge in vivo dosimetry a set of correction factors has to be established carriers. When the detector is operating with zero external volt- to account for variations in diode response in situations different age a potential difference of about 0.7 V exists over this depletion from the reference condition. Dipolar coupling (Magnetic Resonance) Dipolar coupling (also called dipolar- dipolar coupling or DD-coupling) refers to the direct magnetic interactions between nuclear spins. The effect is mutual and can arise both between spins in molecules and between spins in dif- ferent molecules. The dipolar coupling effect is utilised in MR spectroscopy, in which it gives information about the molecular structure (Figure D.48). FIGURE D.46 Symbols of diode (left) and photodiode (right). j i FIGURE D.48 Dipolar coupling between spin i and spin j. Both spins FIGURE D.47 Diode characteristic. are affected mutually. Dirac d-function 271 D irect voltage (DC voltage) The term dipolar coupling is a general word, which can involve DR systems can themselves be further divided into two broad both indirect dipole–dipole coupling and direct dipole–dipole categories: coupling. The indirect coupling, which also involves coupling to orbital electrons, is better known as J-coupling. This indirect cou- • Indirect DR, where the x-ray photons are converted to pling gives rise to multiple spectral peaks in MR spectroscopy. light using a phosphor, usually caesium iodine (CsI), Related Article: J-coupling and then this light is converted into an electrical signal Further Reading: Levitt, M. H. 2008. Spin Dynamics, 2nd using photodiodes. edn., John Wiley & Sons, Chichester, UK. Dirac d-function (General) See Delta function DIR (Deformable image registration) (Radiotherapy) See Deformable image registration (DIR) • Direct DR systems, on the other hand, eliminate the stage of light photon creation and use photoconductors, D such as selenium, to produce an electrical charge. Direct current (DC) (General) Direct current (DC) is form of electrical current which flows in only one direction through an electrical circuit and is typically constant in value. Direct current is generated by batteries and can also be con- verted from AC electrical power which is the usual form of power supplied by the power companies, DC electrical theory is simple, with conductors having only one The basis of both direct and indirect DR systems is a thin film property, their resistance, which relates the DC current flowing transistor (TFT) array which amplifies the incoming signal and through them to the DC potential applied across them (Ohm’s Law): creates the final digital image. The conductor Related Articles: Computed radiography, Digital detectors Further Reading: Allisy-Roberts, P. and J. Williams. 2008. In: Farr’s Physics for Medical Imaging, 2nd edn. DC potential Direct driven across conductor (General) The term direct-drive refers to motors which apply DC current their motion without the need for any gears, levers or pulleys, or through conductor any other indirect linkage. Direct-drive electrical motors have the advantage of providing fast, accurate motion of the driven part, but usually at the Where the potential difference is measured in volts, and the disadvantage of increased cost of manufacture. -current in amperes, the resistance of a conductor is given by The great benefit of few moving parts means a longer lifetime with less servicing need, and the lack of hysteresis, elasticity and Potential (in volts) Resistance R (in ohms) = backlash in any ‘drive chain’ gives greater precision. Current (in amps) Where slow movement/high torque is required, a direct drive V motor will typically be considerably larger. = I Direct voltage (DC voltage) (General) Direct voltage (DC voltage) is a measure of the differ- These units also define the power or electrical energy dissipated ence in electrical potential between any two points, and assumes in conductors as such potentials are constant. It is practically used in measuring the potential difference between points in electrical circuits, Power (in watts) = Potential (in volts)*Current (in amps) when, by using DC electrical theory, the DC through conductors, and the function of electrical circuits can be deduced. Related Articles: Alternating voltage, Alternating current, Direct voltage sources include batteries and ‘DC power sup- Direct voltage plies’ which convert AC electrical power from the domestic elec- trical supply into DC. Direct digital radiography DC electrical theory only holds for CONSTANT on non-vary- (Diagnostic Radiology) Traditional screen film imaging (also ing electrical potentials and currents. (A more complex AC theory sometimes called ‘analogue’) is being quickly replaced by digital is needed to take account of additional phenomena that can occur technologies. In digital technologies the image is captured as a where potentials vary.) signal on a digital detector, rather than OD changes on a film. In DC electrical theory, Ohm’s law states that DC current These digital technologies can be divided into two broad cat- through a conductor is proportional to the DC potential across it; egories: computed radiography (CR) and digital radiography (DR) and that each conductor has only that one electrical property, the systems. In CR systems, the image is stored on a cassette/plate ratio of potential to current, its resistance: and transferred to a readout system where the digital image is produced. In DR systems, the image is produced directly without Potential (volts) Resistance (ohms) = the need for a separate readout process. Current (amps) Dirty radionuclides 272 Discriminator Thevenin’s laws add to DC theory by noting that all the DC voltages where the harmonics ak (which represent the amplitude and phase measured around any closed pathway of conductors will add up to of the different frequency components) can be computed as zero, whilst at any point in an electrical circuit, the sum of DC cur- follows: rents flowing into that point equal the sum of currents flowing out of it. These combine to make the analysis of electrical circuits practical. N -1 Related Articles: Alternating voltage, Alternating current, a = å -(2pi /N )kn k xne Direct current n=0 Dirty radionuclides This representation means that the original sampled signal is fully (Nuclear Medicine) Dirty radionuclides (also referred to as ‘dirty represented by a maximum of N frequencies, i.e. their frequency emitters’) are radionuclides which emit certain particles or gamma response repeats indefinitely in frequency. This ambiguity leads photons that are not suitable for imaging or radionuclide therapy. to the aliasing that is often apparent in discrete sampling. One example of a dirty radionuclide is the positron emitter The DFT has equivalent properties to the continuous Fourier 86Y. PET imaging with this radionuclide can be used to quantify D transform – i.e. linearity, shift theorem, differentiation, integra- the uptake of the therapy radionuclide 90Y. (90Y is a pure beta tion and convolution. emitter and therefore cannot be used for quantitative imaging). The DFT can be computed efficiently using a FFT algorithm. 86Y is considered a dirty radionuclide because it produces 300% FFT algorithms are often more efficient for discrete signals of gamma photons per disintegration (β+ yield is 33%). Imaging the length 2N, and zero-padding can be used to increase the signal 86Y distribution using a PET scanner will give rise to a number of length for greater efficiency. unwanted spurious coincidences between one of the annihilation Applications of the DFT include data compression and MRI photons and one of the gamma photons. Spurious coincidences image reconstruction (k-space). degrade the spatial resolution by misplacing the line of response For more information please see the following external link. (see separate article on Spurious coincidences). Abbreviations: DFT = Discrete Fourier transform and FFT = Related Article: Spurious coincidences Fast Fourier transform. Further Reading: Walrand, S., F. Jamar, I. Mathieu, J. Camps, Related Articles: Fourier transform, Fast Fourier transform, M. Lonneux, M. Sibomana, D. Labar, C. Michel and S. Pauwels. Lossy, k-space 2003. Quantitation in PET using isotopes emitting prompt sin- Hyperlink: http: / /mat hworl d .wol fram. com /D iscre teFou rier gle gammas: Application to yttrium-86. Eur. J. Nucl. Med. Mol. Transform .ht ml Imaging 30(3): 354–361. Discrete wavelet transform Discharge current (General) The most common data analysis transforms are the (General) A discharge current is current caused by a release of Fourier transform and the Laplace transform. The discrete wave- stored electric energy or charge from a capacitor, battery or other let transform is another transform which uses wavelets as ‘base components that store electrical energy. functions’. A typical wavelet may look like a wave that starts with zero Discrete cosine transform (DCT) amplitude, and then grows through to a maximum and then ebbs (General) The most common data analysis transforms are the away back to zero once again. The wavelets used in DWT are Fourier transform and the Laplace transform. The discrete cosine scaled and translated copies of a base wavelet (often denoted the transform (DCT) is closely related to the discrete Fourier trans- daughter and mother wavelets, respectively). form (DFT), but only uses the cosines (the real part) as the basis Wavelets are bounded in time as well as frequency, and this functions. means that they can allow a more accurate description of both Since it only uses the real cosines as a basis, it is designed frequency and time content, especially for non-periodic or non- for ‘real’ data only, with even symmetry, resulting in only ‘real’ stationary signals with discontinuities. values in the frequency domain. It is used especially for image Related Article: Discrete Fourier transform compression (JPEG), where images are analysed for their cosine frequency content, and only data with significant amplitude are Discretisation stored. This is a lossy process, but significant storage savings can (General) Discretisation is the process involved when signals or be made without visually degrading the image. images are sampled and converted into digital form. Abbreviations: DCT = Discrete cosine transform and JPEG = In time varying signals, discretisation occurs in both time and Joint photographic experts group. amplitude – the signal first being sampled at regular intervals, and Related Articles: Discrete Fourier transform (DFT), Lossy each sample then being ‘quantised’ by comparing the signal value against a fixed set of equal steps in amplitude of a reference – as Discrete Fourier transform (DFT) found in an analogue-to-digital converter (ADC). (General) The DFT of a series x refers to its representation in Images may likewise be discretised by using digital cameras the frequency domain. It is the discrete version of the continuous or scanners, where both have multiple small optical sensors which Fourier transform. each provide one quantised value representing the amplitude of Its mathematical formulism is as follows: A discrete sequence the underlying image. of complex numbers x0, x1, …, xn−1 can be expressed in terms of a Abbreviation: ADC = Analogue-to-digital converter. series of harmonics, size a Related Article: Analogue to digital converter k N -1 1 Discriminator x = å 2pi / )k n ake ( N n (Nuclear Medicine) A discriminator is an electrical device which N k=0 discriminates signals below or above a set threshold value. When Disease-free survival 273 Distal Edge Tracking the input signal exceeds the threshold, the discriminator gener- or quantitatively with, for example, a light meter. Example test ates an output pulse. patterns are described in the report of AAPM Task Group 270 and include TG18-QC, used for qualitative assessment of display Disease-free survival performance, and the TG270-ULN series, consisting of images of (Radiotherapy) The time between (a) treatment aimed at cur- varying grey levels, allowing uniformity and luminance response

ing cancer and (b) signs that the cancer has returned, or death. to be quantitatively assessed. Timepoints other than treatment, e.g. diagnosis or trial randomi- The requirements of a display device will depend on its sation point, may also be considered. intended function. For example, displays intended for the use Related Articles: Clinical trial endpoints in the production of a legally binding radiology report (termed ‘primary diagnostics’) must fulfil more stringent quality control Disintegration, radioactive requirements than displays used as part of a wider clinical assess- (Radiation Protection) Disintegration is the transformation of one ment routine. Furthermore, monitors used in primary diagnostics radionuclide into another by radioactive decay. will have performance requirements dependent on the imaging Related Articles: Radioactive decay, Radionuclide modality from which images are assessed. For example, displays used in the reporting of mammography studies have more strin- D gent image quality requirements than those used to report CT Displacement studies. (Ultrasound) When an ultrasound wave propagates in a medium, Related Articles: DICOM (Digital imaging and communi- the particles within the medium will start oscillating. The cations in medicine), Image display, Grayscale standard display displacement is the particle’s distance, at a certain time, from function, Medical Image Display its rest position. A transducer surface moving in a sine shaped Further Readings: AAPM Task Group 270. 2019. Display movement will give rise to a sinusoidal pressure wave. The Quality Assurance, American Association of Physicists in relations between displacement, particle velocity and pressure for Medicine; The Royal College of Radiologists. 2019. Picture such a wave (expressed in one dimension) are Archiving and Communication Systems (PACS) and Guidelines on Diagnostic Display Devices, 3rd edn. Displacement:s = s0cos(wt - kx) Disposal ¶s Particle velocity: v = = -ws0sin(wt - kx) (General) The last phase in the device lifecycle. The possibility of ¶t recycling as much material as possible should be thoroughly con- sidered at the procurement stage. The management of used medi- Pressure : P = Zv = -Zws0sin(wt - kx) cal devices and equipment includes choices about incineration, waste treatment for removing possible risk of infection before If the pressure amplitude p is known, the displacement amplitude recycling or before destruction. can be calculated as In Europe the European Waste Catalogue (EWC) is available, which helps in coding the types of waste. Related Articles: Bidding process, Procurement, Tendering p s0 = with p = 1MPa, Further Readings: Iadanza, E. 2019. Clinical Engineering (Zw) Handbook, 2nd edn., Academic Press, Elsevier, ISBN: Z = .6´106 Rayls, f = 10MHz 9780128134672; Miniati, R., E. Iadanza and F. Dori. 2016. Clinical Engineering (From Devices to Systems), Academic 106 Press, Elsevier, ISBN 9780128037676; Willson, K., K. Ison and s0 = » 0 1 m S. Tabakov. 2014. Medical Equipment Management, Taylor & (1 6´106 ´ 2 ´106 . m . p ) Francis, ISBN: 9780429130373. Hyperlink: EWC: https :/ /ec .euro pa .eu /envi ronme nt /wa ste /f Related Article: Longitudinal wave ramew o rk /l ist .h tm Display quality control Distal (General) The quality control of devices used in the presentation (General) Directional anatomical terms describe the relationship of medical images is a crucial part of the quality assurance of of structures relative to other structures or locations in the body. the entire imaging workflow. Such quality control ensures dis- Distal: Away from, farther from the origin (e.g. the hand is play devices fully utilise the information provided by an imaging located at the distal end of the forearm) modality, ensuring accurate image interpretation. Related Article: Anatomical relationships The characteristics assessed in display quality control include display luminance (i.e. brightness), with ambient lighting in the environment surrounding the display and the luminance of dis- Distal Edge Tracking played images both assessed. Luminance limits are assessed (Radiotherapy) Distal edge tracking was an early method for alongside the performance at all luminance levels, ensuring that delivering intensity modulated proton therapy. As shown in the device, if required, adheres to the DICOM Grayscale Display Figure D.49, the technique tracks the target distal edge with Function (GSDF). Additional tests include the identification of monoenergetic pencil beams to deliver the dose to the whole tar- any display artefacts, colour performance, uniformity, noise, tem- get volume. A uniform dose distribution is constructed by using a poral performance and spatial resolution. number of fields from multiple directions. Testing is completed using ‘test patterns’, images that are The method aims to achieve the greatest possible ratio between displayed on the device and assessed either qualitatively by eye the delivered target dose and the integral dose to the surrounding Distance learning 274 D ivergent beam edge Distribution volume (Nuclear Medicine) In a closed compartment (the tracer cannot escape the compartment) the activity A is the product of the con- centration of tracer in the compartment C and the distribution volume Vd. This is true in a steady state system. The distribution volume in such a system is described by A Vd = (D.10) C This is a suitable method when determining the distribution vol- ume for a closed compartment. Most compartments are in fact open, which means that the tracer can escape. The distribution volume can still be estimated by using the partition coefficient λ. D Related Articles: Tracers, Analogue tracers, Partition coefficient Further Reading: Cherry, S. R., J. A. Sorenson and M. E. Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA, pp. 380–381. FIGURE D.49 A single field delivering dose to the target (grey volume) Dithering via the distal edge tracking (DET) technique. The Bragg peaks of each (Diagnostic Radiology) The term dithering (also called halfton- pencil beam are shown as red circles with green – red gradient showing ing) is related to inkjet printing. This is a method of creating the dose gradient of each beam in the field. colour (or grey) nuances by varying the number of ink drops per area (effectively creating a mix of the ink colours and the white background page). For example, an 8 × 8 dither matrix in grey healthy tissue. However, the method struggles to obtain sufficient halftones will create 64 grey nuances (from 1 ink drop in the area dose uniformity in the target, is at risk of poor robustness due – minimal grey, to 64 drops in the area – maximal grey). to range uncertainty, and places high LET regions in abutting The method is based on the limited resolution of our eyes healthy tissue. For these reasons, it is not currently used to treat (inability to distinguish closely spaced spots). The dithering uses patients in proton therapy. special algorithms producing ink drops patterns (with four, six or Further Reading: Lomax, A. J. 1999. Intensity modulation eight different inks). The result is more colours of the printout (or methods for proton radiotherapy. Phys Med Biol. grey nuances in B/W printing). Good photo printers produce lit- tle, if any, visible dither pattern in the light-coloured areas (where Distance learning they are most noticeable). (General) Distance learning is remotely obtained education. In general this method reduces the spatial resolution (the dither These days distance learning is mainly delivered through Internet area being related to the smallest visible detail). For example, a (in this way transforming itself to e-learning). Among the many 1440 dpi printer using 8 × 8 dither matrix will produce print- definitions for this teaching process is ‘e-learning – the deliv- out 1440/8 = 180 dpi (effective dots per inch), what will produce ery of content via all electronic media, including the internet, approx. 3.5 lp/mm resolution. The method is not anymore used in intranets, extranets, satellite, broadcast, video, interactive TV medical imaging. and CD ROM. E-learning encompasses all learning undertaken, whether formal or informal, through electronic delivery’. Divergence Medical physics is one of the pioneering professions in (Ultrasound) The diffraction pattern of a plane disc source can e-Learning. The training materials EMERALD and EMIT are be divided into a complex near field (near zone or Fresnel zone) typical examples of e-Learning materials, now used in more than that has a cylindrical shape with approximately the same radius 60 countries world-wide. A number of ways to deliver distance as the source, and the far field (far zone or Frauenhofer zone) that learning and e-learning are listed in the following references. diverges with the angle Θ = arcsin(0.61λ/a) where a is the radius Further Reading: Tabakov, S. 2005. Editorial, Special issue of the disc and λ is the wavelength, Figure D.33. The far field on e-Learning in Medical Engineering and Physics. J. Med. Eng. starts at the distance a2/λ from the source. This implies that a Phys. 27(7):543–547. large disc compared to the used wavelength produce a long near Hyperlink: www .emerald2 .eu; www .ltsn .ac .uk field with less divergence angle. Outside the main lobe, at angles wider than Θ, side lobes can Distribution function method be found, Figure D.34. (General) The cumulative probability distribution function Related Articles: Diffraction, Side lobes cpdf(x) describes the probability that a variable y takes on a value less than or equal to a value x. It is constructed from the integral of the probability density function pdf(x) over the interval [−∞, x]: Divergent beam edge (Radiotherapy) Since the radiation used in radiotherapy origi- cpdf (X) = P(y £ X) nates from an effective focal spot the radiation field will diverge along the beam axis. The geometry of this is set by things like X the size and the position of the collimators relative to the focal = ò pdf (X¢)dX¢ spot. The field size is determined at a reference distance from the -¥ focus, e.g. 100 cm focus to axis (centre of rotation or iso-centre) Diverging collimator 275 Dixon Method Focus f 100 cm D(100) D(x + 100) 100 + x cm l b FIGURE D.50 Schematic to show the divergent beam edge. D distance. It will diverge as the distance increases from the focus. θ The length or width (D) of the field will change as in Equation D.11 (see Figure D.50): FIGURE D.51 A schematic overview of a diverging collimator. D (100)(100 + x) D(x+100) = (D.11) (100) The best spatial resolution is acquired when the source is located as close as possible to the collimator face. The geometric efficiency of a diverging collimator can be Diverging collimator expressed as follows: (Nuclear Medicine) Collimators with holes that diverge from the detector face are called diverging collimators. The diverg- 2 2 ing point for such collimators is typically located 40–50 cm 2 æ d ö é d ù é f + l ù g » K çç ÷÷ ´ ê ú ´ (D.1 ) behind the back of the collimator. These collimators project ¢ ( )2 ê ú 4 è leff ø ë d + t û ë f + l + b û a minified non-inverted image on the detector. If the activity distribution extends outside the detector area, a regular paral- where K is a constant depending on the hole shape. The geo- lel-hole collimator would fail to image the entire distribution metric efficiency is defined as the fraction of incident photons in a single view. A diverging collimator allows photons that registered by the detector and it decreases with the distance from originate from an event localised outside the area beneath the the face of the collimator. The collimator is useful in projecting collimator to enter; hence a diverging collimator is useful to large objects onto the detector but at the cost of spatial resolution image an activity distribution greater than the field-of-view of and efficiency. the detector. The relationship between the image size, I, and Related Articles: SPECT, Parallel-hole collimator, Diverging object size, O, depends on the collimator front to convergence collimator, Converging collimator, Collimator, Collimator point distance f, the collimator to source distance b and the col- design, Collimator parameters limator thickness t. The quotient between I and O is called the Further Reading: Cherry, S. R., J. A. Sorenson and M. E. minification factor: Phelps. 2003. Physics in Nuclear Medicine, 3rd edn., Saunders, Philadelphia, PA, pp. 245–246. I ( f - t) = (D.12) O ( f + b) Dixon Method A conclusion drawn from this equation is that an increase in the (Magnetic Resonance) The Dixon method or the Dixon tech- collimator-to-source distance results in a larger minification. nique is a method

to separate MRI signals from fat and water. Differences in object depth result in differences in the minifica- The method takes advantage of the slightly different proton reso- tions which may lead to severe image distortion. nance frequencies of fat and water. Since protons in water and The collimator’s contribution to the system resolution is fat have different resonance frequencies, they will be in-phase with each other at some times after excitation and out-of-phase é d(le¢ff + b) ù é 1 ù é l¢ ù with each other at other times. By setting a specific echo time R + eff coll » ê ú ´ ë le¢ff û ëê q 2 f ú ûú ´ 1 (TE) where the signal from fat and water are in-phase, the signal cos ê ë û (D.13) from fat is added to the water signal to produce the MR signal. A second echo time can be selected where the fat and water are ( - 2m-1) l¢ l¢ l eff eff » = out-of-phase and the signal from fat is subtracted from the water cosq cosq signal. By performing arithmetic operations on the images on a where pixel by pixel basis, a ‘water only’ and a ‘fat-only’ image can be d is the hole diameter acquired. This method is called a two-point Dixon method as two le¢ff is the effective hole length, accounting for septal images are acquired. The technique was extended to a three-point penetration method which accounts for magnetic field inhomogeneity, and θ is the angle between the central axis of the collimator and an this three-point method is the technique implemented on most off centre source as seen in Figure D.51. current MRI scanners. The Dixon method is used extensively as DKI (Diffusion kurtosis imaging) 276 Doppler effect material (e.g. NaI or CsI) with dopant such as Tl, Na, Tb, etc. The resulting crystals in this case will be NaI:Tl, CsI:Na, etc. Doppler angle (Ultrasound) The term Doppler angle is one used to describe the angle between the ultrasound beam and direction of flow in the vessel under examination. The Doppler frequency is proportional to the cosine of the Doppler angle, denoted by θ as shown in the Doppler equation: 2V f co q f = t s d c D where V is the flow velocity ft is the transducer frequency c is the sound velocity in blood FIGURE D.52 Illustration of Dixon method. The angle is also referred to as the beam/flow angle or angle of insonation. The lack of a defined term for this angle and the use a fat suppression technique and as a method to quantify liver fat of other angle terms in imaging, including beam steering angle in (Figure D.52). linear arrays, can lead to considerable confusion for those learn- Related Articles: FATSAT (fat suppression), Magnetic field ing ultrasound. This is unfortunate in view of its fundamental inhomogeneity, Echo time (TE) importance in the production of good sonograms and for accurate Further Readings: Dixon, W. T. 1984. Simple proton spec- measurement of velocities from the sonogram. troscopic imaging. Radiology 153:189–194; Glover, G. 1991. At Doppler angles approaching 90° there are poor qual- Multipoint Dixon technique for water and fat proton and suscepti- ity sonograms as Doppler frequencies decrease. Low Doppler bility imaging. J. Magn. Reson. Imaging 1:521–530. angles give higher Doppler frequencies and good colour and spectral Doppler images. With high Doppler angles there are increasing errors possible in the measured velocities as a result DKI (Diffusion kurtosis imaging) of operator errors and inherent errors in Doppler ultrasound. As (Magnetic Resonance) See Diffusion kurtosis imaging (DKI) a pragmatic limit to this, many operators choose not use Doppler angles above 60° so as to avoid large errors in measured velocity DNA targeting (Figure D.53). (Nuclear Medicine) This term refers to the targeting of the cell Related Articles: Doppler equation, Pulsed wave Doppler DNA by certain radiopharmaceuticals. One example is 18F-FLT ([18F]3′-fluoro-L-3′-deocythymidine) that is an alternative to Doppler effect 18F-FDG when investigating the therapeutic response. The accu- (Ultrasound) The Doppler effect is the change in frequency and mulation rate of 18F-FLT in the cell is proportional to DNA pro- wavelength caused by relative motion between a wave source and duction, hence proportional to cell proliferation rate. A decrease an observer/detector. in DNA production is one of the first responses to tumour ther- apy so 18F-FLT accumulation acts as a good indication of tumour response. Related Articles: Tracer kinetic modelling, Receptor target- ing, Antigen targeting, Neuroreceptor targeting, Glycolysis tar- geting, Apoptosis targeting, Hypoxia targeting Further Reading: Imam, S. K. 2005. Molecular nuclear imaging: The radiopharmaceuticals (review). Cancer Biother. Radiopharm. 20(2): 163–172. Doped (Diagnostic Radiology) See Doping Doping (Diagnostic Radiology) Doping is the process of introducing impurities into pure materials. The crystal structure of doped materials contains defects which change the properties of the material. These defects are the reason for the formation of energy traps in the forbidden zone. Semiconductors of n-type and p-type are produced through the process of doping (e.g. boron is a p-type FIGURE D.53 The angle between the indicated Doppler beam and the dopant and phosphorous is a n-type dopant). The sensitivity of direction of flow is marked with an arc and is measured as 60° as shown luminescence detectors can also be improved by doping the pure in the data above the sonogram. Doppler equation 277 Doppler equation Static source, moving observer Observer velocity v Transmitted ultrasound ft Reflected ultrasound fr Speed of sound c fr = f c + ν t c Speed of sound in medium = c θ fr = f c + v cos θ . c c + ν cos θ fr = f t t c c – v cos θ c θ v D FIGURE D.54 Static source, moving observer. FIGURE D.56 The reflected ultrasound has a frequency shift as a result of reception and reflection from the moving scatterer. Static observer, moving source Source velocity v c + vcosq c fr = ft × c c - vcosq fd = fr - ft f c r = f t æ c + vcosq c c– v fd = ft ç × - ö 1 è c c - vcosq ÷ ø Speed of sound in medium = c c2 + cvcosq - c2 + cvcosq fd = ft c2 - cvcosq θ 2cvcosq f fr = f c d = ft c2 - cvcosq t c– v cos θ v << c 2 f v o f » t c sq FIGURE D.55 Static observer, moving source. d c Doppler equation The effect of a moving observer or moving source is shown in (Ultrasound) The following relation is ‘the Doppler equation’ in Figures D.54 and D.55. diagnostic ultrasound: When an observer moves towards the source of a wave trav- elling through a medium at speed c, it meets the waves at an 2vf increased frequency by virtue of its own velocity v. Conversely, Df = 0 cosq (D.15) c an observer moving away from the source detects a lower fre- 0 quency. If the direction of the observer velocity is not directly where towards the source then the received frequency is also depen- Δf is the Doppler shift frequency dent on the angle θ between velocity and the direction of the v is the velocity of the target wave as it passes the observer. If θ = 90°, then there is no f0 is the transmitted frequency Doppler effect. c0 is the speed of sound in blood If the source of the waves is moving towards the observer, θ is the angle between the transmitter/receiver combination then the wavelength in the medium is reduced and the frequency and the velocity of the particle detected by the observer is increased (or vice-versa if the source is moving away from the observer). If the direction of the observer It also assumes that the angle between the transmit and receive velocity is not directly towards the source then the received fre- beams is small. quency is also dependent on the angle θ between velocity and the The entire derivation of the Doppler equation is somewhat direction of the wave as it passes the observer. lengthy to include here, but essentially the analysis can be divided In medical ultrasound, commercial Doppler devices use the into two parts. In diagnostic ultrasound, the Doppler effect is the pulse echo technique where the source and detector are stationary result of a change in frequency due to sound being reflected off but the echo is scattered from a moving target. The ultrasound a moving target, the first part of the analysis is to observe the then undergoes a Doppler shift by virtue of the shift as sound Doppler shift when the transmitter is stationary, and the reflec- insonates the scatterer with a further shift back from the moving tor (receiver) is moving. In the second part of the analysis, sound scatterer. The effect is shown in Figure D.56. is transmitted from a moving source to a stationary receiver. By This simplifies as follows to give the familiar ultrasound combining these two cases and assuming that the target veloc- Doppler equation: ity is much smaller than the sound speed, the preceding Doppler Doppler imaging modes 278 Doppler sample volume equation is obtained. A full description with diagrams is given in the article Doppler effect. Related Article: Doppler effect Doppler imaging modes (Ultrasound) The term Doppler imaging covers several different implementations of the basic technique including • Continuous wave Doppler • Pulsed wave Doppler • Colour flow imaging • Power Doppler D See Flow imaging Related Articles: Continuous wave Doppler, Pulsed wave Doppler, Colour flow imaging, Power Doppler Doppler phantom FIGURE D.57 Sample volume of 1.5 mm in a flow phantom. The sono- gram shows high constant velocities from the centre of the flow channel. (Ultrasound) A Doppler phantom is a test object to measure and The sample volume is visible as a pair of lines parallel to the transducer assess the performance of ultrasound Doppler devices including face in the centre of the vessel. continuous wave Doppler systems and colour flow and pulsed wave spectral Doppler in ultrasound scanners. Parameters that may be tested include Spectral Doppler: • Penetration depth/sensitivity • Velocity estimation accuracy • Waveform index estimation accuracy • Volume flow estimation accuracy • Temporal resolution • Intrinsic spectral broadening • Accuracy of angle estimation Colour flow main parameters: • Lowest detectable velocity • Highest detectable velocity • Image spatial resolution and registration accuracy • Temporal resolution • Velocity resolution • Tissue colour suppression performance FIGURE D.58 Sample volume of 6 mm. The sample volume includes the entire vessel. The sonogram shows higher intensities generally and Doppler frequencies from the flow adjacent to the wall. Test objects are broadly divided into • Flow phantoms where a fluid is passed through flow In ultrasound scanners, the sample volume usually refers to the channels within a medium with acoustic characteristics length along the beam over which the operator chooses to investi- that mimic human tissue (Figure D.57) gate flow. The length of the sample volume, also referred to as the • String phantoms where a moving band or thread is range gate, can be varied in a series of specific distances, e.g. from insonated (Figure D.58) 1.5 to 10 mm. This can alter the pulse length and the time over which the echo is processed for Doppler frequency calculation. Doppler sample volume Because pulsed Doppler techniques use narrowband processing, (Ultrasound) The Doppler sample volume is the target volume the minimum sample volume is significantly larger than the mini- from which pulsed wave Doppler echoes are obtained. mum B-mode axial resolution. The sample volume is determined by In operation, the sample volume is selected to suit the appli- cation. A small sample volume can be used to investigate flow • The axial length along the pulsed Doppler beam usually in the centre of the vessel and allows for more precise investiga- displayed on the image as a pair of parallel lines (see tion of vessels where there are several vessels in close proximity. Figure D.55) For mean velocity measurements, the vessel should be uniformly • The beam width of the Doppler beam at the selected insonated. The sample volume should then be enlarged to

include depth the whole vessel, insonating flows in the centre of the vessel and • The slice thickness of the Doppler beam, usually deter- at the vessel walls. However, altering the sample volume only mined by the acoustic lens alters the axial length; the slice thickness remains undetermined Doppler shift 279 Dose area histogram by the operator and in-plane focus, usually set to the sample vol- ume depth, means the vessel under investigation may still not be insonated uniformly. Related Articles: Pulsed wave Doppler, Laminar flow Doppler shift (Ultrasound) See Doppler effect Doppler ultrasound (Ultrasound) The term Doppler ultrasound describes a range of techniques, instrumentation and devices to detect, image and measure blood flow by ultrasound. It is conventionally described as follows: when ultrasound insonates moving targets, the movement causes the reflected echoes to change frequency. The change in frequency is depen- D dent on the relative direction of ultrasound beam and blood flow, the velocity of the targets and the transmitted frequency as shown in the ultrasound Doppler equation: FIGURE D.59 Hand-held continuous wave Doppler ultrasound device 2Vft cosq fd = with different probes operating at different frequencies for specific appli- c cations. The unit gives an audio output of Doppler frequencies. where V is the velocity of the blood ft is the transmit frequency θ is the angle between flow and the beam c is the speed of sound in blood fd is the resulting Doppler frequency The velocity of blood in major arteries and veins in humans range from a few cm/s to around 6 m/s (in diseased arteries). The Doppler frequencies that result from diagnostic ultrasound trans- mitted frequencies are in the audio range. The time changing dis- tribution of Doppler frequencies can be displayed as a sonogram and, if an estimate of beam/flow angle θ is made, the sonogram shows the time varying distribution of velocities. Two basic forms of Doppler ultrasound are commonly used clinically. In continuous wave (CW) Doppler, a transducer consists of separate transmit and receive elements. The difference in frequen- cies is the Doppler frequency which is amplified, played through a loudspeaker (Figure D.59) and, in some systems, displayed as a sonogram (Figure D.60). Pulsed wave Doppler (PWD) systems have the advantage that examination of flow at a specific depth can be made and PWD is used for colour flow imaging and spectral PWD in ultrasound scanners. The processing for PWD differs from CW and, it could be argued, is not a Doppler system in the strict sense of the word, although the Doppler equation still applies. PWD systems are FIGURE D.60 Doppler ultrasound system with a spectral display of the subject to a major limitation which limits the velocity range at sonogram, which shows the time changing distribution of frequencies. greater depths. This is discussed under the relevant articles. Related Articles: Flow imaging, Continuous wave Doppler, accumulated in a patient’s normal tissues is poorly quantified. Pulsed wave Doppler, Colour flow imaging, Doppler equation, Changes in patient anatomy can occur over the course of the Doppler effect treatment, rendering the planned dose distribution inaccurate. Validated deformable image registration algorithms stand to Dose improve the accuracy of dose accumulation, for both the target (Radiation Protection) This is an ambiguous term. It is normally volume and the critical normal tissues. used when referring to absorbed dose. Related Articles: Fractionation, Deformable Image Related Article: Absorbed dose Registration (DIR) Dose accumulation Dose area histogram (Radiotherapy) In radiotherapy, dose is ‘accumulated’ fraction (Radiotherapy) 2D or 3D histograms are used in radiotherapy by fraction. Generally, the actual distribution of radiation dose treatment planning. 3D histograms, represented by dose volume Dose area product (DAP) 280 Dose distribution histograms, are used more commonly due to their greater com- Monte Carlo modelling of the absorption and scattering of ionis- prehensiveness. 2D histograms, represented by dose area histo- ing radiation travelling through the body. grams (DAH), provide 2D dose statistical information about the Dose calculations are normally performed either before or given 2D view (differential or cumulative/integral). DAH can after medical investigations involving exposure to ionising radia- show minimum, maximum and mean dose values for the selected tion (x-rays or nuclear medicine) and used to describe the risk to structure. the patient, or to an unborn child of a pregnant patient. Abbreviation: DAH = Dose area histogram. Related Articles: Radiation exposure, Radiation dose, Organ Related Article: Dose volume histogram dose, Effective dose, Equivalent dose Dose area product (DAP) Dose calibrator (Diagnostic Radiology) The dose area product is the most com- (Nuclear Medicine) A dose calibrator is used to measure the mon method of quantifying patient dose in radiology. radioactivity in a sample usually a syringe used for injection in It is the product of the air kerma at a distance from the source a patient or an experimental animal. The word dose calibrator D and the area that the x-ray beam covers measured at that same is somewhat misleading and activity or radioactivity calibrator distance. Its units are typically cGy cm2. would be more appropriate. To measure the activity accurately, In practice the DAP is measured using a DAP meter, which the device has to be calibrated. Efficiency curves must be derived consists of a parallel plate ionisation chamber, attached to the to take into account the geometry of the sample (i.e. the volume) beam exit port of the x-ray tube, and a display, usually located to get the efficacy in the measurement. A common dose calibrator near the control console (Figure D.61). consists of a well ionising chamber but can also consist of other As the exposure is taken the ionisation chamber measures detector materials. the air kerma of the beam. It then multiplies this by the area the Related Article: Activity measurement parallel plates cover to give the dose area product, which is then displayed for the user. In this way DAP meters provide a simple, Dose conformity index non-invasive and instantaneous method of assessing patient dose. (Radiotherapy) A dose conformity index (CI) is a statistic which The simplicity of DAP measurements lie in the fact that the is used to evaluate how well radiotherapy dose conforms to (or focal spot of the x-ray tube can be treated as a point source, so the is tailored to) the shape of the target volume. Dose conformity air kerma will reduce as you move away from the tube in accor- indices may be calculated in a variety of different ways. They are dance with the inverse square law. At the same time the area is typically used in treatment plan comparison. increasing with a square relationship, so the change to the dose and the area effectively cancels out, leaving the DAP measure- Dose constraints ment the same at all points along the beam’s path. Therefore the (Radiation Protection) Dose constraint is defined as a restric- DAP at the patient is the same as that measured at the tube, by the tion on the prospective dose to individuals from a defined source DAP meter. Related Articles: (Figure D.62b), for use at the planning stage in radiation protec- Stationary anode, Rotating anode, Target, tion whenever optimisation is involved. Line focus principle, Biangular anode disk, Focal spot actual, The concept of dose constraint was established by the Focal spot effective, Focal spot International Commission on Radiological Protection (ICRP) in Publication 60. It should be used, where appropriate, within the Dose calculation context of optimisation of radiological protection. For occupa- (Radiation Protection) This is the calculation of the radiation tional exposure, a dose constraint serves as an upper boundary for dose received by an individual. It may be the result of direct mea- the range of options in optimisation (Figure D.63). surement, or indirect measurement with the use of conversion Related Articles: ALARA, Dose limits factors, or by the use of physical or mathematical phantom and Dose conversion factors (Radiation Protection) Radiation dose is rarely measured directly. For example, an ionisation chamber measures exposure – the amount of ionisation caused by the radiation in a volume of air. The various dose quantities can then be derived from the mea- surement by the use of a number of different factors to convert from one quantity to another. Related Articles: Exposure, Radiation weighting factors, Tissue weighting factors, Absorbed dose, Equivalent dose, Effective dose Dose distribution (Radiotherapy) Radiation interacts with matter and in so doing releases a dose at any point of the irradiated object. The repre- sentation of the variation of dose with position in any region of an irradiated object is called dose distribution. The dose distribution may be measured in a phantom or it may be calculated by a TPS. To measure accurately a dose distribution it is necessary to avoid FIGURE D.61 Parallel plate ionisation chamber and display of a DAP that the detector disturbs the distribution and to realise a condition meter. of charged particle equilibrium. The visualisation of a calculated Dose escalation 281 Dose length product D FIGURE D.62 Difference between dose limit and dose constraint: (a) dose limit, (b) dose constraint. (Images courtesy of ICRP.) of the International Commission on Radiological Protection, ICRP Publication 103. Ann. ICRP. 37:2–4. Dose homogeneity (Radiotherapy) In most radiotherapy cases, because the distribu- tion of clonogenic tumour cells is unknown, it is convenient to aim to achieve a homogenous dose distribution within the plan- ning target volume (PTV). However, due to inherent characteris- tics of patients and of treatment beams, this goal is unachievable in a real radiotherapy plan. Therefore, in practice, some hetero- geneity of the dose distribution is present and has to be accepted. ICRU 50 (1993) recommends that this dose variation should be kept within +7% (hot spot) and −5% (cold spot) of the prescribed dose. Cold spots are of concern inside the planning target volume (PTV) and hot spots in those regions where organs at risk are FIGURE D.63 Dose constraints in planned exposure situations. located. Related Article: Hot and Cold spots Further Readings: ICRU 50. 1993. Prescribing, Reporting dose distribution serves to optimise patient planning and to check and Recording Photon Beam Therapy. International Commission the accuracy of the computer generated dose distribution. on Radiation Units and Measurements, Washington, DC. Related Articles: Equilibrium dose distribution, Charged par- ticle equilibrium, Depth dose distribution Dose length product Dose escalation (Diagnostic Radiology) The dose length product (DLP) is a dose (Radiotherapy) Dose escalation studies typically administer descriptor used in CT scanning. It is an approximate indicator of sequentially increasing doses in specific radiotherapy treat- stochastic radiation risk. The DLP is defined as the product of ments to different cohorts of study subjects, until the maximum absorbed dose, as given by the volume CT dose index, CTDIvol, desired dose is reached or a pre-specified event (e.g. level of tox- and the scan length – see the following equation and Figure D.64. icity) occurs. Dose escalation aims to improve tumour control It has units of milli-Gray centimetres: rates. DLP = CTDIvol ´ L (mGy cm), Dose equivalent (Radiation Protection) Prior to the publication of ICRP 60 where L = Scan length ‘1990 Recommendations of the International Commission on Radiological Protection’ published in 1991 Dose Equivalent was Doubling the DLP in a given anatomical area will double the sto- the parameter used to describe whole body effective dose due chastic radiation risk. Generally, stochastic radiation risk is given to a partial body irradiation. ICRP 60 renamed the parameter by the effective dose, E (milliSieverts). Conversion factors are Equivalent Dose. available to convert DLP to E (Table D.4). Related Article: Equivalent dose Related Articles: Computed tomography dose index (CTDI), Further Readings: ICRP. 1991. Recommendations of the Effective dose International Commission on Radiological Protection, ICRP Further Reading: 2004 CT Quality Criteria (MSCT 2004), Publication 60. Ann. ICRP. 21:1–3; ICRP. 2008. Recommendations online from: www .msct .eu /CT _Quality _Criteria .htm Dose limiting tissue 282 Dose model evaluation Pitch 1 Pitch 2 N rotations N rotations CTDIvol1 CTDIvol2 L1 L2 CTDI L2 = 2L1 and CTDIvol2 = vol1 2 D DLP1 =~DLP2 FIGURE D.64 Examples of calculating dose length product. radioisotopes) that give ingestion and inhalation dose coefficients. TABLE D.4 This means that it is possible to evaluate the committed effective Conversion Factors from DLP to Effective Dose, E dose per unit intake via ingestion corresponding to different gut transfer factors, for various chemical forms, and the committed

Region of Body Conversion Factor, EDLP (mSv mGy−1cm−1) effective dose per unit intake (via inhalation) for the lung absorp- Head and neck 0.0031 tion, considering also the component cleared from the lung to the Head 0.0021 gastrointestinal track. Neck 0.0059 In the case of apprentices of 16–18 years old, who are under Chest 0.014 training with activities involving exposures, the following limits Abdomen and pelvis 0.015 shall be respected: an effective dose to whole body of 6 mSv per Trunk 0.015 year; an annual equivalent dose to the lenses of eye of 50 mSv and Source: 2004 CT Quality Criteria (MSCT 2004), www .msct .eu /CT an annual equivalent dose to the extremities and skin of 150 mSv. _Quality _Criteria .htm 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; with the maximum dose per year remaining 50 mSv (circum- stances shall be reviewed when any worker reaches 100 mSv). Dose limiting tissue Alternatively, the annual limit shall not exceed 50 mSv with a (Radiotherapy) Radiation treatment inevitably affects normal tis- temporary change period not exceeding 5 years. sue, and the severity of these radiation induced effects may limit Dose Limits for Public Exposures: The averaged estimated the dose that can be safely delivered to the tumour. Usually one doses to critical groups of the public, attributable to practices healthy organ or tissue in the treatment field is more radiosensi- shall not exceed the following limits: (1) effective dose to the tive than the others; hence the patient cannot be exposed to levels whole body: 1 mSv per year, in special circumstances averaged of radiation higher than the tolerance of this dose-limiting tissue. over a period of 5 years with an effective dose of 5 mSv per year; For further details see the articles on Adverse effects, Therapeutic (2) annual equivalent dose to the lenses of the eye of 15 mSv and effect and Tolerance. (3) annual equivalent dose to the skin of 50 mSv. Related Articles: Adverse effect, Therapeutic effect, Tolerance Dose Limits for Comforters and Visitors of Patients: The dose of any voluntary comforter or visitor of patients shall be Dose limits constrained so that it will not exceed 5 mSv during the period of (Radiation Protection) The following dose limit values apply to the diagnostic investigation or the treatment. The dose to children exposure from practices, with the exception of exposures from visiting patients who have ingested radioactive material, shall be medical practices (diagnostic and therapy) and from natural limited to less than 1 mSv. sources. Further Reading: International Basic Safety Standard Dose Limits for Occupational Exposures: Occupational for Protection against Ionizing Radiation and for the Safety of exposures of any worker shall be controlled in order not to exceed Radiation Sources, Safety Series No. 115. 1996. International the following limits: (1) effective dose to the whole body: 20 mSv Atomic Energy Agency, Vienna, Austria. per year averaged over a period of 5 years with an effective dose of 50 mSv in any single year; (2) effective dose to the embryo or Dose model evaluation foetus: 1 mSv; (3) annual equivalent dose to the lenses of eye: 150 (Radiation Protection) Dose response curves such as the linear mSv and (4) annual equivalent dose to the skin and extremities: response curve, threshold dose response curve and non-linear 500 mSv. response curve, etc. are used either separately or in combination For occupational exposure from radionuclides, with risk as models to describe the response of the human body to expo- of internal contamination, there are tables (for the various sure to various types on ionising radiation from both internal and Dose modulation 283 Dose optimisation external exposure, and at high and low doses and dose rates. Such 2. Depend on the level of ambient dose equivalent and responses are dependent on the cellular responses to such differ- activity concentration, taking into account fluctuations, ent types and energies of ionising radiation. likelihood and magnitude of potential exposures These radiobiological models range from the very simple – e.g. the linear no-threshold model, which assumes that the response The programme shall specify: to exposure to ionising radiation is proportionate, and for which 1. Quantities to be measured there is no threshold – to the more complex to take into account 2. Frequency and position of measurements observed data on bystander effects and adaptive responses. In 3. Appropriate methods and procedures some circumstances the different responses may compete against 4. Reference levels and actions to be taken if they are each other. exceeded It is therefore necessary to continue to observe immediate acute and late chronic effects of radiation exposure to evaluate the If appropriate, records of the finding shall be kept and made avail- dose models in order to more appropriately apply them to specific able also to the workers. exposure situations. Health Surveillance: In accordance with the rules established For more information, see the articles on Radiobiological by the National Authority, the employers, registrants and licens- D model, Bystander effects and Adaptive responses. ees shall make arrangements for appropriate health surveillance. Related Articles: Radiobiological model, Bystander effects, Health surveillance shall be based on general principles of occu- Hormesis pational health and designed to assess the initial and continuing fitness of the workers for their respective assignments. Records: Employers, registrants and licensees shall maintain Dose modulation exposure records for each worker described earlier under indi- (Radiotherapy) A multileaf collimator permits the modulation of vidual monitoring and exposure assessment. the photon fluence by the motion of jaws and leaves and these The exposure records shall include: results in a dose modulation in the irradiated volume. 1. Information on the nature of work 2. Doses, exposures, intake values and data upon which Dose monitoring the dose assessment has been based (Radiation Protection) 3. Information on dates of employment and relative doses, Individual Monitoring and Exposure Assessment: exposures and intakes in each employment, in case of a Registrants and licensees, employers of workers as well as self- worker who is or has been occupationally exposed with employed individuals shall be responsible for the occupational more than one employer exposure of workers, with individual monitoring when appro- 4. Records of any doses, exposures or intakes due to emer- priate, and shall assure that adequate arrangements be made for gency interventions or accidents, which shall be distin- appropriate dosimetry and services. guished from those during normal work Individual monitoring shall be provided, when appropriate, to workers, normally or occasionally employed in controlled areas, Employers, registrants and licensees shall: who may receive significant occupational exposures. In case indi- 1. Provide for access by the workers to their own records vidual monitoring is not feasible or inappropriate, the exposure of 2. Provide for access by the supervisor of the health sur- the workers should be assessed on the basis of the results of the veillance programme by the regulatory authority and workplace monitoring and information on duration and location relevant employer of each worker. 3. Facilitate the provision of copies to new employment Individual monitoring shall not be provided to workers when necessary employed in a supervised area or who enter only occasionally a 4. Make arrangements for the retention of record, as controlled area, The occupational exposure of this kind of work- appropriate, when a worker ceases to work ers shall be assessed on the result of the workplace monitoring, 5. Ensure maintenance and confidentiality of data location and time required for the assigned work. The frequency, precision, nature and therefore choice of Exposure records for each worker shall be kept during the dosimeters shall be made considering the nature of the exposure, worker’s working life and afterwards until the worker attains or likelihood and magnitude of consequent potential exposure and would have attained the age of 75 years, anyhow for not less than risk. 30 years. The effectiveness of the protection devices provided, shall Further Readings: International Basic Safety Standards for be demonstrated, with particular attention to the respiratory Protection against Radiation and for the Safety of Radiation equipment used by workers who may be exposed to radioactive Sources. 1996. Safety Series No.115, International Atomic Energy contamination. The intake of radioactive substances or the com- Agency, Vienna, Austria; 1990 Recommendations of International mitted dose, as appropriate must be assessed. Commission on Radiological Protection, Pergamon, Oxford, UK, Monitoring of the Workplace: Registrants and licensees ICRP 60, 1991. shall establish, maintain and keep under review a programme for monitoring of the workplace, under the supervision, (following Dose optimisation the requirements of the regulatory authority) of qualified experts (Radiotherapy) Dose optimisation in radiotherapy is a treat- or radiation protection officers. ment planning procedure in which the treatment beam param- The nature and frequency of monitoring shall: eters and machine parameters are adjusted so as to optimise the 1. Enable (a) to evaluate the radiological conditions in all dose and dose distribution to better meet with the dose prescrip- workplaces, (b) to assess the exposure in controlled and tion. In planning of conventional treatment, dose optimisation is supervised areas and (c) to review the classification a manual process, in which the planner manually modifies the Dose painting 284 D ose profile beam parameters such as beam angle, wedge angle, beam size, dosimeters (TLDs). Figure D.66 shows once such profile obtained and beam energy to achieve the dose specification. In IMRT plan- with radiotherapy verification x-ray film and scanned on a scan- ning, dose optimisation is achieved through inverse planning (see ning microdensitometer. The optical densities have been con- Inverse treatment planning). verted to exposure readings by using a calibrated film. Related Articles: Slice thickness, Slice profile Dose painting (Radiotherapy) Dose painting describes the delivery of a non- Dose profile homogeneous dose to the target. This allows, for example, ele- (Radiotherapy) The profile of a photon beam consists of two dis- vation of the dose in certain (e.g. more radio-resistant) parts of tinct regions, the umbral and the penumbral region. The umbral the tumour. In photon therapy, dose painting is achieved through region is where the beam profile is unaffected by the collimators intensity modulated radiation therapy. In proton therapy, dose while in the penumbral region the field defining collimators affect painting can be achieved using pencil beam scanning to create an the beam profile. There is no exact indication where the transition intensity-modulated plan. between the two regions occurs but in general a nominal position D Dose re-painting describes the repeated delivery of the same from 1 to 1.5 cm inside the geometric field edge, which is consid- proton/ion field with a fraction of the planned dose. Dose re- ered the 50% dose level, is generally accepted as the approximate painting is a strategy for motion mitigation. location where the transition occurs. In the penumbral region the Related Articles: Multi-field optimisation, Motion mitigation primary photons are shielded by the collimator jaws and therefore in this region there is a sharp gradient fall-off, sigmoid in shape Dose profile (Diagnostic Radiology) In CT the term ‘dose profile’ is used to describe the dose distribution, along the axis of rotation (z-axis), from a single sequential irradiation. The irradiated ‘slice’ is defined as the full width at half maximum (FWHM) of this distribution. The dose profile should not be confused with the slice, or z-sensitivity profile. The width and shape of the dose profile are determined by the primary collimator opening and geometry, and the focal spot size (Figure D.65). Figure D.65 shows the dose profiles resulting from different configurations in green. (a) an ideal rectangular profile from a point source, (b) a realistic trapezoidal profile caused by the pen- umbra produced by a finite focus, (c) is for a similar configuration as (b) but with a reduced collimator opening and (d) the increased penumbra resulting from collimators closer to the focus. –10 0 10 Dose profiles are generally measured in air, on the scan- z-axis distance (mm) ner’s axis of rotation, with x-ray film or radiochromic film. They can also be measured in a phantom with thermoluminescent FIGURE D.66 CT scanner z-axis dose profile for 20 mm. Focus Collimators z-axis Axis of rotation Detectors Full width at half maximum (a) (b) (c) (d) FIGURE D.65 Dose profiles along