metadata
license: apache-2.0
tags:
- vision
- image-classification
(Brain) Common Brain Tumors and Surrounding Tissue Elements
This model can additionally be run on our pathology reports platform
Credits: Kevin Faust & Michael Lee and Colleagues (University Health Network, Canada)
Introduction
This H&E brain tumor classifier was developed using transfer learning on a histology optimized version of the VGG19 CNN (DOI: 10.1038/s42256-019-0068-6) and trained to recognize the four most common brain tumor types and other surrounding tissue elements.
H&E-stained images patches (dimensions: 1024 x 1024 px) and annotations were taken from previous publicly available datasets (PMID: 35156037). Validation testing was carried out on non-overlapping cases from TCGA.
Classes
- Acute hemorrhage
- Blank space
- Cerebellar cortex
- Chronic hemorrhage
- Crush Artifact
- Fibrocollagenous tissue
- Glial histology
- Gray matter
- Normal hematolymphoid tissue
- Meningothelial Histology
- Epithelial pattern
- Skeletal muscle
- Necrosis Radiation
- Necrosis
- Schwannian/spindled Histology
- White matter
This information can be found in the inference.json file
Evaluation Metrics
Classifier validation can be found on the pathology reports platform