Interested in digital pathology image analysis but not sure where to start? Check out the new set of interactive python notebook tutorials demonstrating how to use open-source TIAToolbox with the slide microscopy images available from Imaging Data Commons: GitHub - ImagingDataCommons/idc-tiatoolbox: Tutorials to help you get started applying TIAToolbox to IDC data. You will learn how to:
- load DICOM WSI images with TIAToolbox
- perform stain normalization
- extract image patches suitable for analysis
- semantically classify image patches and regions
- segment and type individual nuclei
IDC contains over 45TB in publicly available digital pathology images (>70K digitized slides), including those accompanied by rich genomic information (e.g., collected by the The Genotype-Tissue Expression (GTEx) and Childhood Cancer Data (CCDI) initiatives). Hugely under-explored space for those looking for meaningful challenges applying AI to cancer imaging research! Screenshot attached is composed of the analysis results examples that you will find in the notebooks.
IDC Claude skill will help you navigate and use IDC content: GitHub - ImagingDataCommons/idc-claude-skill: Natural language interface to NCI Imaging Data Commons
