owkin/phikon
Feature Extraction • 86.4M • Updated • 40.5k • • 40
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
The NCT-CRC-HE dataset consists of images of human tissue slides, some of which contain cancer.
The dataset contains tissues from different parts of the body. Examples from each of the 9 classes can be seen below
NCT biobank (National Center for Tumor Diseases) and the UMM pathology archive (University Medical Center Mannheim). Images were normalized using Macenko normalization.
CC-BY-SA
Owkin claims no ownership of the dataset. This is simply an upload of the original dataset onto HF.