Kather 2016 colorectal cancer histology textures (5,000 H&E tiles, 8 tissue classes)
- Provider
- Jakob Nikolas Kather et al., Heidelberg University
- License
- CC-BY-4.0 (source)
- Size
- 3 files
- Format
- Parquet (in sampled files)
- Stored at
- Hugging Face (https://huggingface.co/datasets/1aurent/Kather-texture-2016)
- Access
- Free and public. No account with the data provider is needed.
- DOI
- 10.5281/zenodo.53169
- Category
- Medical Imaging
- Source
- https://doi.org/10.1038/srep27988
About this dataset
A collection of 5,000 RGB histology image tiles of human colorectal adenocarcinoma, 150 by 150 pixels each (74 by 74 micrometres at 0.495 micrometres per pixel), digitised at 20x on an Aperio ScanScope from formalin-fixed paraffin-embedded primary tumours at the Institute of Pathology, University Medical Center Mannheim. Each tile belongs to exactly one of eight tissue classes: tumour epithelium, simple stroma, complex stroma, immune cells, debris, normal mucosa, adipose tissue and background, 625 tiles each. It is a small, openly licensed benchmark for testing pathology slide encoders and tile classifiers. Mirrored from the original Zenodo record (doi 10.5281/zenodo.53169); cite Kather et al., Scientific Reports 2016. CC BY 4.0.
How to load this dataset in Python
from datasets import load_dataset
ds = load_dataset("1aurent/Kather-texture-2016")
print(ds)Data structure
File types in sampled files: Parquet
Files
Example file: data/train-00000-of-00001-969839c50ce26272.parquet
Analyze this dataset without downloading it
On Scigantic, Kather 2016 colorectal cancer histology textures (5,000 H&E tiles, 8 tissue classes) opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.