Cancer3.AIAI in OncologyDatasets › ISIC Archive — International Skin Imaging Collaboration
Registry / portal Open download CC-0 / CC BY-NC per contributor

ISIC Archive — International Skin Imaging Collaboration

The public backbone of skin-cancer AI; strongly skewed towards light skin tones, which every model card built on it should say.

At a glance

ProviderISIC (Memorial Sloan Kettering and partners)
AccessOpen download
LicenceCC-0 / CC BY-NC per contributor
dermoscopic images70 000
Size notestens of thousands of images with diagnosis; annual challenge subsets (e.g. 2020: 33,126 images)
FormatsJPEG, CSV

Labels and annotations

Diagnosis (melanoma, nevus, BCC, ...), often histopathology-confirmed; some subsets have lesion segmentation and dermoscopic attribute masks; patient metadata (age, sex, site).

Details

Not yet documented on this card.

Models trained or evaluated on it

No model in this catalog is linked to this dataset yet.

Sources

  1. ISIC Archive

This page is educational — it is not medical advice and does not replace consultation with an oncologist. Diagnostic and treatment decisions are made solely by specialist physicians.