125,091 DICOM files from a cancer journey released publicly for deep learning research
typewriters · x · 2026-10-20
A longitudinal, multi-modal medical imaging dataset from a full cancer journey is now public: 125,091 DICOM files, viewable online via OHIF.
- Images are dated and linked to the treatment timeline, spanning initial diagnosis, standard of care, first remission, "founder mode cancer care", and second remission — enabling cross-stage comparison
- The authors explicitly invite anyone in deep learning for radiology to apply their methods and models to the data; findings can be sent to [email protected]
- Context: patients have a legal right to their DICOM files but hospitals make them hard to obtain; the promoters argue the usual excuses (file size, security, technical complexity) no longer hold and want to build the case for setting imaging data free
More from Research
- Berkeley's EyeRobot 2.0 Replaces Wrist Cameras with Active Gaze for Precise Manipulation — berkeley_ai · 2026-10-20
- Ramez essay: a million parallel agents won't yield superintelligence — returns diminish fast — chrisalbon · 2026-10-20
- Good Papers: a Rotten Tomatoes for ML papers with 20 AI reviewers — shangbinfeng · 2026-10-20
- Taco: hardware-agnostic LLM compression with one hash across 7 backends — phoenix10701 · 2026-10-20
- A multi-year deep dive into measuring AI slop across subreddits — Uhlo · 2026-10-20
- Scaling past 1T params: routing, hybrid attention, and 1M-token contexts — philipkiely · 2026-10-20