LO
CapabilityCognitive automation2022-01-13
Algorithms from a 1,290-developer challenge reached pathologist-level Gleason grading on validation sets from two continents, researchers reported
Pathologistoccupation page →
Event date / reported
2022-01-13
Evidence stage
CapabilityA demo, benchmark or paper shows the task can be done. Updates what the technology can do — not what employers will do.
Tasks this bears on
Detecting and grading cancer
Finding cancer in biopsies and assigning its grade, such as the Gleason grade for prostate cancer.
Being augmented✓ Evidence-backed
Where this applies
An international competition with 1,290 developers and 10,616 digitised prostate biopsies. Submitted algorithms reached pathologist-level performance on independent cross-continental cohorts, with agreements of 0.862 and 0.868 (quadratically weighted kappa) with expert uropathologists on US and European validation sets; the authors say this warrants evaluating AI grading in prospective clinical trials. Retrospective validation, not clinical use.
What this means
Grading, the judgement pathologists most often disagree on, can be matched by algorithms on retrospective slides.
What it does not yet show
Retrospective, prostate only; it does not show safety in clinical use.
What you can check
Open the Nature Medicine PANDA challenge article and find "pathologist-level performance".
Does it change the assessment?
No — and this stage does not move it either. A "Capability" record is real evidence, but it does not upgrade a task judgement on its own. The 1 linked judgement above stand where they were.
Source
Bulten et al. — Artificial intelligence for diagnosis and Gleason grading of prostate cancer: the PANDA challenge, Nature Medicine 28 (published 13 January 2022) · verified 2026-09-30 · Claude (VOLO agent) · interpreted 2026-09-30 · Claude (VOLO agent)
Primary source — published by the party that did this, or the authority of record. No co-signature needed.