Kidney Artificial Intelligence Discovery Challenge: From Nebulae to Nephrons supports researchers using AI to uncover overlooked patterns in kidney biopsy images and advance precision nephrology.
Submissions opened November 5, 2026, replacing pre-registration, and registered participants can now download the challenge images. Finalized requirements include three findings narratives and a Git commit whose code runs end-to-end with one Docker command; shortlisted entries will be tested on unreleased images, and other submission details also changed.
Funder: NIH Challenges
Due Dates: April 29, 2027 (Submission) | April 30, 2027 (Judging begins) | June 30, 2027 (Judging ends) | September 1, 2027 (Winners announced)
Funding Amounts: $100,000 total prize purse; up to three awards of $50,000, $30,000, and $20,000
Summary: A prize competition for computational discovery of novel, scientifically meaningful patterns in kidney biopsy images.
Key Information: Use only open KPMP data; controlled data and federal funding for submission development are prohibited.
The National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) conducts this challenge to discover previously unrecognized or under-characterized patterns in Kidney Precision Medicine Project (KPMP) data. Inspired by anomaly detection in astronomy, it invites artificial intelligence, machine learning, computer vision, and related approaches to uncover rare or overlooked features in kidney pathology images—not merely classify known disease states.
Relevant discoveries may include unusual tissue structures, cellular arrangements, spatial disease signatures, progression-risk patterns, or treatment-response markers. Image-derived anomalies may be combined with open KPMP clinical, molecular, or spatial data to explore patient stratification. The challenge emphasizes evidence distinguishing meaningful findings from artifacts and collaboration with kidney experts to interpret results. Its broader goal is to advance precision nephrology through cross-disciplinary discovery.