RSNA Benchmarks is an open initiative. We welcome contributions from radiologists, AI researchers, radiology AI vendors, regulators, and institutions worldwide.
Share anonymized cases from your institution to strengthen benchmark diversity and clinical representativeness.
Learn More→Help design evaluation frameworks, define ground truth protocols, and build the technical infrastructure.
Get Started→Run your models against our benchmarks and contribute results to the growing body of evaluation data.
Coming Soon→RSNA Benchmarks draws data and expertise from institutions worldwide.
Participating institutions to be announced.
Upcoming events, workshops, and presentations related to the RSNA Benchmarks initiative.
Event schedule coming soon.
Any radiologist, AI researcher, radiology AI vendor, regulator, or institution can participate. We welcome contributions of data, expertise, and technical development from the global radiology and medical AI community.
All contributed cases must be fully anonymized and de-identified in accordance with institutional IRB requirements and applicable data protection regulations before submission.
Each benchmark is overseen by a working group with representation from contributing institutions. A steering committee provides overall governance for the initiative.
Yes. Once a benchmark is published, evaluation protocols and submission guidelines will be made publicly available so that any team can assess their model's performance.