Rse Grand Challenge

Grand Challenge is a non profit cloud platform for developing, benchmarking and deploying machine learning solutions in biomedical imaging. It is built for researchers, clinicians and developers working across radiology, pathology and ophthalmology, with support expanding to modalities such as genomics and cardiology. The platform covers the whole workflow. Teams can upload and securely share imaging data, which is validated and converted to standardized formats, and view it in browser based workstations. Reader studies let organizers train expert annotators with sets of questions and immediate feedback, then gather annotations using customizable hanging protocols and overlays. Challenge organizers can host algorithm submission competitions, with submitted algorithms run on demand in a secured environment, scored by evaluation code the organizer provides, and ranked on live leaderboards. User management includes account verification and cheating detection, and organizers can choose GPU and memory options and set time limits. Algorithms submitted to a challenge can remain available afterward for clinicians and researchers to try or host. The platform's data, algorithms and challenges follow the FAIR principles, and a monthly cycle report describes new features. Pricing is published for challenge organizers. Hosting a challenge requires a base contribution of 6,000 euros, invoiced up front and non refundable, which covers up to 1,000 euros of compute and storage and funds platform maintenance, development and support. Compute and storage beyond that is covered by a reserved capacity budget, with additional charges rounded up to the nearest multiple of 250 euros. The page also lists expert review, personalized onboarding and dedicated support during the challenge.