Grand Challenge

Grand Challenge is a platform for end-to-end development of machine learning solutions in biomedical imaging. It offers a comprehensive set of tools for developing, managing and deploying medical imaging algorithms. It is used by a community of more than 101,000 users and has more than 350 challenges, with more than 4,357 algorithms available to use.

Grand Challenge offers several key features for developing and managing machine learning solutions:

• Data Management: Upload medical imaging data easily and securely, controlling access and using browser-based workstations available anywhere.
• Annotator Training: Create sets of questions for annotators, invite experts to participate, and get immediate feedback on their performance.
• Annotation Gathering: Customize hanging protocols and overlays, and use intuitive workstations for viewing and reporting images.
• Algorithm Benchmarking: Manage annotated data sets, collect machine learning solutions, and objectively evaluate their performance.
• Algorithm Deployment: Upload algorithm container images, manage access, and run data on the platform's infrastructure.

The platform is designed to accommodate a wide variety of medical imaging challenges, including algorithm submission challenges like AutoPET III, The LEOPARD Challenge and Abdominal Circumference Operator-agnostic UltraSound measurement. Featured algorithms available on the platform include DataScientX, Deep Learning to Estimate Pulmonary Nodule Malignancy Risk, and Pancreatic Ductal Adenocarcinoma Detection in CT.

The cost for running a challenge on Grand Challenge is a base contribution of €5,000 and a variable capacity reservation. These costs cover the costs of running the platform, including maintenance, development and infrastructure. The capacity reservation is non-refundable and expires in the same calendar year as the invoice.

Grand Challenge is geared for medical imaging researchers and professionals, streamlining the development and deployment of machine learning solutions while maintaining data security and scalability. With robust tools for user management, discussion forums and automated evaluation, the platform fosters a community focused on driving innovation in biomedical imaging.

Published on August 5, 2024