A machine-learning challenge to promote the development of algorithms for the detection and localization of COVID-19 on chest radiographs is being launched by RSNA, the Society for Imaging Informatics in Medicine (SIIM), and the Foundation for the Promotion of Health and Biomedical Research of Valencia Region (FISABIO) in Spain.
SIIM partners HP and Intel will provide $100,000 U.S. (82,000 euros) in prizes for the challenge, which will be hosted on Kaggle and will begin one week before SIIM's annual meeting on 24-27 May. The winning models will then be made available as open source to improve patient care, according to the societies.
The competition will utilize augmented annotations on chest radiographs from the Medical Imaging Data Resource Center (MIDRC) -- RSNA International COVID-19 Open Radiology Database (RICORD), as well as from the BIMCV-COVID-19 dataset produced by an international group of volunteer radiologists from Brazil, Spain, and the U.S.
The challenge is also being supported by the National Science Foundation (NSF) Convergence Accelerator Grant that SIIM and its collaborators were awarded in September 2020. SIIM is calling for open-source AI models to populate the prototype "Model Zoo" built as a result of phase 1 of this grant. The Model Zoo will then serve as the basis for clinical research testing of a collaborative model-centric AI platform designed to support the validation and translation of AI in medical imaging, according to the societies.
Winners of the COVID-19 detection and localization challenge will be announced September 19-20 at SIIM's 2021 Conference on Machine Intelligence in Medical Imaging.
