March 13, 2025
Ian Pan, MD | Brigham and Women's Hospital
Artificial intelligence (AI) in radiology has been slowly gaining traction over the past decade. To promote the creation of clinically relevant AI tools in radiology, organizations such as the Radiological Society of North America (RSNA) have sponsored AI challenges where participants compete to create the best models for a given task. Challenge tasks are designed to tackle real-world clinical problems and have covered various domains, including pediatric bone age assessment, intracranial hemorrhage detection, and lumbar spine degenerative disease classification. In this talk, we will walk through the history of these challenges through the lens of an experienced participant and top performer, and discuss lessons learned, clinical applications, research developments, and future directions.
