AI in Breast Imaging Accelerates Toward Regulated Automation

AI is moving beyond detection into mammographic risk prediction, MRI interpretation and triage, ultrasound, workflow automation, and reported autonomous normal-mammogram identification. A prior-cancer mammogram risk model reported AUC 0.824, while Vara's reported September 2026 CE authorization and the FDA's generative-AI discussion paper are major policy signals; clinical benefit, generalizability, calibration, and postmarket safety remain unresolved.