This session explores how agentic AI systems will reshape radiology by autonomously coordinating and executing multi-step tasks such as exam protocoling, scheduling, quality optimisation, preliminary interpretation, communication of critical findings, and follow-up management. It highlights both the innovation potential and the risks of autonomous multi-action agents in clinical environments.
Chairperson's introduction
Susan Cheng Shelmerdine, London / United Kingdom
Understanding agentic AI: capabilities, risks, and the leap from assistance to orchestration
Kostas Marias, Heraklion / Greece
Agent-based workflow automation and real-world prototypes of agentic radiology
Kevin Groot Lipman, Amsterdam / Netherlands
Clinical safety and governance for agentic AI
Christina Malamateniou, London / United Kingdom
Q&A: Beyond copilots: Agentic AI systems and the future of radiology operations