Building Confidence in Clinical AI: Clinical Risk Management in AI Training at the Royal Society of Medicine

Newton’s Tree’s advanced training workshops give attendees practical strategies for managing clinical risks in AI. The course not only emphasises the critical importance of safety in AI systems but also crucially empowers professionals to take these insights back to their organisations and implement safer AI-driven solutions.

On 18th September 2024, Newton’s Tree hosted their second advanced training course on clinical risk management in AI, designed to help those working on AI projects in healthcare including CMOs, CCIOs and CNIOs, clinical scientists and safety officers from both NHS and industry gain the skills to manage the unique risks posed by AI applications. With a agenda featuring case studies, workshops, and expert-led discussions, the event saw a blend of theory, practical application, and networking opportunities.

The day began with an introduction to the importance of clinical risk management in AI, laying the groundwork for understanding the complexities of managing AI risks in a healthcare setting. The first workshop of the day tackled the clinical risk management of a diagnostic AI system. Participants engaged in interactive sessions where they identified potential hazards, performed evaluations, and proposed hazard control strategies based on real-world examples.

Following a networking lunch, the second workshop zoomed in on the challenges of managing risks in a large language model (LLM)-based AI system. These systems, which power applications like AI-driven chatbots and ambient intelligence tools, present new - and dynamic - layers of risk that will evolve quickly as time moves on.

The course provided invaluable lessons on integrating clinical risk management into AI healthcare systems learning the interplay between ISO and DCB Standards, and highlighted AI-specific risks, particularly those using machine learning due to their dependency on data, potential instability, and feedback loops. Finally the workshop outlined practical tools for participants to identify hazards, estimate risks, and implement controls in AI systems.

Newton’s Tree’s advanced training workshops give attendees practical strategies for managing clinical risks in AI. The course not only emphasises the critical importance of safety in AI systems but also crucially empowers professionals to take these insights back to their organisations and implement safer AI-driven solutions.

Newton’s Tree is committed to increasing knowledge sharing and advancing the safe integration of AI in healthcare. We’ll be putting on another course early next year (date TBC), so sign up to our newsletter below to receive updates or follow us on linkedin.

Also, if you would like to discuss a bespoke course for your organisation feel free to get in touch.