AI tools are entering healthcare rapidly, yet safe adoption remains difficult. Only 2% of ICU AI models reach real use, largely due to limited monitoring once deployed. This session introduces the Generative AI Control Center, developed at Erasmus MC using SAS Viya, which continuously evaluates AI tools through automated LLM-based assessments and visual risk scoring. Attendees will learn how to implement transparent, real-time monitoring that builds clinician confidence, identifies when tools are helpful or harmful, and enables responsible and scalable AI use in the ICU. We also show how ICU Maps and the Respiratory Dashboard support daily workflow, illustrating how continuous oversight helps hospitals meet upcoming regulatory demands while ensuring AI truly supports patient care.
What does it really take to turn AI in healthcare from hype into real, life-changing impact? In this compelling talk, researchers from the REAL Lab expose why only a tiny fraction of models ever reach clinical practice—and reveal the hidden barriers of trust, ethics, and workflow integration. Through a powerful ICU case study, you’ll see how responsible AI, stakeholder involvement, and rigorous validation can transform promising ideas into tools that actually help clinicians and patients. If you want to understand what’s missing in today’s AI revolution—and how to fix it—this is a must-watch.
Watch the recording
