Beyond the Model: Engineering Trust in Human-AI Systems

As healthcare AI matures, the biggest challenges are no longer solely about building more accurate models—they're about designing systems that help clinicians make better decisions. Success depends not only on algorithm performance, but also on human behavior, workflow integration, and thoughtful implementation.

Join Maureen Canellas, MD, MBA, FACEP, emergency physician, Associate Chief Medical Officer at UMass Memorial Medical Center, and MIT Research Affiliate in Healthcare AI, for an interdisciplinary exploration of the next frontier in healthcare AI. Drawing on psychology, human factors engineering, systems engineering, implementation science, and operations management, Dr. Canellas will explore how AI developers and healthcare leaders can design and evaluate human-AI systems to improve clinician trust.

What You'll Learn

• Why healthcare AI has matured from primarily a machine learning problem into a systems engineering problem
• How human factors—including automation bias, appropriate reliance, and cognitive offloading—influence clinician decision-making
• Why implementation, workflow design, and organizational systems often determine AI success more than model performance alone
• Apply a practical human-AI systems framework to evaluate and design AI tools beyond model performance alone