When Does AI Earn a Doctor's Trust?

The Data Science Discipline of the School of EECS is hosting the following guest seminar:

Title: When Does AI Earn a Doctor's Trust?

Speaker: Professor Patrick Mikalef, NTNU – Norwegian University of Science and Technology

Abstract: AI promises substantial gains in healthcare, yet the opacity of clinical AI systems makes trust a precondition for their safe use. Most research treats trust as something users grant — or withhold — once an AI system is deployed. But what happens earlier, while the system is still being built? Drawing on an exploratory case study of six machine-learning and deep-learning pipelines for clinical decision support at a Norwegian university hospital, this talk examines how multidisciplinary teams of clinicians, AI developers, and biomedical engineers cultivate trust during pipeline development, and how explainable AI (XAI) shapes that process. It opens up a different way of thinking about XAI: not as a technical property of finished systems, but as an ensemble of sociotechnical practices through which trust is built, tested, and rebuilt as AI pipelines take shape. The talk traces what these practices look like in clinical settings, why trust-building during development is far from a steady accumulation, and what this means for how we design and govern healthcare AI.

Short bio: Patrick Mikalef is a Professor in Data Science and Information Systems at the Department of Computer Science. In the past, he has been a Marie Skłodowska-Curie post-doctoral research fellow working on the research project "Competitive Advantage for the Data-driven Enterprise" (CADENT). He received his B.Sc. in Informatics from the Ionian University, his M.Sc. in Business Informatics for Utrecht University, and his Ph.D. in IT Strategy from the Ionian University. His research interests focus the on strategic use of information systems and IT-business value in turbulent environments. He has published work in international conferences and peer-reviewed journals including the Journal of Business Research, British Journal of Management, Information and Management, Industrial Management & Data Systems, and Information Systems and e-Business Management.