Human-AI Medical Decision-Making Under Uncertainty

Abstract

AI-based decision support systems are becoming increasingly common in healthcare, promising improved accuracy and efficiency in diagnosis, treatment, and prediction. At the same time, due to the high-stakes and highly regulated context of medical decisions, uncertainty is neither an exception, nor a mere individual decision characteristic. Rather, uncertainty is a continuously constructed and enacted phenomenon by both focal medical professionals and the array of actors and technologies involved in healthcare tasks. The purpose of this panel symposium is to engage a group of panelists in a formal, moderated, interactive discussion on how uncertainty is dealt with by (a) medical doctors and healthcare professionals, (b) AI systems, and (c) human-AI ensembles. We would like to stimulate discussions around the changing nature of medical decision-making revolving around the new types of uncertainty emerging in the interactions between human and AI technologies. We would also like to examine the consequences of such interactions for the way medical decision making is enacted and organized.

Ähnliche Arbeiten