Designing a User Interface for Counterfactual Simulations of Adaptive Treatment Strategies

Abstract

AI research has proven potential to revolutionize the healthcare industry by introducing AI-enabled tools that can aid a clinician’s decision-making process in how to diagnose and treat patients. It can be especially useful in the field of personalized medicine, particularly for simulating the outcomes of adaptive treatment strategies applied to specific patients. In order for such a tool to be used in practice, however, it needs to be properly designed. Most decision support tools fail to reach the hands of doctors because they fail to address the needs of the user. The research in this paper uses principles of HCI to design a tool that can be used by clinicians and researchers to simulate strategies. The end results show that it is possible to design a medical tool that addresses the needs of multiple users and insights into the most important aspects that should be considered when designing such tools.