AI applications in health care are prone to biases that could perpetuate health disparities. In this paper we study the ways in which AI may maintain, perpetuate, or worsen inequitable outcomes in health care. We review current approaches to evaluating and mitigating biased AI and potential applications of AI to address health equities. Finally, we discuss current incentives for equitable AI and potential changes in the regulation and policy space. As AI becomes increasingly embedded in the daily operations of health care systems, it is imperative that we understand its risks and evaluate its impact on health equity.
Copyright:
Reproduced with permission of the copyright holder. Further use of the material is subject to CC BY-NC-DC license. (More information)