Bioethics Principles in Machine Learning-Healthcare Application Design: Achieving Health Justice and Health Equity

Health technologies featuring artificial intelligence (AI) are becoming more common. Some healthcare AIs are exhibiting bias towards underrepresented persons and populations. Although many computer scientists and healthcare professionals agree that eliminating or mitigating bias in healthcare AIs is needed, little information exists regarding how to operationalize bioethics principles like autonomy in product design and implementation. This short course is framed with a Social Determinants of Health lens and a health justice and health equity stance to support computer scientists and healthcare professionals in building and deploying ethical healthcare AI. In this short course we introduce the bioethics principle of autonomy in the context of human-centered design (Module 1) and share options for design thinking models, suggesting four activities to embed ethics principles during design (Module 2). We then discuss the importance of gaining the perspectives of diverse groups to minimize harm and support the fundamental human values of underrepresented persons in support of health equity and health justice ideals (Module 3).