Artificial Intelligence in Healthcare: Considerations for Adoption and Adaptation in Academic Medical Settings

Abstract

Discussions around artificial intelligence (AI) and machine learning (ML) and their applicability within academic medicine have become prominent over the past several years. Various end-user-focused AI/ML tools have emerged, offering opportunities to enhance efficiency and improve outcomes in biomedical research and medical education. While AI holds the promise of revolutionizing many aspects of academic medicine, the gravitas of the medical field necessitates scrupulous consideration and forward planning when implementing AI/ML in medical settings. Consequently, frameworks to guide AI/ML implementation discussions within academic medicine are crucial for mitigating the inherent pitfalls of such technology. This chapter proposes a framework to assist decision-makers in the academic medicine ecosystem with AI/ML implementation decisions. The framework emphasizes [A] understanding the functionality of different types of AI (Large Language Models, Computer Vision, and Omics Learning Models) to identify inherent use cases and limitations; [B] considering regulatory constraints and ethical principles specific to the implementation context; and [C] evaluating the overall costs and benefits of AI/ML implementation. Proactively balancing innovation with human oversight is essential to leveraging AI’s benefits while mitigating risks. As AI in healthcare evolves, ongoing research, collaboration, and regulations will be vital to ensure AI is aligned with the goal of advancing healthcare responsibly.