A survey on ethical digital twins in healthcare

Abstract

Digital Twins for Health (DT4H), defined as dynamic, data-driven, in silico representations of individuals, organs, medical devices, clinical workflows, or healthcare systems, are emerging as a promising paradigm for transforming healthcare. DT4H systems have the potential to advance personalized medicine, treatment optimization, disease prediction, drug discovery, clinical trials, and health system planning. However, their real-world development and deployment also raise important ethical, legal, and social implications (ELSI). Because DT4H systems rely on sensitive longitudinal health data and may directly inform clinical decision-making, they introduce significant concerns related to their development, use, and ongoing governance. Many of these concerns are not entirely new, as they are grounded in established bioethical principles and substantially overlap with ELSI frameworks developed for AI for healthcare (AI4H). Accordingly, rather than examining DT4H-related ELSI issues in isolation, this survey situates them within existing bioethical and AI4H frameworks while identifying the distinctive challenges that arise from the unique characteristics of DT4H and must be addressed to support their ethical, responsible, and trustworthy development and deployment.