Human Digital Twins for Personalized Healthcare: A Review of Computing-Networking Architecture, Enabling Technologies and Open Challenges

Abstract

Rising healthcare expenditure, documented waste within care delivery, and system-wide shocks such as the COVID-19 pandemic have intensified interest in care models that are proactive, continuous, and tailored to the individual patient. The digital twin (DT) paradigm - a continuously synchronized virtual counterpart of a physical entity, first established in industrial and manufacturing settings - has increasingly been extended to human-centric systems through the concept of the human digital twin (HDT): a dynamic, data-driven computational representation of an individual patient's physiology and health trajectory. Because an HDT must represent a living, constantly changing human body rather than a mechanical asset, realizing it depends on more than modeling technique alone; it requires a networking architecture capable of continuously acquiring, transporting, computing on, and managing high volumes of sensitive, often latency-critical patient data. This review synthesizes the literature on digital twin fundamentals, the emerging HDT vision, and the networking technologies - edge-cloud compute continua, sixth-generation (6G) networks, and digital twins of the network itself - positioned to support it. We further examine commercial and research implementations, relevant consortiums, and cross-domain lessons from industrial, automotive, and structural-health digital twins, before outlining open challenges in security, interoperability, and resource allocation, and directions for future work