Building a Digital Health Twin for Personalized Intervention: The EPI Project

Abstract

The Enabling Personalized Interventions (EPI) project, part of the COMMIT2DATA top sector initiative, brings together research on data science, software-defined network infrastructure, and secure and trustworthy data sharing, executed within the healthcare domain. The project applies the digital twin paradigm, in which data science-driven algorithms monitor and perform functions on a digital counterpart of a real-world entity, to enable proactive responses based on predicted outcomes. The EPI project applies this paradigm in the healthcare context by developing and testing applications that can act as personalized digital health twins for self/-joint management. The EPI project addresses several challenges to digital twin applications in the healthcare domain, such as: 1) strict health data sharing policies often lead to data being locked in silos, 2) legal, policy and privacy requirements make data processing increasingly more complex, and 3) significant limitations on infrastructure resources may apply. In this paper, we report on the use cases the EPI used as the basis to develop possible solutions to these challenges. In particular, we describe algorithms and tools for algorithmic real-time response and analysis of distributed data at scale. We discuss the automatic enforcement of legal interpretations and data-sharing conditions as executable policies. Finally, we investigate infrastructural challenges by implementing and experimenting with the EPI Framework - consisting of a distributed analysis infrastructure and BRANE for orchestrating multi-site applications. We conclude by describing our Proof of Concept (PoC) and showing its application to one of the EPI use cases.