CONNECTED: A Knowledge Graph-Driven Platform for Clinical Data Harmonization and Personalized Digital Twin-Based Healthcare

The digital healthcare innovation surge requires frameworks integrating heterogeneous clinical data to support real-time, personalized care. However, existing solutions often face interoperability, scalability, and adaptability limitations, restricting their utility in predictive, precision medicine. This paper introduces CONNECTED (COmpreheNsive and staNdardized hEalth-Care plaTforms to collEct and harmonize clinical Data), a multi-layer, microservices-based platform designed to facilitate Digital Twins (DTs) for patient monitoring and personalized treatment. The solution’s core innovation lies in using knowledge graphs, which harmonize diverse clinical data sources and link patient information to AI models through automated, manifest-driven interfaces. This approach enables adaptive, patient-specific simulations by integrating real-time and historical data. We validate CONNECTED by implementing a stroke risk classifier, demonstrating the platform’s potential to provide patient-specific predictions supporting early intervention strategies. CONNECTED thus offers a scalable, flexible foundation for precision medicine, equipping clinicians with actionable insights across varied clinical applications.