Sustainable Edge Deployment of Consumer-Centric Digital Twins for Healthcare 5.0: A Resource-Efficient Optimization Framework

Healthcare 5.0 emerges as a transformative paradigm positioning patients as active participants through consumer-centric digital twin technologies that enable proactive health management and personalized interventions. Consumer-centric digital twins (CCDTs) function as high-fidelity virtual patient replicas continuously processing physiological data streams from wearable sensors and medical devices to predict health trajectory deviations and recommend timely corrective actions. While patient CCDTs exhibit heterogeneous requirements for processing electrocardiograms, glucose monitors, and mobility data. This study formulates sustainable CCDT deployment as a binary integer programming problem minimizing total system latency under computational capacity, storage limits, synchronization requirements, and collaboration needs, proving its NP-hardness. The proposed sustainable patient-centric edge deployment (SPCED) algorithm combines branch-and-bound initialization through linear relaxation with matching-theory refinement using move and swap operations, achieving bilateral exchange stability. Simulation results demonstrate that SPCED outperforms six baseline algorithms, including state-of-the-art decentralized federated learning approaches, achieving 99.81% computational efficiency improvement while maintaining superior latency performance across varying patient populations, resource configurations, and collaboration patterns, validating its effectiveness for sustainable healthcare 5.0 CCDT ecosystems.