A hypergraph convolution-based intelligent healthcare platform for aging population management.

Introduction: The growing aging population imposes increasing demands on healthcare systems, particularly in managing chronic diseases among older adults. However, existing approaches face significant challenges in integrating multimodal data and analyzing complex disease associations effectively. Methods: This study proposes an intelligent healthcare platform based on Hypergraph Convolutional Networks (HGCN) to address these limitations. The platform collects real-time multimodal data-including physiological signals, behavioral records, and environmental parameters-via wearable and IoT devices. These data are integrated into a dynamic medical knowledge graph, and analyzed using HGCN and hierarchical feature learning to facilitate health condition monitoring and inter-institutional collaboration. Results: Experimental evaluations demonstrated the platform's effectiveness, achieving an 87.26% accuracy and a 0.831 F1-score in disease risk prediction. The system also maintained a 100% request success rate under 480 concurrent users, with minimal response latency. Discussion: The proposed platform significantly improves personalized care for older adults, enhances the efficiency of healthcare resource allocation, and offers a scalable solution for intelligent healthcare services.