Optimizing Ultrasound Point-of-Care Resource Allocation with Complex Network and Dynamic Demand Mining

Objective To analyze the discrepancy between the demand for point-of-care ultrasound consultations and resource allocation, investigate its dynamic patterns, and propose optimized resource allocation strategies.Methods Based on ultrasound data from the 2024 HiscanPACS system at Yixing People's Hospital, a static network involving patients, requesting departments, and ultrasound physicians was constructed to analyze its structural characteristics. A time-series network was developed using the NetworkX library in Python, and Z-score analysis was applied to identify abnormal periods and examine changes in collaborative structures. Key nodes and high-frequency pathways were identified from sub-networks of positive results. A super-network was constructed from inter-departmental requests to apply community detection algorithms to identify collaborative communities among departments and their associated physician configurations.Results The static network exhibited good connectivity (average path length 1.91, network diameter 2) and modular structure (modularity 0.322, average clustering coefficient 0.444), yet overall sparsity remained (graph density 0.002). Time-series analysis identified days with abnormally high demand (threshold: 15.06 events/day). During these periods, the network scale decreased by 84.9%, but collaboration within the core team strengthened, showing a decentralized structural trend. High-frequency doctor-patient interactions and key physician nodes (e.g., U2080) were recognized in the positive sub-network. A high degree of modularity (Q = 0.662) was observed in the cross-departmental collaboration network, with the Louvain algorithm partitioning it into five functionally distinct communities. Notably, Community 0 exhibited a significantly higher workload compared to all other communities.Conclusion Using complex network analysis and dynamic demand mining, this study reveals key issues in point-of-care ultrasound resource allocation and proposes refined management strategies—including scheduling optimization, equipment sharing, and process referral—to provide a basis for improving service levels in healthcare institutions.