Presentation (CTI)

OBJECTIVES: The integration of healthcare and elderly care begins with the collection and fusion of data to eliminate silos between medical and elderly care information systems, enabling seamless data sharing. However, this process is fraught with challenges, including the multi-source and heterogeneous nature of the data, high levels of specialization, and diverse modalities. Furthermore, stakeholders' low willingness to share data and operational difficulties faced by healthcare and elderly care institutions exacerbate these challenges. To address these issues, this study proposes a task-oriented multi-source heterogeneous data fusion framework, validated through experimental applications in scenarios such as disease prediction and chronic disease management.
METHODS: A mixed-methods approach was employed, incorporating literature review, theoretical analysis, case studies, and experimental validation. A task-oriented multi-source heterogeneous data fusion framework was proposed and validated through experimental applications in scenarios such as disease prediction and chronic disease management.
RESULTS: The task-oriented multi-source heterogeneous data fusion framework includes three layers: data collection, data fusion, and data service. The framework has been validated using real-world datasets and has proven effective in improving the performance of disease prediction tasks, demonstrating its feasibility and practicality.
CONCLUSIONS: The task-oriented multi-source heterogeneous data fusion framework offers a foundational step toward advancing the integration of healthcare and elderly care systems. By overcoming challenges related to data silos, heterogeneity, and operational barriers, the framework improves the practical application of data sharing and supports the development of data-driven elderly healthcare services.