An ICT-based supporting system for home-nursing services

These days, the demand for home nursing in Japan has increased significantly. Along the increase, the shortages of staff and inefficiency of operations have become serious problems. This study aims to develop a system to support the operational efficiency of home nursing. As a thing being of particular importance, we focus on producing a summarized report of a patient's presenting symptom for a medical doctor with several nurse's notes that include unstructured data. A manual process of making the report imposes excessive burdens on nurses. However, the unstructured data contains crucial information, the course of symptoms. In this paper, we demonstrated an analysis of symptom courses, which includes extracting sentences about symptoms and classifying the status of symptoms. Experimental outcomes show that status classification works well, but there is room for improvement in both symptom extraction and status classification. As a future work, to improve the accuracy of symptom name extraction by human annotation and symptom classification by learning word dependencies. The symptom analysis system will contribute to support the time-consuming paperwork in home nursing.