Japan Association for Medical Informatics

[2-B-1-03] Toward Care Forecasting Services in Nursing and Medical Care

*Sozo Inoue1 (1. Kyushu Institute of Technology)

In this presentation, a research on future forecasting of care in nursing care and medical care, under the JST Future Society Creation Project, are introduced.
For care recipients and patients, predicting deterioration of physical condition, frailty, pressure ulcers, and end-of-life care will contribute to their QoL. For service providers, predicting excretion, nurse calls, accidents, loitering and BPSDs, and the moods of users and staff will also lead to improving productivity.
In response to these short-term and long-term forecast needs, we researched on future forecasting by using machine learning, combining nursing care record big data and sensors at nursing homes.
In this regard, we will also introduce recent efforts such as simultaneous behavior prediction of multiple people from cameras in public areas and advanced efforts such as active data acquisition through cooperation with generative AI.
In addition, as a future plan, we will describe a concept to realize the care forecast in the whole community, such as visits and homes, not only in medical facilities and nursing care facilities.