Fujifilm and Juntendo Hospital have developed an AI that uses hospital data to accurately predict the risk of falls for outpatients.
The research team collected over 500 characteristics associated with falls, such as age and prescription history, from hospital data collected by Fujifilm’s CITA Clinical Finder. This data platform is used to centralize data from various hospital departments, including EMR, radiology, and endoscopy.
The AI model is then trained based on these features. Predict an individual’s risk of falling. The technology displays your predicted fall risk as a percentage and identifies characteristics that may be potential risk factors.
The researchers then tested the AI’s effectiveness on data from about 70,000 outpatients at Juntendo Hospital, and the latest findings showed that the AI achieved 96% accuracy in predicting and generating fall risk.
According to the media release, Fujifilm and Juntendo will continue testing the AI with the aim of early clinical application.
Why is this important?
In Japanese hospitals, it is said that it is difficult to identify patients at risk of falling because outpatients have limited hospitalization time. They also realized that the number of outpatients who need to be evaluated for fall risk is greater than inpatients. There is an increasing demand for efficient and accurate solutions to prevent falls in medical settings.
Market Snapshot
Research is already being conducted in Japan on using AI to prevent falls in hospitals and nursing homes. For example, Fujitsu and Wakayama Medical University have been working on trials combining sensors and AI technology to detect falls while protecting patient privacy.
Meanwhile, Fujifilm’s latest AI innovation joins the company’s expanding portfolio of elderly care technologies. In 2022, the company will AI to predict progression of Alzheimer’s disease in patients with mild cognitive impairment.
