(This is a joint announcement with Wakayama Medical University, so there is a possibility that the information will be duplicated.) Fujitsu Limited
Fujitsu and Wakayama Medical University Commence Joint Verification Test of Monitoring Technology Using Millimeter Wave Sensors to Detect Elderly Falls
Fujitsu Limited (*1) (hereafter, Fujitsu) and Wakayama Medical University (*2) (hereafter, Wakayama Medical University) have jointly developed a technology to estimate human posture using millimeter-wave sensors instead of cameras. Then, using monitoring technology that utilizes Fujitsu’s proprietary AI technology “Behavior analysis technology Actlyzer” (Note 3), which detects behavior in detail from changes in posture, situations such as falls can be detected early and fractures can be detected. From November 21, 2022, we will start a joint demonstration experiment aimed at reducing the risk of aggravation such as in hospital rooms and other facilities with high privacy.
Fujitsu will analyze the data obtained from the demonstration experiment, and work to improve monitoring technology based on the analysis results evaluated by Wakayama Medical University using knowledge from the medical field. In addition, using this technology, we will develop a monitoring service that immediately detects movements such as falls that can lead to serious disability in the elderly at hospitals and other facilities, and notifies the staff, aiming to provide it in FY2023. Through the results of this
demonstration experiment, both parties will contribute to the creation of an environment in which the elderly can live a safe and secure life in a super-aging society.
According to the All Japan Hospital Association (Note 4), there will be 290 falls in hospitalized patients per month in 2021 at 18 hospitals in Japan. Falling in the elderly has a high risk of leading to a serious disability and requires monitoring. On the other hand, it is difficult to introduce monitoring technology that installs cameras in hospital rooms, etc., from the perspective of impairing patient privacy.
In order to solve the above problems, the two parties will use privacy-conscious monitoring technology that does not use cameras to detect conditions such as falls in elderly people in hospitals at an early stage and take appropriate measures quickly. We will jointly conduct a demonstration experiment to reduce the risk of aggravation such as bone fractures.
[About the joint demonstration experiment]
1. Implementation period:
November 21, 2022 (Monday) to March 31, 2024 (Sunday)
(1) Demonstration technology: monitoring technology using millimeter wave sensors
Fujitsu developed in July 2022 a technology that uses a
millimeter-wave sensor to estimate a person’s posture, and Fujitsu’s proprietary AI technology “Behavior Analysis Technology Actlyzer” that recognizes complex behavior from changes in a person’s posture. We will conduct technical verification of monitoring technology that can detect movements such as falls while considering privacy.
(2) Demonstration experiment content: Data measurement of elderly people Both will conduct demonstration experiments at actual facilities such as hospitals, inviting collaborators from elderly patients and people requiring nursing care as test subjects. The patient’s posture is collected as point cloud data from millimeter-wave sensors installed in the room, and Fujitsu’s monitoring technology detects falls and movements that lead to falls. By comparing it with logs and camera images installed for demonstration experiments, we will evaluate and improve the effectiveness of detection, such as whether it is properly detected.
(3) Post-test: Application of privacy-friendly monitoring technology as a service
Based on point cloud data collected from millimeter-wave sensors, Fujitsu analyzes body movements specific to falls and those that lead to falls. The point cloud data is acquired by radio wave irradiation and reflection from the target person, so unlike camera images, it does not contain information that identifies individuals, making it possible to monitor with consideration for privacy. Wakayama Medical University will conduct a technical evaluation of Fujitsu’s analysis results using knowledge from the medical field, and based on the evaluation results, Fujitsu will make further improvements to its monitoring technology. We aim to develop a monitoring technology service that considers
3. Both roles:
・Fujitsu: Provision of monitoring technology, data collection and analysis, technology improvement based on technology evaluation ・Wakayama Medical University: Technical evaluation and technical improvement proposals from the perspective of medical practice [Image
Fig. Demonstration concept
Proper nouns such as product names listed are trademarks or registered trademarks of their respective companies.
(Note 1) Fujitsu Limited: Headquarters: Minato-ku, Tokyo; President: Takahito Tokita. (Supplement: Depending on the specifications of the publication media and viewing environment, the character “Takashi” may not be displayed correctly. Correctly, “Ichi” is placed above “Nama” in “Takashi”.)
(Note 2) Wakayama Medical University Public University Corporation: Location: Wakayama City, Wakayama Prefecture; Chairman: Kazuhisa Miyashita.
(Note 3) Behavior analysis technology Actlyzer: This technology is based on Fujitsu’s AI image analysis solution “Fujitsu Cognitive Service GREENAGE” (https://www.fujitsu.com/jp/solutions/business- technology/tc/sol/greenages-cs/) has already been commercialized as behavior detection.
(Note 4) All Japan Hospital Association: Headquarters: Chiyoda-ku, Tokyo, President: Yuji Inoguchi. Reported the number of falls and falls at participating hospitals in 2021
(https://www.ajha.or.jp/hms/qualityhealthcare/indicator/42/). (Note 5) Bed-leaving sensor: A device that prevents and counteracts falls, falls, and wandering in hospitals and other facilities. There are types that detect people’s movements using infrared rays, and types that physically detect people’s movements, such as mat sensors and bed sensors.
【 Related Links 】
・Development of new technology for highly accurate estimation of human posture from point cloud data collected by millimeter wave sensors (Fujitsu press release dated July 6, 2022):
・Development of AI technology “Behavior analysis technology Actlyzer” that recognizes various human actions from images (Fujitsu press release on November 25, 2019): https://pr.fujitsu.com/jp/news/2019/11 /25.html
≪Inquiries regarding this matter≫
Research Division Converging Technology Laboratory
Wakayama Medical University Hospital
Department of Medical Information
The product prices, specifications, service details, etc. described in the press release are current as of the date of the announcement. Subject to change without notice. Please note.
Details about this release: