Development of an objective evaluation tool that captures signs of dementia in the elderly using non-contact gait measurement technique and artificial intelligence

2023 Fiscal Year Final Research Report

• Project/Area Number: 21H03280
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 58080:Gerontological nursing and community health nursing-related
• Research Institution: Kindai University
• Principal Investigator: Kurita Koichi 近畿大学, 工学部, 教授 (90455171)
• Project Period (FY): 2021-04-01 – 2024-03-31
• Keywords: 歩行計測 / 認知症 / 非接触計測 / 静電誘導

Outline of Final Research Achievements

The purpose of this research was to develop a technique to detect the signs of dementia in elderly people from the decline in walking function. To achieve this, we first developed an ultrasensitive electrostatic induction sensor. Furthermore, using the walking signals detected by the electrostatic induction sensor as learning data, deep learning was performed to detect left-right asymmetry in walking movements, and hemiplegia walking movements, and to identify the degree of disability. To simulate gait disorders, an ankle weight was attached to the right ankle of a healthy person, and gait signals for four tasks were detected depending on the weight of the ankle weight. As a result, it was revealed that the average classification accuracy rate using a convolutional neural network was 83.0%. As a result, we were able to confirm that the slight asymmetry between left and right gait that appears during walking motion can be easily detected without contact.

Free Research Field

計測工学

Academic Significance and Societal Importance of the Research Achievements

本研究は、歩行動作により誘起される静電誘導電流を検出することにより、認知症や片麻痺等の高齢者の歩行機能の低下をAIを用いて識別する手法に学術的意義がある。歩行動作により誘起された静電誘導電流には、被験者の足の接地・離地の際の床に対する足裏接触面積の時間微分に相当する詳細な情報が含まれている。提案手法を用いることで、非接触かつ無装着で歩行動作の僅かな差異を簡便に得ることができることを明らかにした。これにより、従来の歩行パラメータによる評価に比べて高精度な歩行動作の検出が可能となり、高齢者の健康状態の変化の「兆し」を捉える客観的評価ツールとして期待でき、社会的な意義があると考えている。