Development of tele-rehabilitation technology for dementia prevention using AI

2023 Fiscal Year Final Research Report

• Project/Area Number: 21K19741
• Research Category: Grant-in-Aid for Challenging Research (Exploratory)
• Allocation Type: Multi-year Fund
• Review Section: Medium-sized Section 59:Sports sciences, physical education, health sciences, and related fields
• Research Institution: Niigata University of Health and Welfare
• Principal Investigator: Noto Shinichi 新潟医療福祉大学, リハビリテーション学部, 教授 (00339954)
• Project Period (FY): 2021-07-09 – 2024-03-31
• Keywords: アルツハイマー病 / AI（人工知能） / 音声 / 会話 / 機械学習 / リハビリテーション

Outline of Final Research Achievements

In this study, we attempted to develop an artificial intelligence (AI) to discriminate the differences between the speech of Alzheimer's disease patients and healthy older adults by analyzing the speech of both groups. In the experiment, Alzheimer's disease patients and ordinary older adults were asked to perform a cognitive function test, and their conversations were recorded and analyzed by a computer using a machine learning method. As a result, it was found that speech distortion occurred in Alzheimer's disease patients, and AI was able to determine the possibility of Alzheimer's disease by detecting the distortion. On the other hand, since the number of cases analyzed in this study was only about 100 for each of the two groups, further machine learning was considered necessary to improve the performance of the AI.

Free Research Field

リハビリテーション医学

Academic Significance and Societal Importance of the Research Achievements

本研究の最大の成果は，会話の音声という非侵襲的な検査によって，アルツハイマー病発症の可能性を予見できる可能性を示したことである．アルツハイマー病の早期発見には体液バイオマーカーの手がかりとした検査が有効とされているが，時間とコストがかかる．またPETによる検査も同様でスクリーニングには適していない．本研究がAI（人工知能）の開発を目的としたことも時宜を得ており，これが実用化されれば，安価でアルツハイマー病の早期発見やリハビリテーションの効果判定に利用可能となる．社会的にも，例えば，金融機関などにおいて資産管理能力の判定に用いることを通して，認知機能が低下する前に相続対策に役立てることができる．