Development of Swallowing Monitor by Artificial Intelligence of Cervical Auscultation

2021 Fiscal Year Final Research Report

• Project/Area Number: 19K10424
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 57080:Social dentistry-related
• Research Institution: Nagasaki University (2020-2021); Iwate Medical University (2019)
• Principal Investigator: Tamada Yasushi 長崎大学, 病院(歯学系), 助教 (50633145)
• Co-Investigator(Kenkyū-buntansha): 佐々木 誠 岩手大学, 理工学部, 准教授 (80404119)
• Project Period (FY): 2019-04-01 – 2022-03-31
• Keywords: 人工知能 / 頸部聴診法 / 摂食嚥下リハビリテーションｎ / 嚥下モニター

Outline of Final Research Achievements

When a patient has dysphagia, the amount of food or drink swallowed at one time is often regulated by limiting the amount of food or drink in a mouthful to prevent aspiration. However, even if the bite size is limited, if there is still food or drink in the oral cavity due to cognitive dysfunction or other reasons, the amount swallowed at one time may increase with the addition of a second bite. Therefore, we thought that a non-invasive method to estimate the amount of food or drink to be swallowed at one time was needed. In this study, data from a laryngeal microphone and electromyograph attached to the neck were digitally processed, and the data was learned by artificial intelligence, making it possible to estimate with high accuracy the amount swallowed at a time and characteristics of the way the person swallows. In addition, an artificial intelligence program specialized for analyzing the state of swallowing was developed and presented at academic symposiums.

Free Research Field

摂食嚥下リハビリテーション

Academic Significance and Societal Importance of the Research Achievements

マンパワーの不足、介助者の知識・技術不足などを理由に摂食時の正しい見守りや介助が行われない場合がある。侵襲の無いモニターの開発は、これらの問題解決法の一つとなる。本研究で開発した喉頭マイクおよび筋電計からのデータを人工知能が学習し、１回嚥下量や嚥下の特徴を高精度で推測するシステムは、今後も続く超高齢社会の食事場面での問題を解決する１手法となり得る。