Experimental analysis of automatic discrimination performance between simulated bruxism and non-bruxism using electromyography and machine learning

2023 Fiscal Year Final Research Report

• Project/Area Number: 20K23107
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Multi-year Fund
• Review Section: 0907:Oral science and related fields
• Research Institution: Okayama University
• Principal Investigator: Omori Ko 岡山大学, 大学病院, 医員 (30884879)
• Project Period (FY): 2020-09-11 – 2024-03-31
• Keywords: 筋電図 / ブラキシズム / 機械学習 / 生体情報

Outline of Final Research Achievements

It remains questionable whether bruxism can be reliably diagnosed using conventional evaluation criteria. Therefore, we wondered whether it would be possible to objectively distinguish electromyograms during different types of bruxism, swallowing, scratching, and body movements. In this context, we attempted to discriminate electromyograms by applying a method to discriminate changes in vector values (feature values) by converting electromyograms into high-dimensional vectors.
As a result, it was shown that this classification system can discriminate teeth contact bruxism from non-bruxism with high accuracy using masseter muscle EMG. In addition, it was shown that an analysis model that included bilateral infrahyoid muscles and skin-transmitted sound further improved the accuracy of discrimination.

Free Research Field

歯科補綴学

Academic Significance and Societal Importance of the Research Achievements

現在，簡易筋電計による睡眠時の筋活動測定が広く行われている。そのため，本研究成果を基盤としてブラキシズムの詳細な測定精度の向上，社会実装が可能となった暁には，歯への機械的負荷を定量的に把握することが可能となる。これにより補綴装置の予後予測や適応症の診断，歯根破折リスクの診断が可能となることから，歯科臨床を大きく改変する可能性を有している。
また，睡眠時/覚醒時ブラキシズムの生理学的理解や各種の治療法への反応性について検討を行う際にも，本研究結果は評価方法として活用できることから，口腔生理学，睡眠歯学，口腔運動学への学問的貢献も大きいと言える。