2019 Fiscal Year Final Research Report

Development of the cleft palate speech assessment and treatment system using a new neural network theory

Research Project

PDF

Project/Area Number	17H04407
Research Category	Grant-in-Aid for Scientific Research (B)
Allocation Type	Single-year Grants
Section	一般
Research Field	Surgical dentistry
Research Institution	Kagoshima University
Principal Investigator	NAKAMURA Norifumi 鹿児島大学, 医歯学域歯学系, 教授 (60217875)
Co-Investigator(Kenkyū-buntansha)	上田裕市熊本大学, 大学院先端科学研究部(工), 教授 (00141961) 手塚征宏鹿児島大学, 医歯学域歯学系, 助教 (50759777) 坂田聡熊本大学, 大学院先端科学研究部(工), 助教 (80336205) 三浦尚子鹿児島大学, 医歯学域附属病院, 言語聴覚士 (50715331)
Project Period (FY)	2017-04-01 – 2020-03-31
Keywords	口蓋裂 / 異常構音 / 言語治療 / ニューラルネットワーク / 音声可視化
Outline of Final Research Achievements	In this study, in order to analyze and visualize the speech characteristics of cleft palate patients, we created an editing tool for 50 words of articulation test, and built a prototype system with recording, editing, analysis, and visualization functions. As a result of visualizing the 5 vowels in color with the speech data of 10 cleft palate children and 5 healthy children, and analyzing the relationship between the nasalance score and nasopharyngeal endoscopic findings, it was found that the colors of 5 vowels correlated with the nasalance score. Furthermore, using the constructed neural network (NN) system, we observed changes in articulation characteristics of eight cleft palate children before and after fistula closure. As a result, it has become possible to visualize that as the nasopharyngeal closure function, which was associated with fistula closure, was improved, displacement of the articulation place were gradually improved and closed to that of a healthy subjects.
Free Research Field	口腔外科
Academic Significance and Societal Importance of the Research Achievements	口蓋裂術後の言語異常に対する予知性の高い言語訓練法は未だない。言語訓練には、患者の異常言語の認識（弁別）と中枢におけるフィードバック機構が良好に働くことが重要であるが、しばしば聴覚異常を有する患者では聴覚的なフィードバックを活用することは困難となる。今回開発した新規ニューラルネットワーク(NN)システムは口蓋裂児の鼻咽腔閉鎖機能やその変化に合わせた調音位置の変動を詳細に観察可能にできたことで、口蓋裂における異常構音の発生メカニズムの解明やより効果的な視覚フィードバックを応用した言語訓練の開発が期待できる。音声可視化システムは、口蓋裂児の開鼻声の程度ならびに構音異常の診断に有用なツールといえる。