A Novel Designing System of Speech Aid from Imaging Analysis of Oral Cavity Volume for Cerebrovascular Accident and Oral Cancer Patients

正常人和脑血管意外言语障碍患者参加了这项研究。受试者在发[ta]、[na]、[ka]、[ra]音前保持舌位时进行头部磁共振成像。根据这些MR图像，在140个点处测量每个受试者的舌表面和腭之间的距离。用正常人口腔平均容积数据库（[ta]、[na]、[ka]、[ra]）建立口腔平均容积数据库。然而，脑血管意外患者的MR成像质量较差。由于脑血管意外患者在发音时的MR图像拍摄时间（一次发音3分钟）过长，无法在发音前保持舌位，因此，正常人发音时口腔容积的数据库可作为脑血管意外或口腔癌患者发音时口腔容积分析的后续研究的重要信息。从即时为这些患者设计语音辅助器的角度来看，预计基于语音中的口腔体积的数据库是非常需要的。

Normal subjects and patients with speech disorder with cerebrovascular accident participated in this study. The magnetic resonance(MR) imaging of the head of the subjects were taken during the subject kept the tongue position right before pronounce [ta],[na],[ka],[ra]. From these MR images, the distance between the tongue surface and the palate was measured at 140 points for each subject. The data base of average oral cavity volume for each pronounce ([ta],[na],[ka],[ra]) were made from the data of normal subjects. However, it was difficult to take fine MR Images of the patients with cerebrovascular accident. Because the time of taking MR images (3 minutes for one pronounce) seemed to be too long for the cerebrovascular accident patients to keep the tongue position right before pronounce.The data base of oral cavity volume during pronounce in the normal subjects could be an important information for continuous studies about analysis of oral cavity volume during pronounce in the cerebrovascular accident or oral cancer patients. From the view point of, for instantly, designing the speech aid for those patients, it is predicted that the data base based from the oral cavity volume in speech could be highly needed.