文章基本信息

标题：Support Vector Machines for Classification of Temporomandibular Disorders from Facial Pattern Values
本地全文：下载
作者：Mansoureh Ghodsi ; Saeid Sanei
期刊名称：Journal of Data Science
印刷版ISSN：1680-743X
电子版ISSN：1683-8602
出版年度：2011
卷号：9
期号：3
页码：373-388
出版社：Tingmao Publish Company
摘要：The aim of this study is to develop a method for detection oftemporomandibular disorder (TMD) based on visual analysis of facial movements.We analyse the motion of colour markers placed on the locations ofinterest on subjects faces in the video frames. We measured several featuresfrom motion patterns of the markers that can be used to distinguish betweendierent classes. In our approach, both static and dynamic features are measuredfrom a number of time sequences for classication of the subjects. Ameasure of nonlinear dynamics of the variations in the movement of colourmarkers positioned on the subjects faces was obtained via estimating themaximum Lyapunov exponent. Static features such as the number of outliersand kurtosis have also been evaluated. Then, Support Vector Machines(SVMs) are used to automatically classify all the subjects as belonging toindividuals with TMD and healthy subjects.
关键词：Temporomandibular disorder; maximum Lyapunov exponents;support vector machine.