Study of a Vocal Feature Selection Method and Vocal Properties for Discriminating Four Constitution Types

Abstract

The voice has been used to classify the four constitution types, and to recognize a subject's health condition by extracting meaningful physical quantities, in traditional Korean medicine. In this paper, we propose a method of selecting the reliable variables from various voice features, such as frequency derivative features, frequency band ratios, and intensity, from vowels and a sentence. Further, we suggest a process to extract independent variables by eliminating explanatory variables and reducing their correlation and remove outlying data to enable reliable discriminant analysis. Moreover, the suitable division of data for analysis, according to the gender and age of subjects, is discussed. Finally, the vocal features are applied to a discriminant analysis to classify each constitution type. This method of voice classification can be widely used in the u-Healthcare system of personalized medicine and for improving diagnostic accuracy.

A Novel Method for Classifying Body Mass Index on the Basis of Speech Signals for Future Clinical Applications: A Pilot Study

“…All features were extracted using 5 vowels (A, E, I, O, U) and 1 sentence [38]. For speech feature extraction, we extracted 65 features from the collected data set.…”

Recent Progress in Voice-Based Sasang Constitutional Medicine: Improving Stability of Diagnosis

“…Many statistical pattern recognition methods were applied to SC classification, such as logistic regression [4], linear discriminant analysis [5], and support vector machine [16]. Although many attempts were made in the past, none of them analysed the features from each frame separately.…”

“…Although the accuracies of voice component were low, voice information was still helpful to increase the performance of the integrated model. Another research using the same voice data of Do et al was carried out by Kim et al [5]. They employed linear discriminant analysis as a classifier, and the training accuracies were 51% and 47% for male and female, respectively.…”

Recent progress and prospects in Sasang constitutional medicine: A traditional type of physiome-based treatment

“…Chae et al attempted to determine the relationship between personality assessment and constitution (Chae et al, 2003(Chae et al, , 2009(Chae et al, , 2011(Chae et al, , 2012). Koo and Kim performed studies selecting distinctive features from facial images and voice records in a manner consistent with the principles of the Sasang constitutions (Koo et al, 2009;Kim et al, 2012a). Jang attempted to develop a practical diagnosis using vocal features and improved the reliability (Jang et al, 2013a;b).…”