Research on Objective Auscultation of TCM Using Wavelet Packet and Support Vector Machine

Jianjun Yan, Xiaojing Shen

The goal of this study is to provide objective analysis and quantitative research for the auscultation in Traditional Chinese Medicine (TCM) using wavelet packet transform (WPT) and support vector machine (SVM). Based on WPT, the voice signals are decomposed into six layers wavelet coefficients. This paper proposed Shannon entropy as feature parameter extracted from wavelet packet coefficients to make analysis of health, qi-vacuity and yin-vacuity subjects. Then the feature values used as vectors were put into SVM to be trained and predicted, and the classification results showed that the overall accuracy of the health group, qi-vacuity group and yin-vacuity group reached to 80.84%. It is proved that our method is effective for auscultation research of TCM.