Study On Extraction And Classification Of Electrical Characteristics Of Some Acupoints Of Human Arms And Identification Of Meridian Model

The study of electrical properties of meridian acupoint can make people more in-depth understanding of meridian acupoint. Found that the specificity of acupoint compared to non-acupoint. which can be used to locate acupoint. Identification the mathematical model of the meridian, understanding the parameters of the model, which makes a preliminary exploration for the future application in the clinical diagnosis of disease using meridian model parameters.Therefore the study of electrical properties of the meridian acupoint has important theoretical and practical significance. This study targeted the electrical properties of Tai-Yin Lung Meridian and a number of acupoints along this meridian, this study includes feature extraction and classification of the acupoint and non-acupoint based on electrical stimulation, the specificity of the acupoint electrocardiosignal compared to non-acupoint. and the exploration on model identification of Tianfu acupoint to Kongzui acupoint part of the Tai-Yin Lung Meridian.By designing an electrical stimulation and signal acquisition experiment on Tai-Yin Lung Meridian,the signal data collected on Kongzui acupoint and a nearby non-acupoint was used to feature extraction of the signal by the method of wavelet packet decomposition and reconstruction, and support vector machine method was employed to the classification of characteristic values acquisition obtained. The results showed that in the case of electrical stimulation of the human body experiment, the Kongzui acupoint and a nearby non-acupoint at the inner side on Tai-Yin Lung Meridian can make a very good classification,indicates that there is a significant difference between acupoint and non-acupoint under electrical stimulation.Two electrocardiosignal acquisition experiments are designed on Chize,Kongzui acupoint and a number of non-acupoint around the acupoint on Tai-Yin Lung Meridian,then studies the characteristics of electrocardiosignal obtained from acupoint distinguished of that from non-acupoint both in the point of view of wavelet-packet decomposition and statistics. First, by means of three-layer wavelet-packet decomposition and reconstruction,then calculates the norms of reconstruction coefficient of the third layer nodes,it is found that the acupoint is greater than the non-acupoint in energy at the low frequency band, the energy of non-acupoint on the meridian line is greater than that of the non-acupoint not on the meridian line at the lowest frequency band. Second, through the statistical analysis of electrocardiosignal obtained, the specificity of statistical quantity of electrocardiosignal is found on Chize, Kongzui acupoint relative to its surrounding four non-acupoints, namely, second-order statistics of electrocardiosignal at the acupoint is significantly higher than its surrounding four non-acupoints,and conclusion obtained above is took to the test using the small-sample test method, results show that the conclusion does have a very high confidence level.Using the experimental data of electrical stimulation for the study of model identification of the Tianfu to Kongzui acupoint part of Tai-Yin Lung Meridian,this paper attempts a variety of identification methods, it is found that the identification results are not satisfactory by using both linear ARMAX and ARX model as identification methods, recognition accuracy of the model is only about80%,using non-linear NARX model identification method is able to reach more than90%,it is indicated that the Tai-Yin Lung Meridian channel is non-linear.