Nonlinear Characteristics Aanalysis And Classification On Acupuncture Point Potential

As the core of traditional chinese medicine(TCM) theory, meridians and acupoints play an important role in disease diagnosis and the process of acupuncture treatment. Acupoints’ electrical characteristics are similar to human bioelectricity in many ways. Therefore the reserch of acupuncture point theory which use the modern biological signal analysis technique from the angle of electrical study will make significance to scientific verification of TCM. diagnosis and treatment of diseases. Current research of acupoints’ potential signals still exists many problems.On one hand, the strong background noise which is caused by the problem of the target signal’s weekness and non-stationary leads to difficulties in the process of signal’s acquisition and denoising.On the other hand, the lack of targeted method for signal processing,feature extraction and pattern classification caused the phenomenon of a low classification accuracy.Consider those above problems,classification of acupoints’ and non-acupoints’ potential signal is studied.The main research results obtained as follows:1) The potential signals of human body have been studied. Potential signals of PC7, PC3, LU5 and non-acupuncture points near them of 20 healthy people are sampled. And then, de-noise the sampled signals with the help of wavelet.2) Chaotic nonlinear characteristics of acupoints’ potential signals have been analyzed. Considering that the bioelectricity signals often have nonlinear characteristics, chaotic theory is applied to analyze acuponits’ potential signal.The time delay is calculated by mutual information method and embed dimension is estimated by Cao method.Then phase spaces of time series are reconstructed.Based on the reconstructed phase spaces, correlation dimension and largest lyapunov index of both acupoints’ and non-acupoints’ potential signals are analyzed.Experiment results reveal that either acupoints’ or non-acupoints’ potential time series have chaotic nonlinear characteristics,and there exists obvious differences of correlation dimension between acupoints’ and non-acupoints’ potential time series.3) A new method which combines wavelet packet technology and nonlinear parameter analysis has carried out.For the acupoints’ and non-acupoints’ potential signals have different degree of difference in different frequency bands, wavelet packet analysis technique which owns excellent local analysis ability is applied in the process of nonlinear analysis. Firsty, potential signals are decomposed to the third layer which has eight frequency bands, then correlation dimensions of the 8 sub band signals which are the local characteristics of potential signals are caculated.4) Classification of acupoints ’and non-acupoints’ potential signals has been reserched. A kind of feature vector is constructed with general nonlinear characteristic parameters and correlation dimensions of each sub-band signal.In order to prove the performance of new feature vector, another feature vector combined by the global nonlinear characteristics is also applied for contrast.Then the grid search method based support vector machine (SVM) is used to classify two groups of feature vector in order to complete the classification of the acupoints and non-acupoints. The classification results show that the feature vector which combine both local and global nonlinear characteristic parameters has a higher and more stable test accuracy.Its average test accuracy is 89.29% which makes a 17.11% increase to the reference group,thus providing a new way to construct feature vector for the classification of acupuncture potential signals.