The Nonlinear Dynamics Analysis Of Heart Sounds Based On The Chaos Theory

Although the progress of people’s living standard has been continuously improved,the incidence of cardiovascular disease remains high.As one of the most significant physiological signal of the cardiac signals,heart sounds can reflect the operating state of the heart.Through the analysis of the characteristics of heart sounds we can extract useful features from the primitive signal.These features will boost the diagnosis of heart disease.But current research of heart sounds still exists many problems.On one hand,the strong background noise which is caused by the problem of the weakness of the target signal and non-stationary leads to difficulties in the process of signal’s acquisition and denoising.On the other hand,it is hard to find the targeted method for signal processing,feature extraction.Meanwhile pattern classification accuracy is too low.Consider those above problems,classification of normal and abnormal heart sounds is studied.The normal and abnormal heart sounds have been collected and studied.We had collected some normal subjects and abnormal subjects in the hospital with the help of LITTMANN stethoscope.And then,de-noise the sampled signals with the help of wavelet.The main findings are as follows:1)Chaotic nonlinear characteristics of heart signals have been analyzed.Considering that the bioelectricity signals often have nonlinear characteristics,chaotic theory is applied to analyzed heart sounds.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 normal and abnormal heart sounds are analyzed.Experiment results reveal that either normal or abnormal heart sounds have chaotic nonlinear characteristics,and there exists obvious differences of correlation dimension between normal and abnormal heart sounds.2)A feature vector construction method based on improved surrogate data method is proposed for the first time.This method combines the surrogate data method with the correlation dimension to get a new feature-normalized sigma variability.This new feature describes the distinct between the normal heart sounds and abnormal heart sounds which is better than correlation dimension.3)Pattern recognition is carried out with the feature vector.Firstly,this paper uses the nonlinear method to analyze all the samples which include the normal and abnormal heart sounds and calculates the time delay,the correlation dimension,the first max lyapunov index and the normalized sigma of the signals.Then support vector machine(SVM)is used to classify two groups of feature vector in order to complete the classification of the normal and abnormal heart sounds.The results shows that the new feature—normalized sigma variability raises the accuracy of the classification.New feature vector has a higher test accuracy which is 86.67%while the other vector’s accuracy is 76.67%.