The Research Of Valvular Disease Diagnosis Based On The Time Series Of Heart Sound System

Heart sound signal has the nonlinear and complexity characteristics. In previous studies, scholars usually modeled it as an ideal linear system, using linear method to analyze this ideal model. However, this is not enough to fully display its characteristics. Chaotic time series is a very important nonlinear system sport mode, with excellent performance in revealing inherent randomness nonlinear processes, thus the subject proposed chaotic time series method to study the heart sound signal of valvular diseases, reflects the inherent laws deeper heart sound signals in essence, using this method to solve the auxiliary diagnosis of heart valve disease problems.This paper describes the mechanism of normal heart sound signals and valvular heart disease heart sound signals, the relationship between the changes of heart sound and pathological changes of heart, and expounds the application of heart sound in the diagnosis of heart valve disease. Heart sound signal is the research object, using wavelet transform method to denoise heart sounds, then using Hilbert transform to extract the denoised heart sounds signal’s envelope, then analysis the autocorrelation of envelope to find out the cardiac cycle length, this length is used to segment the original heart sounds; Chaotic time series is used to analysis valvular heart sound signal, quantitative calculate chaos characteristic values such as the correlation dimension and largest Lyapunov index.Heart sound signal analysis system is based on the LabVIEW platform designed by this paper:including the heart sound signals acquisition, preprocessing, chaotic time series analysis. Using this system to preprocess normal heart sounds and several common heart valve disease heart sound signal, the results showed that the heart is a typical chaotic system, the chaotic characteristics of normal between valvular heart disease signals have a very significant difference, in which the correlation dimension and the largest Lyapunov exponent of the pathological changes of heart sound signals are higher than normal heart sound signals, that means, compared with normal heart sound signals, the pathological changes of heart sound signal has high complexity, while the differences between the various valvular disease can lay the foundation for the next classification. Chaotic characteristics for heart sound signals have an important role. In addition, the develop products used on this platform can run on the user’s computer system, non-LabVIEW development environment and use the user’s display shows the results can provide a supporting role for doctors and patients.