A heart sound signal classification and recognition method

Abstract

The invention discloses a method for classifying and identifying heart sound signals. Firstly, discrete wavelet decomposition is performed on preprocessed heart sound signals to obtain detailed wavelet coefficients and approximate wavelet coefficients of different frequency bands; and then the detailed wavelet coefficients and approximate wavelet coefficients are sequentially calculated Normalize the average Shannon energy envelope and autocorrelation function to obtain the autocorrelation feature of the detail wavelet coefficient envelope and the autocorrelation feature of the approximate wavelet coefficient envelope. Approximate autocorrelation features are used to perform nonlinear feature dimension reduction, and the detailed features and approximate features after dimension reduction are fused to obtain fusion features; finally, the fusion features are used as the input of the support vector machine for classification and recognition; adopting the present invention can avoid heart sound signal Segmentation processing is performed to improve the accuracy of heart sound feature extraction, which has a positive effect on the analysis and feature extraction of pathological heart sounds.

Description

technical field [0001] The invention relates to a heart sound signal classification and recognition method, in particular to a heart sound classification and recognition method based on an autocorrelation function and a local linear embedding algorithm without segmentation. Background technique [0002] Heart sounds mainly come from the opening and closing of heart valves and the turbulent flow of blood. It contains physiological information about various parts of the heart such as atria, ventricles, great vessels, cardiovascular and the functional status of each valve, and can reflect the mechanical activities and mechanisms of the heart. , with biological characteristics such as universality, stability, uniqueness and collectability. In modern medicine, heart sound auscultation is often used as a means of preliminary diagnosis of heart disease. However, heart sound is a very weak, low-frequency signal, and the results of auscultation are easily affected by the doctor's hea...

AI Technical Summary

Problems solved by technology

For the segmentation of heart sounds, ECG signals are often used to assist the segmentation of heart sounds. However, using ECG signals to assist segmentation requires simultaneous recording and synchronous processing of ECG signals and heart sound signals, which is very inconvenient; However, there are two problems in segmenting by this method: (1) In the case of pathological heart sounds or background noise, there will inevitably be omission of peak points or detection errors; (2) Using peak points to determine the period of heart sound signals The premise is that the systolic period of the heart sound signal is shorter than the diastolic period, but this condition is not always true, especially for some pathological heart sounds

Images