Single-channel heart and lung sound separation method and system

Abstract

The invention discloses a single-channel heart and lung sound separation method and system. The method comprises the following steps of step 1, processing the collected heart and lung sound mixed signals x(t) to obtain the time-frequency spectrum X of the heart and lung sound mixed signals x(t); step 2, taking the column average of 20 Hz to 100 Hz portions of the time-frequency spectrum X and obtaining an amplitude model time series template H0 of heart sounds; step 3, constructing heart sound time series structure regularization terms according to the amplitude model time series template H0 of the heart sounds, performing non-negative matrix factorization on the time-frequency spectrum X, extracting the estimation Xc of the time-frequency spectrum of the heart sound signals and the estimation Xr of the time-frequency spectrum of lung sound signals; and step 4, reconstructing time domain signals of the heart sounds and the lung sounds according to the estimation Xc of the time-frequency spectrum of the heart sound signals and the estimation Xr of the time-frequency spectrum of the lung sound signals. The single-channel heart and lung sound separation method and system can effectively separate overlapping components of heart sounds and lung sounds in a frequency domain, require only single-channel heart and lung sound mixed signals as input to achieve effective separation of heart and lung sounds, have strong clinical applicability, do not need clustering, and have stable effect of separating the heart and lung sounds.

Description

technical field [0001] The present invention relates to the technical field of heart-lung sound separation, in particular to a single-channel heart-lung sound separation method and system Background technique [0002] Heart sound and lung sound signals contain rich cardiopulmonary health information. Cardiopulmonary sound signal detection is of great significance for the diagnosis and prevention of cardiopulmonary diseases. Under clinical conditions, the heart sound and lung sound signals collected by a stethoscope placed on the surface of the chest cavity are often mixed together and mixed with external environmental noise, which is not conducive to doctors' clinical diagnosis. Therefore, it has important clinical significance and practical value to separate the clean heart sound signal and lung sound signal from the collected mixed signal and eliminate the interference of environmental noise. [0003] Generally, the frequency range of heart sounds is 20Hz-600Hz, while th...

AI Technical Summary

Problems solved by technology

In view of the above advantages of NMF, researchers have proposed a method for separating cardiopulmonary tone-deafness based on NMF basis function clustering, but the clustering process of this method is easily affected by singular points far away from the center, and the timing of heart sounds is not used. structure, making the effect of heart-lung sound separation unstable, which is not conducive to practical application

Images