A new method for obtaining the heart sound signals source components

Abstract

This paper introduces a new method which using synchronized acquisition of four-channel heart sounds to acquire a heart sound signal source. First of all, we introduces the generation mechanism of heart sounds in detail, and puts forward that the heart sound signals have the characteristics of time-sharing. Then, according to the basic characteristics of heart sound transmission in the thoracic cavity, we model the mathematical process of heart sound transmission in the thoracic, and propose a mathematical model about heart sound transmission. Finally, according to the mathematical characteristics of the model and the main feature of time-sharing in the heart sound signal, we propose a source separation method for the four-channel heart sound signal in order to obtain the source of heart sounds. The purpose of this algorithm is to eliminate the interaction between simultaneous acquisition of four-channel heart sound signal, which makes the acquired heart sound signal show the characteristics of time difference, energy concentration and low frequency of the spectrum so that we can obtain the source composition of heart sound signal. The experimental results show that the algorithm used in this paper can better obtain single heart sound source components, and single heart sound source components include four major components of the first heart sound (S1), the A and P components of the second heart sound (S2), which have a positive meaning in analyzing the original composition and production mechanism of the heart sound.