Heart Sound Analysis for Diagnosing Cardiovascylar Disorder

Abstract

Assisting physicians in the auscultation of the patients by providing less costly, automatic and accurate signal processing module for the heart event detection and recognition is of great importance. This paper presents a complete heart sound analysis system covering from the segmentation of beat cycles to the final determination of heart conditions. A basic task for the diseases diagnosis from the phonocardiogram is to detect the exact timing location of the events present in the cardiac cycle, especially in pathological cases. This paper also presents a new algorithm for segmentation of S1 and S2 heart sounds without using ECG as a reference signal. The decision making process is insured by support vector machine (SVM) classifier utilizing the features based on wavelet coefficients so that the signals of PCG are classified into five categories: normal heart sound, AS, AI, MS and MI using a small training dataset and tested with an enormous testing dataset to show the generalization capability of the scheme.