Segmentation and Classification of Cardiac Sound Signals and Their Use in the Diagnosis of Heart Disease

Abstract

The PCG Phonocardiogram signal represents the recording of heart sounds. The study of intracardiac hemodynamic makes it possible to understand the nature and origin of these normal and pathological heart sounds. The classification of PCG signal beats into different pathological cases is a very complex recognition task, which has prompted researchers to develop techniques for the automatic classification of cardiovascular diseases for proper diagnosis. In this manuscript, we propose an automatic classification of PCG signals using the Deep Learning ANN algorithm based on PCG signal segmentation to extract features from PCG signals. The results allow us to find the type of disease with an accuracy of 96%.