Detection and identification algorithm of the S1 and S2 heart sounds

Abstract

This article attempts to contribute to heart abnormalities detection by suggesting a segmentation algorithm of S1 and S2 heart sounds. It describes the steps followed and the results achieved during the realization of the first PASCAL (Pattern Analysis, Statistical Modelling and Computational Learning) Classifying Heart Sounds Challenge. The performed work leads to improved results, compared to the results found by the three finalists of this challenge, using the envelope of Shannon energy and other methods, to detect and identify first and second heart sound of phonocardiogram signals.