A generic end-point detection algorithm for cardiac sounds

A generic algorithm for the segmentation of cardiac sounds is presented in the paper. It is based on the extraction of simple parameters from the heart beat. The proposed approach does not require synchronization with electrocardiogram. It is designed to work even in the case of cardiac anomalies. Segmentation represents a first step toward an automatic diagnosis of cardiac diseases. All procedures were implemented in a software tool. Segmentation algorithm was tested with actual cardiac sounds, showing a good success rate