Comparative Analysis of S1 and S2 Sounds Localization Methods in a Phonocardiogram

The phonocardiogram (PCG) signal is an acoustic signal resulting from the mechanical functioning of the human heart. Although neglected today, its use with various technological advances brings a lot of information about the state of the heart. The aim of our work is to identify the most efficient method for the detection of B1 and B2 sounds in a PCG signal. Three methods were analyzed throughout our work. These are the Shannon Energy, the Hilbert Transform and the Continuous Wavelet Transform. Our analysis was based on 90 PCG recordings from normal subjects from the Pascal Classifying Heart Sound Challenge database. Our results showed that the Shannon energy exhibits the least detection error. Indeed, out of 1420 noises to be detected on the 90 recordings of PCG signals used, the implementation of the Shannon energy that we made presented 59 detection errors while the Hilbert transform and the continuous wavelet transform presented respectively 84 and 185 detection errors.