Proposition of a Time Transform for the Detection of Cardiac Sounds in the Phonocardiogram of a Healthy Patient

Collecting patient health data is a task that has become very common today. Indeed, thanks to everyday objects such as watches, cell phones, etc., collecting data on a patient is not a very complex task. However, to collect and process this data with such simple objects, it is important to develop simple and effective methods for the processing the data collected. This article is part of this dynamic and proposes a temporal transformation of the phonocardiogram signal (PCG) for the detection of cardiac sounds. This transform is based on the derivative of the square of the PCG signal. This approach has been confronted with other existing temporal transformations. The results obtained show that the proposed method highlights cardiac noises in the PCG signal. Also, this approach guarantees a better difference in amplitude between heart sounds S1 and S2 (0.5267 for the method proposed against 0.339 for the method of the literature guaranteeing the best difference of amplitude namely the Energy of the Signal) which favors the identification of each type of noise and therefore an extraction of the cardiac constants by simple methods such as thresholding.