Емпіричне дослідження ефективності використання методів вейвлет-перетворень для аналізу фонокардіограми серця

Abstract

This research presents results of analyzing effectiveness of number of wavelets on the task of preprocessing phonocardiogramm (PCG) for further statistical analysis and classification as PCG of heart ill with aortic stenosis or PCG of healthy heart. Two types of samples (ill and healthy heart PCGs) were subjected to discrete-wavelet transform with Haar, Daubeches, symlets and Discrete Meyer wavelets as basis function, resulting sequences were statistically analysed. Level two Daubeches wavelet found out to be best suited for this particular task.