Mitral regurgitation PCG-signal classification based on adaptive Db-wavelet

Abstract

An auscultation method is an important diagnostic indicator for hemodynamic anomalies. Heart sound classification and analysis play an important role in the auscultative diagnosis. This study uses a combination of wavelet decomposition of continuous transform for systolic and diastolic pattern; with a combination of system identification based on ARMAX techniques to efficiently extract the features for pre-processed heart sound in order to classify and estimate hemodynamic properties of the heart. A system was developed for the interpretation of heart sounds acquired by phonocardiography using wavelet decomposition. As an advanced method to integrate the automated auscultation computer aided diagnosis (CAD). The task of feature extraction was performed using three methods: time domain feature, Short-Time Fourier Transforms (STFT) and Debauchies methods and ARMAX. The performances of these feature extraction methods were then compared and interpreted. © 2008 Springer-Verlag.