Abstract:Phonocardiogram (PCG) plays an important role in evaluating many cardiac abnormalities, such as the valvular heart disease, congestive heart failure and anatomical defects of the heart. However, effective cardiac auscultation requires trained physicians whose work is tough, laborious and subjective. The objective of this study is to develop an automatic classification method for anomaly (normal vs. abnormal) detection of PCG recordings without any segmentation of heart sound signals. Hybrid signal processing and artificial intelligence tools, including tunable Q-factor wavelet transform (TQWT), variational mode decomposition (VMD), phase space reconstruction (PSR) and neural networks, are utilized to extract representative features in order to model, identify and detect abnormal patterns in the dynamics of PCG system caused by heart disease. First, heart sound signal is decomposed into a set of frequency subbands with a number of decomposition levels by using the TQWT method. Second, VMD is employed to decompose the subband of the heart sound signal into different intrinsic modes, in which the first four intrinsic modes contain the majority of the heart sound signal's energy and are considered to be the predominant intrinsic modes. They are selected to construct the reference variable for analysis. Third, phase space of the reference variable is reconstructed, in which the properties associated with the nonlinear PCG system dynamics are preserved. Three-dimensional PSR together with Euclidean distance has been utilized to derive features, which demonstrate significant difference in PCG system dynamics between normal and abnormal heart sound signals. Finally, PhysioNet/CinC Challenge heart sound database is used for evaluation and the synthetic minority over-sampling technique method is applied to balance the datasets. By using the 10-fold cross-validation style, experimental results demonstrate that the proposed features with dynamical neural networks based classifier yield classification performance with sensitivity, specificity, overall score and accuracy values of 97.73\(\%\), 98.05\(\%\), 97.89\(\%\), and 97.89\(\%\), respectively. The results verify the effectiveness of the proposed method which can serve as a potential candidate for the automatic anomaly detection in the clinical application.

Envelope extraction algorithm and phonocardiogram signal application based on wavelet transform

[Study of Biometric Identification of Heart Sound Base on Mel-Frequency Cepstrum Coefficient].

Improved Methods for Detecting Main Components of Heart Sounds.

Comparison Of Six Envelope Extraction Methods Based On Abnormal Heart Sounds

Application of Simplicity-based Envelope Extraction in Heart Sound Segmentation

Heart sound signal envelope extraction based on modified Viola integral method

Quantitative analysis of disease features of a class of pulse signals based on wavelet

A Robust Envelope Extraction Algorithm for Cardiac Sound Signal Segmentation

The Study of Three-Dimensional Phonocardiogram Using Wavelet Transform Method

Envelope Extraction and Recognition of Heart Sounds Based on Mathematical Morphology

Heart sound feature extraction and classification based on integration of wavelet packet analysis and chaos theory

Time-frequency Analysis of Heart Sounds Based on Continuous Wavelet Transform

Refined Analysis of Heart Sound Based on Hilbert - Huang Transform

Identification of the Normal and Abnormal Heart Sounds Using Wavelet-Time Entropy Features Based on Oms-Wpd

Envelope Extraction Of Anaesthesia Breathing Sound Signal On Hilbert Huang Transform

Detection And Classification Of Abnormities Of First Heart Sound Using Empirical Wavelet Transform

Deep Wavelets for Heart Sound Classification

Analysis Method and Application of Heart Sounds

Classification of Heart Sound Signals Based on AR Model.

A new approach for the detection of abnormal heart sound signals using TQWT, VMD and neural networks

Analysis of Heart Sound Based on Wavelet Transform