A new learning and classification framework for the detection of abnormal heart sound signals using hybrid signal processing and neural networks

The objective of this study is to develop an adaptive learning and classification framework for anomaly (normal vs. abnormal) detection of Phonocardiogram (PCG) recordings without any segmentation of heart sound signals. First, heart sound signal is decomposed into a set of frequency subbands with a number of decomposition levels by using the tunable Q-factor wavelet transform (TQWT) method. Second, variational mode decomposition (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. Third, three-dimensional (3D) phase space reconstruction (PSR) together with Euclidean distance (ED) has been utilized to derive features. Fourth, an adaptive learning and classification framework is constructed based on deterministic learning theory to model, identify and classify the normal and abnormal patterns in the dynamics of PCG system between normal people and patients with heart diseases. Finally, PhysioNet/CinC Challenge heart sound database is used for evaluation. By using the 10-fold cross-validation style, the proposed method achieves the classification performance with sensitivity, specificity, overall score and accuracy values of 97.46%, 97.67%, 97.57%, and 97.56%, respectively.