ECG arrhythmia classification based on variational mode decomposition, Shannon energy envelope and deterministic learning

Document Type


Date of Original Version



Electrocardiography (ECG) signals play an important role in the cardiac disorder diagnosis and arrhythmia detection since they reflect all the electrical activities of the heart and include information about heart function and heart conditions. Due to the subtle alterations in the amplitude, duration and morphology of the ECG, the development of computer-based intelligent system in the arrhythmia diagnosis field is attractive in terms of the amount of data and the importance of the data it contains for the classification of heartbeats from different types of arrhythmias. In the present study we propose a novel technique for the automatic detection of cardiac arrhythmia with one-lead ECG signals based upon variational mode decomposition (VMD), Shannon energy envelope, phase space reconstruction (PSR) and deterministic learning theory. First, VMD is employed to decompose the ECG signals into different intrinsic modes, in which the first four intrinsic modes contain the majority of the ECG signals’ energy and are considered to be the predominant intrinsic modes. Second, Shannon energy is used to extract the characteristic envelope of predominant intrinsic modes. Third, phase space of the Shannon energy envelope (SEE) is reconstructed, in which properties associated with the nonlinear ECG characteristics are preserved. Three-dimensional (3D) phase space reconstruction (PSR) together with Euclidean distance (ED) has been utilized to derive features, which demonstrate significant difference in ECG system dynamics between normal versus abnormal individual heartbeats. Fourth, neural networks are then used to model, identify and classify ECG system dynamics between normal (healthy) and arrhythmic ECG signals based on deterministic learning theory. Finally, experiments are carried out on the MIT-BIH arrhythmia database to verify the effectiveness of the proposed method, in which 626 ECG signal fragments for one lead (MLII) from 28 persons of five classes of heartbeats were extracted. These five classes are normal sinus rhythm (NSR), premature ventricular contraction (PVC), paced beat (PB), left bundle branch block (LBBB), and right bundle branch block (RBBB). By using the 10-fold cross-validation style, the achieved average classification accuracy is reported to be 98.72%. Experimental results verify the effectiveness of the proposed method and indicate that it has the potential to serve as a candidate for the automatic detection of myocardial dysfunction in the clinical application.

Publication Title

International Journal of Machine Learning and Cybernetics