ECG Beats Classification Based on Ensemble Feature Composed of Independent Components and QRS Complex Width

A novel method is proposed in this paper for the feature extraction of electrocardiogram (ECG). Different with other algorithms, the proposed method utilizes independent component analysis (ICA) and wavelet transform to get an ensemble feature composed of ICA-based features and the QRS complex width feature. The QRS complex is the most characteristic waveform of an ECG signal and its width has been a diagnostic criterion of cardiac arrhythmia. Therefore, our ensemble feature consisting of QRS complex width would provide much more information on cardiac diseases than other methods. The formed ensemble feature is fed into an artificial neural networks classifier. To validate the proposed method, we applied it to the MIT-BIH arrhythmia database. The experimental results have shown the effectiveness of the proposed method.