A Method for Automated J Wave Detection and Characterisation Based on Feature Extraction

Abstract

J waves are low-amplitude, high-frequency waveforms which look like notches or slurs appearing in the descending slope of the terminal portion of the QRS complex in electrocardiogram (ECG). J wave is related to early repolarization syndrome (ERS), idiopathic ventricular fibrillation (IVF) or Brugada syndrome (BrS). Patients with the three syndromes are susceptible to cardiac arrhythmias and sudden cardiac death. Accordingly, J wave detection presents a non-invasive marker for some cardiac diseases clinically. In this report, 12-lead ECG record with higher signal-to-noise ratio (SNR) is formed using multi-beat averaging method. Then, we define five feature vectors including three time-domain feature vectors and two wavelet-based feature vectors. Those feature vectors are processed by principle component analysis (PCA) to reduce its dimensionality. Finally, a Hidden Markova model (HMM), trained by a proper set of these feature vectors, is employed as a classifier. Compared with other existing methods, the results show the proposed method reveals high evaluation criteria (accuracy, sensitivity, and specificity) and is qualified to detect J waves, suggesting possible utility of this approach for defining and detection of other complex ECG waveforms.