Predicting imminent episodes of ventricular tachyarrhythmia using an entropy-based feature in the EMD domain

Efficient prediction of Ventricular tachyarrhythmia (VTA)particularly ventricular tachycardia (VT) and ventricular fibrillation (VF) is very important for clinical purpose, as they are the most serious cardiac rhythm disturbance that can be life threatening. A reliable predictor of an imminent episode of ventricular tachycardia that could be incorporated in an implantable defibrillator capable of preventive therapy would have important clinical utilities. However, there are several methods which have separated pre arrhythmia and control subjects, but there are only a few methods to predict VF/VT by tracing whole the signal from beginning to end and providing us a quantitative predictor by the time. In this paper we tried to present an quantitative predictor by finding an entropy-based pattern in T-wave of ECG signals which has the most important role in ventricular activity of heart using Empirical Mode Decomposition (EMD). As this pattern rarely occurs in control records it can be considered as a useful index for probability occurrence of VF/VT. so physicians can apply an aptly timed electrical shock or it can be used to improve Implantable cardiac defibrillators and thus it yields to increase the probability of saving many cardiac patients. The developed algorithm can reach sensitivity of 84% and specificity of 93% in online/VT prediction.