Ventricular Tachyarrhythmia Prediction based on Heart Rate Variability and Genetic Algorithm

Abstract

Predicting ventricular tachyarrhythmia (VTA) provides opportunities to reduce casualties due to sudden cardiac death. However, prediction accuracy is still need improvement. In this paper, we propose a method that can predict VTA events using support vector machine (SVM) that trained with HRV features from heart rate variability (HRV). The Spontaneous Ventricular Tachyarrhythmia Database (Medtronic Version 1.0), comprising 106 pre-VT records, 26 pre-VF records, and 135 control data, is used.  Fifty percent of the data was used to train the SVM, and the remainder was used to verify the performance. Each data set was subjected to preprocessing and HRV feature extraction. After correcting the ectopic beats, 5 minutes RR intervals prior to each event was cropped for feature extraction. Extraction of the time domain, spectral, non-linear and bispectrum features were performed subsequently. Furthermore, both t-test and genetic algorithm (GA) were used to optimize the HRV feature subset. With optimized feature subset by GA, proposed method of current work able to outperform previous works with 77.94%, 80.88% and 79.41 % for senstivity, specificity and accuracy respectively.