Prediction of ventricular tachycardia by a neural network using parameters of heart rate variability

In this paper, we propose a classifier that can predict ventricular tachycardia (VT) events using artificial neural networks (ANNs) trained with parameters from heart rate variability (HRV) analysis. The Spontaneous Ventricular Tachyarrhythmia Database (Medtronic Version 1.0), comprising 106 pre-VT records and 126 control data, was used. Each data set was subjected to preprocessing and parameter extraction. After correcting the ectopic beats, data in the 5 minute window prior to the 10 second duration of each event was cropped for parameter extraction. Extraction of the time domain and non-linear parameters was performed subsequently. Two-thirds of the database of extracted parameters was used to train the ANN, and the remainder was used to verify the performance. ANNs for classifying the VT events was developed, and the sensitivities of the ANN was 82.9% (71.4% specificity). The normalized area under the receiver operating characteristic (ROC) curve of the ANN was 0.75.