Early Detection and Classification of Cardiac Ventricular Arrhythmia

This study proposes an algorithm for predicting and classifying fatal ventricular arrhythmias (VA) by applying the fuzzy c-means clustering (FCM) and phase-space diagram. A feature index, based on the box-counting technique applied to ECG phase-space trajectories, predicts VA 298.3 seconds before onset with 98.4% accuracy. This proposed method has shown the possible outcome for predictingthe impending VA prior to actual occurrence in time sufficientlyand also it shows the possibilityof classification of VAs type. This covers a novel result compared to existing approaches that is useful for clinical practice that may be proved to be helpful to preempt the fatal occurrence of VA that causes Sudden Cardiac Death. Apart from VA, the proposedalgorithmis useful other arrhythmias types that are precursors to other cardiac problems as well..