Phase Space Reconstruction and Ensemble Voting to Localize Arrhythmogenic Zones of the Ventricles

Premature ventricular contraction (PVC) and ventricular tachycardia (VT) are two types of cardiac rhythm irregularities originating from the lower chambers of the heart. In severe cases, these arrhythmias could lead to sudden cardiac arrest. For patients unresponsive to pharmacological treatments, catheter ablation becomes a necessary intervention. Precise localization of arrhythmia origins significantly improves the surgical success rates, reduces procedure time and minimizes complications. This study introduces an electrocardiogram (ECG) analysis framework to predict ventricular arrhythmia origin zones. The framework employs phase space reconstruction to map ECG signals into higher-dimensional spaces, followed by a machine learning ensemble voting mechanism to identify arrhythmogenic ventricular locations. This approach aims to enhance the precision and efficiency of ventricular arrhythmia localization, contributing to improved clinical outcomes.