Scar-related myocardial tissue can lead to ventricular arrhythmia (VA), a global concern for sudden cardiac death. Repolarization dispersion due to electrical remodeling within infarcted territory often triggers ventricular arrhythmias. However, evaluating ventricular repolarization globally is clinically challenging and there is no gold-standard approach. This paper introduces a new method using body-surface leads to automate activation (ATs) and recovery times (RTs) detection in unipolar electrograms (UEs). Multilevel Discrete Wavelet Transform sets activation and recovery detection windows based on external lead data. Then, Wyatt's method is used to compute ATs and RTs. By analyzing catheter-based intracardiac electrograms from n=9 infarcted swine, we compare ARI values between healthy tissue and arrhythmia-prone border zones (BZ), and between normal sinus rhythm (NSR) and pacing. Results reveal significant ARI differences between NSR and pacing, and among tissue types, with heterogeneous ARI values highlighting the repolarization complexity within BZ. This emphasizes the necessity for new automated approaches in assessing and treating cardiac arrhythmias, acknowledging the diverse electrophysiological individual profiles.