Three-dimensional modeling of localized ventricular ischemia for relative quantitative analysis of ECG alterations

OBJECTIVE: This study aims to investigate the impact of localized ventricular ischemia on cardiac electrical activity using a three-dimensional computational modeling framework, and to evaluate the robustness of relative ECG-based quantitative markers under varying ischemic conditions. APPROACH: A three-dimensional heart torso model was developed incorporating anisotropic myocardial conductivity and rule-based fiber orientation in the ventricular regions to further enhance physiological plausibility while preserving computational efficiency. Electrical propagation was modeled using modified FitzHugh-Nagumo equations within a bidomain formulation. Localized ischemia was introduced through controlled alterations in intracellular conductivity, resting membrane potential, and ionic kinetics. Simulations were performed under healthy and ischemic conditions, and ECG signals were analyzed using both qualitative waveform assessment and normalized quantitative metrics, including ST/QRS and T/QRS ratios. A sensitivity analysis was conducted by varying ischemia severity, location, and size. Model outputs were compared with clinical ECG signals from the European ST-T Database using correlation and RMSE after normalization and temporal alignment. MAIN RESULTS: The simulations revealed clear lead-dependent ECG alterations associated with ischemia, including ST-segment deviations and T-wave modifications. The ST/QRS ratio showed consistent sensitivity to ischemia severity while remaining relatively stable across variations in ischemic size and location. Quantitative comparison with clinical ECG signals demonstrated moderate agreement, with good agreement of the QRS complex and larger discrepancies in repolarization features. SIGNIFICANCE: This study highlights the importance of incorporating anisotropic conduction and fiber orientation in computational ECG modeling. The proposed framework provides a reproducible and computationally efficient approach for investigating ischemia-induced ECG changes and supports the use of relative ECG metrics as robust descriptors for methodological analysis and synthetic data generation.