Multiscale Electrophysiological Remodeling in Ischemic Heart Failure: Simulation Analysis of Ventricular Ectopic Activation
TL;DR
These analyses support numerical convergence and parameter-range-specific sensitivity, not biological population variability, and represent conditional cellular remodeling and generic geometry-constrained wave dynamics rather than a causal molecular-to-organ mechanism or individualized arrhythmia prediction.
Abstract
Background/Objectives: Premature ventricular contractions (PVCs) frequently accompany ischemic chronic heart failure (I-CHF), but ionic remodeling and organ-scale propagation cannot be interpreted as a continuous mechanism when cellular and tissue models are not explicitly coupled. Methods: We used a scale-separated workflow comprising a literature-constrained human ventricular cellular model and an independent FitzHugh–Nagumo-type reaction–diffusion model in one CT-derived ventricular geometry. Results: The I-CHF-motivated cellular scenario prolonged APD90 from 280 to 350 ms in EPI-labeled cells and from 300 to 390 ms in ENDO-labeled cells, reduced peak ICaL by 40–45%, and increased peak INaCa by approximately 105%. At the organ scale, prescribed septal stimulation generated outer-wall peak-to-peak local potentials of 1.894, 3.758, and 1.941 mV at Sites I–III, respectively. The paired inner-wall traces retained two-thirds of these amplitudes and identical maximum-|dφe/dt| landmark times of 109.4, 132.2, and 347.6 ms, providing no evidence of independently resolved transmural heterogeneity. Parameter uncertainty was evaluated using one deterministic baseline and 55 additional configurations generated by Latin hypercube sampling. Tissue diffusivity was most strongly associated with landmark time (PRCC = −0.78), whereas stimulus amplitude was most strongly associated with peak-to-peak amplitude (PRCC = 0.72). Mesh- and time-step-refinement differences decreased to 0.7% and 0.5%, respectively. Conclusions: These analyses support numerical convergence and parameter-range-specific sensitivity, not biological population variability. Because the two model levels remained mathematically independent and the ventricular model omitted a patient-specific ischemic substrate, the findings represent conditional cellular remodeling and generic geometry-constrained wave dynamics rather than a causal molecular-to-organ mechanism or individualized arrhythmia prediction.