Abstract
Diabetic cardiomyopathy increases the risk of ventricular tachycardia and fibrillation through ionic and structural remodeling, but whether the reentrant substrate arises from a diffuse slow-conduction (fibrotic) substrate, a focal scar, or from repolarization remodeling remains unclear. Using a three-dimensional monodomain model (truncated-ellipsoid left ventricle; ToR-ORd ionic model with six diabetic current-scaling factors), I performed a controlled ablation study (in-silico knockout, not catheter ablation) in which diffuse fibrosis and connexin-43 remodeling were reduced tissue conductivity, patchy fibrosis discrete inexcitable elements, and a focal scar an inexcitable transmural wedge, with reentry tested by an S1-S2 protocol. Diabetic remodeling slowed conduction (≍60 to 13 cm/s) and prolonged the QT interval (255 to 301 ms). Sustained reentry required slow conduction and a focal scar: a scar at normal conduction did not reenter, and a slow-conduction substrate without a scar did not - at any patchy-fibrosis density, nor with spatially correlated or strandlike fibrosis textures (0/14); only their combination produced scar-anchored reentry. Disabling ionic remodeling did not abolish reentry, and a conductivity sweep confined reentry to a conduction-velocity window, identifying conduction slowing, not repolarization, as the driver of reentry maintenance; findings reproduced at a finer, independent discretization, though conduction-velocity and block thresholds are effective, resolution-dependent values (coarse grid; robustness check, not mesh convergence). Because the focal scar reflects concomitant ischemic injury and the diffuse slow-conduction substrate is invisible to late-gadolinium-enhancement (LGE), detecting this two-hit substrate would require pairing LGE with diffuse-fibrosis (T1/extracellular-volume) mapping for risk stratification in diabetic patients with a scar.