Multiscale Electrophysiological Remodeling in Ischemic Heart Failure: Simulation Analysis of Ventricular Ectopic Activation

Figures

• 

  Figure 1. Corrected conceptual representation of the intracellular and extracellular continua. The membrane separates φi and φe, and the transmembrane balance contains capacitive, aggregate ionic, and applied-stimulus terms. Di and De denote the transport tensors in the retained coefficient-form notation. This schematic replaces the original panel, which incorrectly labeled the capacitive branch as −ICaL.

• 

  Figure 2. Computational workflow and prescribed ectopic stimulation sites. (a) Geometry preparation and bidomain/reaction–diffusion simulation workflow; (b) candidate septal stimulation positions evaluated in the ventricular model.

• 

  Figure 3. Deterministic three-dimensional ventricular potential propagation under control and I-CHF-motivated model conditions. Panels show the retained potential maps at the indicated simulation times. (a) 3D ventricular potential diffusion about normals; (b) Potential cloud map of representative sections of the heart segments about normals; (c) 3D ventricular potential diffusion about I-CHF patients accompanied by PVCs; (d) Potential cloud map of representative sections of the heart segments about I-CHF patients accompanied by PVCs.

• 

  Figure 4. Deterministic cellular action-potential outputs under control and I-CHF-motivated parameterizations. (a) Condition-wise action-potential comparison; (b) EPI- and ENDO-labeled outputs.

• 

  Figure 5. Deterministic cellular-current outputs for ICaL (a) and INaCa (b) under control and I-CHF-motivated parameterizations in EPI- and ENDO-labeled cells.

• 

  Figure 6. Deterministic local extracellular potentials at three ventricular sampling sites. (a) Outer-wall (OW; solid) and inner-wall (IW; dashed) traces; dotted lines mark the maximum-(|dφe/dt|) landmarks. (b) Peak-to-peak amplitudes. (c) Landmark times. OW and IW denote geometric sampling surfaces, not distinct transmural phenotypes. The traces are local outputs from a single baseline simulation rather than ECG leads, and the landmarks are not interpreted as tissue activation times.

• 

  Figure 7. Parameter uncertainty, sensitivity, and numerical convergence analyses. (a) Median local-potential trajectory and pointwise 2.5th−97.5th percentile envelope obtained from 55 Latin hypercube parameter samples, together with the deterministic baseline trajectory. (b) Partial rank correlation coefficients between the sampled organ-scale parameters and the maximum-(|dφe/dt|) landmark time, peak-to-peak amplitude, and absolute waveform area. (c) Relative differences in the selected output metric across the mesh and time-step refinement sequences, calculated with respect to the reference solution. The dashed line denotes the predefined 1% convergence criterion.