Fibrosis Microstructure Modulates Reentry in Non-ischemic Dilated Cardiomyopathy: Insights From Imaged Guided 2D Computational Modeling

Figure 1. Computational model creation from an LGE-CMR image. (A) LGE-CMR image. (B) Image segmentation into enhanced and non-enhanced myocardium. (C) LV myocardial pixel intensity histogram and example intensity to fibrosis probability mapping. (D) Computational model with LGE intensities registered to image. (E) Zoomed in view of simulated fibrosis micro-structure. (F) Four conduction slowing zones in LGE. (G) Rule-based circumferential fibers.

Figure 2. Simulated fibrosis micro-structures with varying type and density. Black lines are separating edges through which current cannot cross, whereas white areas are electrically inactive and non-conducting.

Figure 3. Examples of three canonical reentry mechanisms. (Top) Paths of reentrant circuits in green, with blue boxes indicating zoomed-in areas used for the voltage maps below. (Middle) Transmembrane voltage maps, with time (ms) after the final stimulus displayed in the top left corner. (Bottom) Pseudo-electrograms measured 4 cm anterior to the geometry showing 1s of electrical activity.

Figure 4. Incidence of reentry mechanisms by fibrosis texture, density, and with normal and reduced conductivity in the LGE.

Figure 5. Activation maps around the LGE for three types of fibrosis micro-structure. Red lines are isochrones spaced at 10 ms intervals. The stimulus site is shown in green in the fibrosis-free control model (bottom row).

Figure 6. (A) Mean (solid line) and standard deviation (transparent area) of trans-septal activation time (TAT) for four coupling intervals (CI) vs. fibrosis density, fibrosis type, and LGE conductivity. The reentry incidence is overlaid in red. (B) Segmented LGE-CMR image with locations of stimulus and TAT recording site.

Figure 7. ROC curves for predicting reentry inducibility based on transmural activation times (TAT). TAT500 and TAT210 are the transmural activation times with 500 and 210 ms coupling intervals respectively, while ΔTAT is the difference. The green dot shows the point of maximum sensitivity and specificity. AUC stands for area under curve.

Figure 8. Effect of pacing location on transmural activation times and reentry incidence. Transmural activation times are given as a mean (solid line) and standard deviation (transparent area) for four coupling intervals (CI), taken over 10 random realizations of each fibrosis density. The reentry incidence is overlaid in red for comparison. All results were obtained with replacement fibrosis texture and reduced conductivity. The results for site 1 are the same as in Figure 6 (top right), whereas sites 2–5 are new locations shown for comparison.