Computer-aided characterization of the arrhythmogenic substrate after myocardial infarction

Figures

• 

  Figure 1. Conceptual frameworks for re-entry initiation. (A) The traditional triangle of arrhythmogenesis by Coumel identifies three major factors for arrhythmia initiation. (B) Schematic illustration of re-entry in tissue with early and late repolarization time (RT) due to short or long action potential duration, respectively. A premature beat coming from an early RT region (1) may block against tissue that has a longer RT and is still refractory (2) and cannot induce activation in the late RT region (3). If the early RT region is large enough, the activation wave may propagate through that region (4) in a time span that allows the late RT region to repolarize and become excitable, allowing re-activation of the tissue (5). The wave may then travel back to the region that was still refractory but which has become excitable in the meantime (6). By that moment, the tissue from which the premature beat originated has become excitable again, and the wavefront from the late RT region will propagate to that tissue, restarting this circuit (7). (C) The Circle of Re-entry extends the triangle of Coumel and proposes four requirements for re-entry arising from spatiotemporal interactions between…

• 

  Figure 2. Advanced techniques for structural characterization of the arrhythmogenic substrate. (A) Thresholding of LGE-CMR provides a 3D model of slow-conducting channels and interface area: healthy myocardium (blue), border zone (yellow), and scar core (red) of the left ventricle, including slow conduction channels (black arrows), and the interface area between border zone and core scar.18 (B) Heterogeneity (entropy) of pixel signal intensities in LGE-CMR images, without prior thresholding.24 (C) LGE-CMR adipose tissue quantification at different locations (atrioventricular and interventricular grooves, right ventricle free wall).22 (D) Hypoattenuation of ischaemic area in dual-energy LIE-CT image.25,26 (E) Ce-CT tissue heterogeneity in ischaemic cardiomyopathy, showing a 3D model of the left ventricle in patients with high (arrhythmic) and low (arrhythmia-free) heterogeneity.27 (F) Ce-CT image and 3D model of the left ventricle divided into non-injured tissue (red), infiltrating adipose tissue (yellow), and fat–myocardium admixture (dark yellow).28 Ce, contrast enhanced; CMR, cardiac magnetic resonance imaging; CT, computed tomography; LGE, late gadolinium enhancement; LIE, late iodine…

• 

  Figure 3. Overview of novel invasive electrical characterization mapping approaches of the arrhythmogenic substrate for VT ablation. (A) Ripple mapping42: a channel of electrograms extended from base to apex through the scar (red to purple), showing progressively delayed local activation along its course. (B) Coherent mapping44: automated conduction velocity map, indicating a basal inferolateral wall isthmus. (C) Directed graph mapping45: isthmus location of a single-loop re-entry and a possible inner-loop. (D) Dynamic voltage mapping46: changes in local voltages distinguishing uniform tissue (purple) and disarray (blue).

• 

  Figure 4. The diffusion reaction Eikonal alternant model (DREAM) combines the computationally expensive (A) reaction–diffusion model99 with the fast (B) Eikonal model7 to simulate re-entrant cardiac arrhythmias, both functional and around anatomical obstacles such as scar. (C) DREAM98 achieves faster computing times compared to monodomain simulations with RD models at mesh resolutions <0.5 mm, while maintaining accuracy even on coarse meshes at resolutions >0.5 mm where monodomain simulations can’t be performed with classic RD models. AT, activation times; Idiff, diffusion current; RT, recovery times; Vm, transmembrane potential voltage.