Cardiovascular Magnetic Resonance-Based Three-Dimensional Structural Modeling and Heterogeneous Tissue Channel Detection in Ventricular Arrhythmia.

ABSTRACT

Geometrical structure of the myocardium plays an important role in understanding the generation of arrhythmias. In particular, a heterogeneous tissue (HT) channel defined in cardiovascular magnetic resonance (CMR) has been suggested to correlate with conduction channels defined in electroanatomic mapping in ventricular tachycardia (VT). Despite the potential of CMR for characterization of the arrhythmogenic substrate, there is currently no standard approach to identify potential conduction channels. Therefore, we sought to develop a workflow to identify HT channel based on the structural 3D modeling of the viable myocardium within areas of dense scar. We focus on macro-level HT channel detection in this work. The proposed technique was tested in high-resolution ex-vivo CMR images in 20 post-infarct swine models who underwent an electrophysiology study for VT inducibility. HT channel was detected in 15 animals with inducible VT, whereas it was only detected in 1 out of 5 animal with non-inducible VT (P < 0.01, Fisher's exact test). The HT channel detected in the non-inducible animal was shorter than those detected in animals with inducible VTs (inducible-VT animals 35 ± 14 mm vs. non-inducible VT animal 9.94 mm). Electrophysiology study and histopathological analyses validated the detected HT channels. The proposed technique may provide new insights for understanding the macro-level VT mechanism.