Integrating multipolar and wavefront annotation with computed tomography-derived wall thinning analysis to identify critical substrate in scar-related ventricular tachycardia
M Hachisuka, Y Iwasaki, M Kobayashi, H Hirayama, Y Toida, S Okajima, N Ito, Y Fujimoto, H Murata, Y Aizawa, K Yodogawa, W Shimizu, K AsaiAbstract
Introduction
Catheter ablation (CA) for scar-related ventricular tachycardia (VT) remains challenging due to high recurrence and long procedures. CT-derived wall thinning correlates with abnormal electrograms in ischemic cardiomyopathy [1,2] but shows limited predictive value for identifying deceleration zones [3]. Although thinned scar indicates conduction slowing, it alone may not define VT-critical sites. Digital-twin imaging provides three-dimensional (3D) substrate characterization [4], but its direct correspondence with VT mapping remains unclear.
Purpose
This study assessed the electrophysiological and anatomical characteristics of VT circuits by integrating activation-based annotation algorithms with CT-derived wall-thinning analysis.
Methods
From 2024 to 2025, five patients (3 men, age 60 ± 21 years) with sustained monomorphic VT underwent CA. In two ischemic cardiomyopathy cases, CT data were used to generate wall-thinning maps. Mapping data were retrospectively annotated with multipolar and wavefront algorithms and merged with 3D CT models. VT-critical sites were defined as radiofrequency points that terminated VT.
Results
A total of 24,017 and 37,084 mapping points were analyzed with multipolar and wavefront annotation, respectively. In the multipolar analysis, VT-critical sites (n=256) had lower unipolar (2.6 ± 3.4 vs 4.4 ± 2.8 mV, P<0.001) and bipolar voltages (0.4 ± 0.3 vs 0.6 ± 1.3 mV, P=0.009) than non-critical sites (n=23,761) and more often coincided with wall thinning ≤5 mm (28.9% vs 21.4%, P=0.004) and wall-thickness channels (21.1% vs 16.5%, P=0.047). In the wavefront analysis, VT-critical sites (n=445) showed lower unipolar (3.4 ± 1.6 vs 3.6 ± 2.4 mV, P=0.023) and bipolar voltages (0.4 ± 0.7 vs 0.6 ± 1.1 mV, P<0.001) and shorter fractionation durations (94.4 ± 38.2 vs 109.5 ± 42.8 ms, P<0.001). Concordance with wall thinning ≤5 mm (13.0% vs 23.7%, P<0.001) and wall-thickness channels (7.9% vs 11.1%, P=0.032) was lower, as shown in Figure, suggesting VT-critical sites may not always overlap with thinned regions on CT. Within wall-thickness channels, 3,965 multipolar and 4,091 wavefront points were analyzed. VT-critical sites on these channels (54 and 35 points) were compared with non-critical sites. In the multipolar analysis, no voltage differences were seen in unipolar (3.1 ± 7.0 vs 3.7 ± 2.5 mV, P=0.102) or bipolar values (1.1 ± 7.1 vs 0.4 ± 1.8 mV, P=0.437). In the wavefront analysis, VT-critical sites showed lower unipolar (2.4 ± 0.8 vs 3.2 ± 1.8 mV, P<0.001) and bipolar voltages (0.2 ± 0.2 vs 0.4 ± 0.7 mV, P<0.001) and shorter fractionation durations (97.2 ± 31.1 vs 112.5 ± 42.3 ms, P=0.049).
Conclusion
VT-critical sites showed low voltage across annotation algorithms. CT-derived wall thinning alone was insufficient for precise localization, but integration with multipolar annotation improved target identification. Identifying wall-thickness channels before CA may further aid localization of VT-critical sites.