Near real-time in silico pace mapping accurately localized focal ventricular arrhythmias with a distance of 8.2 mm (6.9-12 mm) from clinical ground truth sites across 270 locations.
Observational (n=18)
Does near real-time in silico pace mapping (InSPM) accurately localize focal and re-entrant ventricular arrhythmias compared to clinical pace mapping in patients with structural heart disease?
In-silico pace mapping provides a rapid and accurate computational modeling approach to localize focal and re-entrant ventricular tachycardias, with a median localization error of 8.2 mm.
BACKGROUND: Catheter ablation of ventricular tachycardia (VT) is characterized by long procedures and frequent recurrence. Personalized image-based computational models may provide noninvasive ablation target guidance but are computationally demanding and cannot localize focal arrhythmias. OBJECTIVE: This study aims to clinically validate our near real-time in silico pace mapping (InSPM) approach, which rapidly localizes both focal and re-entrant arrhythmia site of origins within personalized image-based models. METHODS: Personalized models incorporating scar were reconstructed from imaging data in 18 patients with structural heart disease; 12-lead electrocardiogram (ECGs) were obtained during clinical pace mapping and pacing site locations defined as ground truth. ECG templates of induced monomorphic VT were obtained. Virtual pacing was conducted in models and simulated ECGs correlated with clinical templates to produce high-resolution virtual pace-maps. Distance (d) between clinical ground truth sites and simulation predicted target areas with highest correlation quantitatively assessed InSPM accuracy for localizing focal activations. For re-entrant VT, predicted targets were compared with surrogates of VT site of origin and mapped VT circuits. RESULTS: Intrinsic resolution of clinical pace mapping was approximately 4 mm for similarly correlated ECGs (mean correlation coefficient >0.99). Across 270 clinical pace-mapping locations, d was 8.2 mm (6.9‒12 mm), relatively insensitive to cardiomyopathy, but with increased accuracy in right vs left ventricles. Patient-specific ECG electrodes alongside accurate scar representation, particularly in patients with ischemia, were important for optimizing InSPM accuracy. InSPM created from clinical ECG VT templates reliably identified re-entrant VT exit sites. CONCLUSION: InSPM provides a rapid and validated personalized computational modeling ablation technology to accurately localize both focal and re-entrant VTs, which may be practically integrated into clinical workflows.
A Fri, study conducted a observational in Structural heart disease with ventricular tachycardia (n=18). Near real-time in silico pace mapping (InSPM) vs. Clinical ground truth sites was evaluated on Distance (d) between clinical ground truth sites and simulation predicted target areas. Near real-time in silico pace mapping accurately localized focal ventricular arrhythmias with a distance of 8.2 mm (6.9-12 mm) from clinical ground truth sites across 270 locations.