Electrophysiologically-informed digital twins reduced mean cycle length and conduction velocity errors compared to anatomic-only models (14% vs 23% and 37% vs 54%, respectively; p<0.05).
Do electrophysiologically-informed digital twins improve the accuracy of capturing atrial fibrillation cycle length and conduction velocity compared to anatomic-only models in patients with AF?
Integrating functional electrophysiological data with anatomical data in digital twins significantly improves the accuracy of modeling atrial fibrillation phenotypes, representing a step toward precision-guided treatment planning.
Absolute Event Rate: 14% vs 23%
p-value: p=<0.05
Patient-specific digital twins (DTs) are a promising tool for studying cardiac arrhythmias and planning patient-specific therapies. Although recent DTs have begun to incorporate functional data, comprehensive regional calibration of electrophysiological activity during atrial fibrillation (AF) remains a fundamental challenge for improving predictive accuracy. This study presents electrophysiologically-informed DTs (EP-DTs) for atrial fibrillation that integrate detailed patient-specific electrophysiological and anatomical data to emulate the AF electrophysiological phenotype and predict therapy outcomes. Comprehensive EP-DTs of the left atrium (LA) were developed in 12 patients using CT-derived geometries. These models were regionally calibrated using intracavitary AF recordings from multipolar catheters to capture spatial variability. EP-DTs more accurately captured key electrophysiological phenotypes related to prognosis, including AF cycle length (CL) and conduction velocity (CV). EP-DTs reproduced global CL (e.g., 194 ± 15 ms vs. 192 ± 10 ms) and CV (e.g., 376 ± 146 mm/s vs. 372 ± 54 mm/s) referenced to patient data. Regionally, EP-DTs reduced mean CL and CV errors (14 ± 7% vs. 23 ± 9%; 37 ± 12% vs. 54 ± 15%; p < 0.05) compared to anatomic-only models. Notably, EP-DTs were also able to specifically identify proarrhythmic regions representing limited areas of atrial tissue (8.7 ± 8.1% vs. 12.7 ± 6.6% area). Incorporating functional with anatomical data substantially enhances the physiological fidelity of cardiac DTs, reproducing significant biomarkers such as CL and CVs. EP-DTs therefore represent a critical step toward precision-guided treatment planning in cardiac arrhythmias.
Termenón-Rivas et al. (Fri,) conducted a other in Atrial fibrillation (n=12). Electrophysiologically-informed Digital Twins (EP-DTs) vs. Anatomic-only models was evaluated on Mean cycle length (CL) error (p=<0.05). Electrophysiologically-informed digital twins reduced mean cycle length and conduction velocity errors compared to anatomic-only models (14% vs 23% and 37% vs 54%, respectively; p<0.05).