The task-specific convolutional neural network achieved an AUROC of 0.755 for predicting atrial fibrillation recurrence after pulmonary vein isolation using ECG data.
Observational (n=189)
No
Does a task-specific convolutional neural network improve the prediction of atrial fibrillation recurrence after pulmonary vein isolation compared to a foundation model using ECG data?
Both task-specific and foundation neural network models can successfully predict atrial fibrillation recurrence after pulmonary vein isolation using only ECG data, achieving comparable performance.
Absolute Event Rate: 0.755% vs 0.742%
Catheter-based pulmonary vein isolation (PVI) is widely used to treat atrial fibrillation (AF). The procedure isolates the pulmonary vein from the left atrium to suppress irregular heartbeats. If AF recurs, further interventions may be required. This study investigates model-based prediction of AF recurrence using only electrocardiogram (ECG) data. We benchmarked two neural networks for the prediction task. The task-specific convolutional neural network achieved an AUROC of 0.755, slightly exceeding the foundation model (AUROC 0.742). These results confirm earlier findings that ECGs contain predictive information about AF recurrence and they can be extracted by foundation models.
Boudnik et al. (Wed,) conducted a observational in Atrial Fibrillation recurrence after Pulmonary Vein Isolation (n=189). ECG data prediction model vs. none was evaluated on Prediction of atrial fibrillation recurrence. The task-specific convolutional neural network achieved an AUROC of 0.755 for predicting atrial fibrillation recurrence after pulmonary vein isolation using ECG data.