Serial Artificial Intelligence ECG-Derived Atrial Fibrillation Probability and Recurrence Risk After Pulsed Field Ablation

Background Pulsed field ablation is now widely used for atrial fibrillation (AF) ablation, but atrial tachyarrhythmia recurrence remains common and postablation risk tools are limited. We assessed whether a validated sinus‑rhythm artificial intelligence- enabled ECG AF probability score, and its early postablation change, predicts recurrence after pulsed field ablation. Methods We studied consecutive patients undergoing index pulsed field ablation across 3 Mayo Clinic sites (February 2024-March 2025) with ≥1 sin…

By the Truvace deskOctober 17, 20262 min read1 sourcePeer-reviewedTRV-2026-1736· cite & verify

In brief

Researchers at three Mayo Clinic sites followed 1052 consecutive patients who had a first pulsed field ablation between February 2024 and March 2025. They calculated an AI-derived AF probability score from sinus-rhythm ECGs taken before the procedure and tracked how it changed in the weeks afterward. About 30% of patients had a recurrence over 188 days.

The findings suggest an existing ECG-based AI score could add information to the limited tools for judging who is likely to have an arrhythmia return after ablation. The authors present it as an aid to multivariable risk stratification rather than a standalone predictor, and they call for prospective validation before it is used in practice.

Main points

The cohort had 1052 patients, average age 67, and 33% were women.
Median baseline AI ECG AF score was 0.35, and 320 patients (30%) had recurrence.
Each 0.1 increase in baseline score carried a hazard ratio of 1.08 for recurrence (95% CI 1.03-1.14).
Scores rose on day 0 after ablation and then declined, with less improvement among patients who recurred.

The gain

A sinus-rhythm AI-enabled ECG AF probability score, at baseline and in its early postablation change, was associated with recurrence after pulsed field ablation and may support risk stratification.

The rundown

The study enrolled consecutive patients who had an index pulsed field ablation and at least one sinus-rhythm ECG within 180 days before the procedure. The baseline score, which ranges from 0 to 1, was the mean across all baseline ECGs.

Recovery indices were defined as the mean postprocedure score within days 1-30, 1-60, or 1-90 minus baseline. Each index was used to predict recurrence after matching 30-, 60-, or 90-day blanking windows.

The primary outcome was time to the first documented AF, atrial flutter, or atrial tachycardia episode longer than 30 seconds. It was found through clinic ECGs, ambulatory monitoring, and device interrogations. Cox regression adjusted for clinical and echocardiographic covariates, and discrimination was measured with the Harrell C-statistic.