Artificial Intelligence for Postoperative Atrial Fibrillation After Cardiac Surgery: A 2026 Perspective

Postoperative atrial fibrillation remains one of the most common complications after cardiac surgery. Although it is often transient, it can complicate recovery, prolong hospitalization, and identify patients at higher risk of future atrial fibrillation. In daily practice, however, management remains largely reactive, as most patients are recognized only after the arrhythmia has already occurred. This creates an important opportunity for earlier risk stratification. Cardiac surgery is a particularly attractive setting because the event is common, perioperative monitoring is dense, and preoperative 12-lead electrocardiograms are routinely available. Many of these electrocardiograms are recorded in sinus rhythm and may appear normal on clinical review, yet they may still contain subtle signs of latent atrial vulnerability. In this perspective, we argue that postoperative atrial fibrillation after cardiac surgery is a promising use case for clinically meaningful artificial intelligence in 2026. The most compelling next step is to test whether preoperative electrocardiograms add useful information to clinical and biomarker-based risk assessment before the first postoperative episode occurs.