Arrhythmias & electrophysiology
AI-ECG Risk Stratification for Atrial Fibrillation
- multicenter real-world cohort
- development of machine-learning models
- stratified 10-fold cross-validation
- external validation in an independent cohort
+3 more
- development of machine-learning models
- stratified 10-fold cross-validation
- external validation in an independent cohort
Authors' conclusion
AI-ECG provides rapid, low-cost AF risk estimation from a single sinus rhythm ECG; however, its predictive performance is modest compared with clinical scores. At present, AI-ECG may complement, but not replace, traditional risk stratification.
Study context
- Population
- Patients aged ≥40 years who underwent 12-lead ECGs using an AI-enabled electrocardiograph (FCP-9900)patients enrolled
- Participants
- 665 patients
- Exposure
- Machine-learning models vs AI-ECG alone for atrial fibrillation (AF) risk stratification
Main finding
As reported in the abstract
AF prevalence across AI-ECG risk categories
- AF prevalence increased across AI-ECG risk categories, with significantly higher odds in the mid-high and high groups vs low
Observational data: associations, not effects.
Further findings
| Outcome | support vector machine | adaptive boosting | artificial neural network |
|---|---|---|---|
| Discrimination of Model 2 in internal and external validation | AUC 1.00 | AUC 0.97-0.98 | AUC 0.89-0.95 |
- Discrimination of Model 2 in internal and external validation: Model 2 incorporated CHADS2 and CHA2DS2-VASc scores
- Discrimination of Model 2 in internal and external validation: Discrimination (AUC): Model 2 with CHADS2 and CHA2DS2-VASc, internal and external validation, SVM 1.00, adaptive boosting 0.97–0.98, ANN 0.89–0.95; AI-ECG alone 0.64–0.69
SHapley Additive exPlanations feature contributions
- CHA2DS2-VASc the most influential predictor; AI-ECG provided modest incremental value
Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.
Original publication
AI-ECG Risk Stratification for Atrial Fibrillation. JACC Adv. Published 22 July 2026. doi:10.1016/j.jacadv.2026.103036
Summarised from the published abstract; every number is checked against it