An algorithm using artificial intelligence can now detect atrial fibrillation during normal sinus rhythm using input from a standard ECG, US researchers report.

Researchers from the Mayo Clinic in Rochester, Minnesota, used machine learning and neural networks to identify subtle patterns present in a standard ECG acquired during normal sinus rhythm that were due to structural changes associated with a history of or impending AF.

The study included nearly 650,000 normal sinus rhythm ECGs from more than 180,000 patients, with 8.4% having known AF.

With the resulting algorithm, a single 10-second 12-lead ECG could be used to identify a patient with AF with 79.0% sensitivity, 79.5% specificity and 79.4% overall accuracy, the researchers reported in the Lancet.