Deep Learning Model Using Transthoracic Echocardiograms Can Id A-fib

A deep learning model using transthoracic echocardiograms (TTEs) can accurately predict patients with active or occult atrial fibrillation (AF). The model was trained on over 100,000 TTE videos and was able to differentiate between TTEs in AF and those in sinus rhythm with high accuracy. It also predicted the presence of concurrent paroxysmal AF among TTEs in sinus rhythm. The model’s performance was better than traditional methods and was improved when combined with an electrocardiogram (ECG) deep learning model.