Prediction of atrial fibrillation from sinus rhythm

A recent study has observed the accuracy of an artificial intelligence (AI)-based model to predict paroxysmal AF from a normal sinus rhythm single-lead ECG.

A convolutional neural network (CNN) model was trained and evaluated using data from three AF screening studies. A total of 478 963 single-lead ECGs from 14 831 patients aged ≥65 years were included in the analysis.

The Cardiolund ECG Parser was used for automated ECG analysis, and to produce the dataset from which the AI-model training and test data have been extracted.

The study results show that AI-model can predict paroxysmal AF from sinus rhythm with an AUC between 0.62 and 0.8.

The study is reported in the publication: An artificial intelligence–based model for prediction of atrial fibrillation from single-lead sinus rhythm electrocardiograms facilitating screening, by Tove Hygrell, Fredrik Viberg, Erik Dahlberg, Peter H Charlton, Katrin Kemp Gudmundsdottir, Jonathan Mant, Josef Lindman Hörnlund, Emma Svennberg, in EP Europace, 2023;, euad036, DOI: 10.1093/europace/euad036