Predicting atrial fibrillation in primary care using machine learning

Abstract

The optimal time-varying machine learning model exhibited greater predictive performance than existing AF risk models and reflected known and new patient risk factors for AF.