Neural network-based prediction of atrial fibrillation at discharge following cardiac surgery

1. AbstractBackgroundAtrial fibrillation at discharge is an infrequent but clinically relevant complication after cardiac surgery, linked to adverse long-term outcomes. This study aimed to develop and evaluate artificial intelligence-based models for predicting atrial fibrillation using electrocardiographic conduction abnormalities and other clinical risk factors.MethodsA total of 3,405 patients undergoing coronary artery bypass grafting, valvular, aortic, or combined procedures were included. Perioperative 12-lead electrocardiograms and clinical variables were preselected via univariable logistic regression (p