Software for prediction of postoperative atrial fibrillation following bypass surgery

The problem of accurately predicting the onset of sustained postoperative atrial fibrillation (AF) in patients undergoing coronary artery bypass grafting (CABG) remains open. Investigators have reported many clinical indices currently associated with postoperative AF following CABG. Contemporary machine learning techniques are well-suited to recognizing underlying trends in 'training' data consisting of several labeled examples, and to using the results to classify new unlabeled data with remarkable sensitivity and specificity. We propose the development of advanced clinical decision support software capable of automatically gathering and analyzing relevant clinical data from patients undergoing CABG in order to provide physicians with objective and non-invasive insights into the likelihood of sustained postoperative AF so that patient morbidity and mortality, as well as healthcare costs, can be significantly reduced by targeting appropriate preventative therapies.

Geoffrey Seaborn
Faculty Supervisor: 
Drs. Selim Akl & Damian Redfearn