‘From data to diagnosis’: Identification of patients with atrial fibrillation using computational analysis of sinus rhythm electrocardiograms

UK prevalence of atrial fibrillation (AF), the most common treatable form of cardiac rhythm disorder seen in clinical practice is expected to rise with the increase in the ageing population. In addition, risk factors for AF, such as hypertension and obesity are expected to increase globally. Importantly, the prevalence of AF is likely underestimated as a large proportion remain undiagnosed due to the transient (paroxysmal) nature of this abnormality. AF can have major consequences if undetected (such as strokes). Our project is based on using several novel computational algorithms on electrocardiographic (ECG) signals to find visually non-discernible differences. We will explore combining the information from the different algorithms to increase the accuracy of the predictions. This project will create a framework that can facilitate accurate and rapid diagnosis of patients and be a stepping stone towards an automated, low cost and reliable screening method for AF and other cardiovascular diseases.