CNN Based Deep Learning to Detect Low Ejection Fraction Using Single Lead ECG

Early detection of left ventricular systolic dysfunction (LVSD) using minimally invasive electrocardiographic assessment in an ambulatory setting may be important considering the increasing economic burden of heart failure. Recently artificial intelligence (AI) algorithms have been reported to classify reduced ejection fraction using a 12-lead ECG measurement snapshot. However, the studies of AI to identify LVSD using only a single lead ambulatory ECG (aECG) are limited. We developed a convolution neural network based deep learning algorithm (CNN-aECG) to predict left ventricular ejection fraction (LVEF) less than 40% using aECG from implantable loop recorders (ILR). A total of 35,741 unique LVEF-aECG pairs collected from 2,249 ILR patients were used to train the model. The model yielded accuracy, sensitivity, specificity, and AUC of 75%, 70%, 76% and 0.8 respectively in the independent test set of 6,611 unique LVEF-aECG pairs from 750 patients. This proposed method of ambulatory CNN-aECG monitoring may allow for identifying reduced ejection fraction post ILR implant via remote monitoring.