An InceptionTime-Inspired Convolutional Neural Network to Detect Cardiac Abnormalities in Reduced-Lead ECG Data

Abstract

Cardiovascular disease is the leading cause of death worldwide. The twelve-lead electrocardiogram (ECG) is a common tool for diagnosing cardiac abnormalities, but its interpretation requires a trained cardiologist. Thus there is growing interest in automated ECG diagnosis, especially using fewer leads. Hence the PhysioNet-CinC Challenge 2021: Will two (leads) do? The University of Bath team (UoB_HBC) developed InceptionTime-inspired deep convolutional neural networks, using parallel 1D convolutions of varying length, for twelve-, six-, four-, three-, and two-lead models. The twelve-lead model achieved a Challenge metric score of 0.35 on the test set, placing the University of Bath team 23rd out of 39 entries. Though the twelve-lead model performed best, three-lead performance was lower by only 0.25 %, suggesting potential for reliable reduced-lead diagnoses. Furthermore, the three-lead model performed consistently better than the six-lead, highlighting the importance of selection of type of lead, not just their number.