Real-Time CNN Based ST Depression Episode Detection Using Single-Lead ECG

Abstract

A method for real monitoring of the heart for ST-depression episodes is described here. We have developed a convolutional neural network (CNN) based machine learning algorithm for classifying ECG signals into normal or ST-depression episodes of the heart with an accuracy over 92%. Our algorithm is capable of detecting ST-depression episodes of varying duration. The algorithm is evaluated using European ST-T Database. The best results obtained here are 0.95%, 0.98%, and 0.91% respectively for accuracy, sensitivity, and specificity.