Can Non-Heart Diseases like Diabetes and Sleep Apnea be Detected by an ECG?

Abstract

Raul Isea

An electrocardiogram (ECG) is frequently used to identify heart problems, but it can also identify apnea, diabetes, and other conditions that are not heart-related. The study uses deep learning methods from artificial intelligence, particularly the Convolutional Neural Network (CNN) approach, to look for patterns in ECG data. To predict illness, two different methodologies are used. The first approach makes use of an ECG scale, while the second employs a gradient-boosting machine (GBM)-based learning technique. The MIT-BIH Arrhythmia database, the Normal Sinus Rhythm database, the BIDMC Congestive Heart Failure database, the Sleep Apnea database, the type 2 diabetes mellitus dataset, and the dataset for healthy volunteers were the sources of the data used in the study. Finally, 92% of predictions were accurate.

PDF