Low-Cost ECG Monitoring System with Classification Using Deep Learning

Abstract

Electrocardiogram (ECG) signals are widely used as one of the most important tests in medical practice to assess the condition of a patient’s heart by placing electrodes on the body surface. Electrocardiographs are used to detect, process, and visualize changes in the heart’s electrical activity over time. In this study, an ECG acquisition system has been implemented to acquire, process, and classify ECG signals. Deep learning techniques (CNN models) were used for classification. For performance evaluation, the ECG signals have been extracted from the MIT-BIH database. The developed web platform has been programmed to visualize the classification results and to print the analysis, providing the clinician with all the information and data necessary to make his diagnosis and determine the appropriate treatment.