A deep learning algorithm based on convolutional units and time frequency presentations classified ECG signals for congestive heart failure, arrhythmia, and normal beats with 93.75% accuracy.
Heart attacks had been for many years one of the primary public health issues. According to the World Health Organization, ischemic heart diseases are in the top of ten leading causes of death. Thus, automatic detection of abnormal heart conditions may provide a necessary hospitalization, for patients suffering from heart diseases making it possible to save their lives. In the current article, we present a simple but efficient deep learning algorithm based on simple convolutional units and time frequency presentations. The proposed model can classify three types of Electrocardiogram (ECG) signals related to three different cases; namely patients with Congestive Heart Failure, patients with Arrhythmia and others with a normal heart beats. The 162 recordings used in this algorithm was taken from Physio Net databases; 96 of them taken from the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) arrhythmia database, 36 from the MIT-BIH normal sinus rhythm database and 30 from the Beth Israel Deaconess Medical Center (BIDMC) congestive heart failure database. The simplicity and flexibility of the proposed approach would present a real potential for real time monitoring with 93.75% accuracy.
Karboub et al. (Tue,) conducted a other in Congestive Heart Failure, Arrhythmia, and normal heart beats (n=162). Deep learning algorithm based on convolutional units and time frequency presentations was evaluated on Classification accuracy. A deep learning algorithm based on convolutional units and time frequency presentations classified ECG signals for congestive heart failure, arrhythmia, and normal beats with 93.75% accuracy.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: