Publication

Abstract : People, now being busy with their everyday schedule, find it difficult to get proper heart monitoring and diagnosis for their loved ones in emergency calls. Conventional ECG monitoring methods suffer adversely from the increasing quantity of patients and lack of trained clinicians. This is manifested in the form of increased death rates due to unattended cases. Hence, the establishment of a smart assistive technique is more preferred than the traditional technique in the healthcare field. This paper proposes a deep neural network (DNN)-based decision support system which helps to monitor ECG abnormalities. A DNN-based algorithm is proposed to train and test the decision-making system. The trained system looks for the prediction of normal and abnormal conditions of the patient, meanwhile training itself to increase its accuracy rate. Experimental results show that this system has obtained very good performances in terms of low error rate and high accuracy.