Effective classification of ECG signals using enhanced convolutional neural network in IOT (2202.05154v1)

Abstract: In this paper, a novel ECG monitoring approach based on IoT technology is suggested. This paper proposes a routing system for IoT healthcare platforms based on Dynamic Source Routing (DSR) and Routing by Energy and Link Quality (REL). In addition, the Artificial Neural Network (ANN), Support Vector Machine (SVM), and Convolution Neural Networks (CNNs)-based approaches for ECG signal categorization were tested in this study. Deep-ECG will employ a deep CNN to extract important characteristics, which will then be compared using simple and fast distance functions in order to classify cardiac problems efficiently. This work has suggested algorithms for the categorization of ECG data acquired from mobile watch users in order to identify aberrant data. The Massachusetts Institute of Technology (MIT) and Beth Israel Hospital (MIT/BIH) Arrhythmia Database have been used for experimental verification of the suggested approaches. The results show that the proposed strategy outperforms others in terms of classification accuracy.