Real-Time Smart System for ECG Monitoring Using a One-Dimensional Convolutional Neural Network

Medical facilities and equipment have significantly advanced due to technological advancements in healthcare systems and diagnosis. On the other hand, deep learning neural networks are among the most cutting-edge tools and technology in biomedical engineering. In this context, the authors built an intelligent healthcare system to assist doctors in diagnosing cardiac illness by intelligently analysing the ECG signal. The proposed system can automatically detect and classify different cardiovascular diseases using a one-dimensional convolutional neural network (1D-CNN). The CNN model was developed and evaluated using the MIT-BIH ECG database from Physionet, where the achieved accuracy and loss rate was 97% and 0.11 %, respectively. An acquisition node was developed based on Arduino Uno and AD8232 sensor to collect and process the ECG signal. The Raspberry Pi board was designed as the processing unit by implementing the proposed CNN model following the TinyML approach that allows machine learning to be executed on resource-constrained, embedded edge devices. Finally, the Node-Red server was used to test and display the different prediction results.