HEART ECG SIGNAL CLASSIFICATION USING DEEP LEARNIG CNN MODEL AND MIT BIH DATASET ON REALIME DATA CAPTURED

Abstract

Electrocardiogram (ECG) signal classification is an important task in the early diagnosis and monitoring of cardiovascular diseases and cardiac arrhythmias. Manual analysis of ECG signals is time-consuming, requires expert cardiologists, and may produce inaccurate results due to human limitations. Therefore, automated ECG classification systems based on deep learning techniques have gained significant attention in modern healthcare applications.  This research presents a Convolutional Neural Network (CNN)-based deep learning model for real-time ECG signal classification using the MIT-BIH Arrhythmia Dataset and real-time captured ECG data. The proposed system aims to automatically classify different heartbeat patterns into normal and abnormal categories with high accuracy and efficiency.  Initially, ECG signals are preprocessed using noise removal, normalization, and segmentation techniques to improve signal quality and reduce unwanted interference. After preprocessing, the ECG data is fed into the CNN model, which automatically extracts important features from the signals without requiring manual feature engineering. The CNN architecture consists of convolution layers, pooling layers, fully connected layers, and an output classification layer for heartbeat prediction.  The MIT-BIH Arrhythmia Dataset is used for model training and testing because it is one of the most widely used benchmark datasets for ECG analysis. In addition, real-time ECG signals are captured using ECG sensors such as AD8232 integrated with microcontroller devices for practical evaluation of the proposed system.  The performance of the proposed model is evaluated using standard performance metrics including accuracy, precision, recall, and F1-score. Experimental results demonstrate that the CNN-based ECG classification system achieves high classification accuracy and performs effectively for detecting various heartbeat abnormalities. The proposed model also shows strong performance on real-time ECG data.  This research contributes to the development of intelligent healthcare systems by providing an automated, accurate, and real-time ECG signal classification framework. The proposed system can assist healthcare professionals in early disease diagnosis, continuous patient monitoring, and smart healthcare applications.