ECG Signal Processing for Classification of Diseases Using AI and Optimization Techniques

Abstract

: The electrocardiogram ECG is a test that determines the hearts electrical activity. It investigates symptoms of a possible heart problem, such as chest pain, palpitations suddenly noticeable heartbeats, dizziness, and shortness of breath. The two most common types of ECG abnormalities,sinus bradycardia, and sinus tachycardia, known as arrhythmia, have been focused in this work. This abstract provides an overview of various innovative hybrid techniques applied for ECG signal processing to classify different arrhythmia diseases, highlighting their effectiveness in improving accuracy and reliability. The ECG signal processing includes the preprocessing of signals, feature extraction,optimization, and classification. After preprocessing the signal using Discrete Wavelet Transform DWT,feature extraction is performed using QRS complex detection. The MIT-BIH arrhythmia databases ECG records are used to assess the classifiers performance. The first approach proposes a hybrid classifier utilizing a Genetic Algorithm GA for feature optimization and an Artificial Neural Network ANN for classification. Various parameters such as Accuracy, Precision, recall, F measure, Error and Execution time are utilized to evaluate the effectiveness of the proposed classifiers. This classifier achieves an accuracy of 92.94. In the second approach, another classifier comprising Cuckoo Search CS and Genetic Algorithm GA along with ANN and Support Vector MachineSVM is proposed for ECG signal classification. The accuracy achieved using this technique is 94.23. The third approach includes Cuckoo Search CS for optimization and Support Vector Machine SVM with Feedforward Back Propagation Neural Network FFBPNN, namely DWT CS SVM FFBPNN . The fourth technique incorporates Artificial Bee Colony ABC for feature extraction and optimization, coupled with a Convolutional Neural Network CNN for automated ECG signal classification for the detection of arrhythmia diseases.