Advancing cardiac diagnostics: high-accuracy arrhythmia classification with the EGOLF-net model.

Introduction: Arrhythmia, characterized by irregular heartbeats, can range from harmless to potentially life-threatening disturbances in heart rhythm. Effective detection and classification of arrhythmias are crucial for timely medical intervention and management. Methods: This research utilizes the MIT-BIH Arrhythmia Database, a well acknowledged benchmark dataset, to train and validate the proposed EGOLFNet model, Enhanced Gray Wolf Optimization with LSTM Fusion Network. This model integrates advanced optimization techniques with deep learning to enhance diagnostic accuracy and robustness in arrhythmia detection. The methodology includes preprocessing the ECG signals to normalize and filter out noise, followed by feature extraction using statistical methods and wavelet transforms. The distinctive aspect of EGOLF-Net involves using Enhanced Gray Wolf Optimization to select optimal features, which are then processed by LSTM layers to capture temporal dependencies in the ECG data effectively. Results and Discussion: The model achieved an accuracy of 99.61%, demonstrating the potential of EGOLF-Net as a highly reliable tool for classifying arrhythmias, significantly advancing the capabilities of cardiology diagnostic systems. Thus the proposed EGOLF-Net model was developed and validated for accurately identifying heart arrhythmias using electrocardiogram (ECG) data.