A Novel Deep-Learning-Based Framework for the Classification of Cardiac Arrhythmia

Area of Science:

• Biomedical Engineering
• Artificial Intelligence in Medicine
• Cardiology

Background:

• Cardiovascular diseases (CVDs) are a leading cause of mortality worldwide.
• Electrocardiogram (ECG) signals are crucial for diagnosing heart conditions, including cardiac arrhythmia.
• Current diagnostic methods require continuous improvement for accuracy and efficiency.

Purpose of the Study:

• To propose a novel deep learning approach for classifying ECG signals.
• To accurately differentiate between normal heart rhythms and sixteen types of cardiac arrhythmia.
• To enhance the diagnostic capabilities for cardiovascular diseases using AI.

Main Methods:

• ECG signals were preprocessed and transformed into a 2D representation using Continuous Wavelet Transform (CWT).
• A deep convolutional neural network (D-CNN) with an attention block was employed to extract spatial features.
• A novel clump of features (CoF) framework and k-fold cross-validation were used for dimensionality reduction.

Main Results:

• The proposed deep learning framework achieved 99.84% accuracy.
• The model demonstrated 100% sensitivity and 99.6% specificity in classifying ECG signals.
• The algorithm outperformed existing state-of-the-art methods in accuracy, F1-score, and sensitivity.

Conclusions:

• The novel deep learning approach offers a highly accurate and efficient method for cardiac arrhythmia detection from ECG signals.
• The integration of CWT, D-CNN with attention, and the CoF framework provides robust feature extraction and dimensionality reduction.
• This AI-driven technique has the potential to significantly improve the early diagnosis and management of cardiovascular diseases.