A Hybrid Deep CNN Model for Abnormal Arrhythmia Detection Based on Cardiac ECG Signal

Area of Science:

• Biomedical Engineering
• Artificial Intelligence in Medicine
• Signal Processing

Background:

• Electrocardiogram (ECG) signals are crucial for diagnosing cardiovascular diseases (CVDs).
• Environmental noise can significantly impair the accuracy of ECG signal analysis.
• Robust algorithms are needed for reliable ECG interpretation.

Purpose of the Study:

• To develop a robust algorithm for accurate ECG signal classification.
• To enhance classification accuracy in the presence of environmental noise.
• To compare the performance of 1D and 2D CNN models for ECG analysis.

Main Methods:

• A 1D Convolutional Neural Network (CNN) with two convolutional and two down-sampling layers was designed.
• ECG data was transformed into 2D images for analysis with a 2D CNN model.
• The proposed models were trained and tested on the MIT-BIH arrhythmia database.

Main Results:

• The 1D CNN model achieved a classification accuracy of 97.38%.
• The 2D CNN model achieved a classification accuracy of 99.02%.
• Both models demonstrated superior performance compared to existing state-of-the-art algorithms.

Conclusions:

• The proposed 1D and 2D CNN models are effective for robust ECG signal classification.
• The 2D CNN approach shows improved accuracy for noisy ECG data.
• These models offer a promising solution for automated CVD diagnosis.