Application of stacked convolutional and long short-term memory network for accurate identification of CAD ECG...

Area of Science:

• Cardiology
• Artificial Intelligence
• Biomedical Engineering

Background:

• Coronary Artery Disease (CAD) is a leading global cause of heart disease, often asymptomatic in early stages.
• Electrocardiogram (ECG) is a common diagnostic tool for CAD but suffers from low sensitivity due to challenges in interpreting low-amplitude signals.
• Manual interpretation of ECGs can be error-prone, necessitating automated diagnostic solutions.

Purpose of the Study:

• To develop and implement a deep learning model for automated and objective interpretation of ECG signals for CAD diagnosis.
• To enhance the diagnostic accuracy and reliability of CAD detection from ECG data.

Main Methods:

• Implementation of a hybrid deep learning architecture integrating Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNN).
• Training and validation of the model on ECG signals to identify patterns indicative of CAD.
• Utilizing a blindfold strategy to evaluate the model's diagnostic performance.

Main Results:

• The proposed LSTM-CNN model achieved a high diagnostic accuracy of 99.85% in detecting CAD from ECG signals.
• The model demonstrates significant potential for accurate and automated CAD diagnosis.

Conclusions:

• The developed deep learning model offers a promising approach for accurate automated CAD diagnosis using ECGs.
• The prototype is ready for further validation with large-scale clinical databases before potential deployment.