Heart disease risk prediction based on deep learning multi-scale convolutional enhanced Swin Transformer model

Area of Science:

• Cardiology
• Artificial Intelligence
• Machine Learning

Background:

• Heart disease is a major global health concern, necessitating advanced diagnostic tools.
• Early prediction of heart disease is crucial for effective intervention and improved patient outcomes.

Purpose of the Study:

• To develop and evaluate a novel deep learning model for accurate heart disease risk assessment.
• To enhance the interpretability of AI models in cardiovascular risk prediction.

Main Methods:

• A multi-scale convolution-enhanced Swin Transformer (MSCST) model was developed.
• The model utilizes multi-branch convolutional networks with channel attention for feature extraction.
• Swin Transformer modules integrate global and local information via self-attention, with SHAP analysis for interpretability.

Main Results:

• The MSCST model achieved 89.42% accuracy and an AUC of 0.8908 on the Cleveland Heart Disease dataset.
• Performance surpassed traditional machine learning and existing deep learning approaches for heart disease prediction.

Conclusions:

• The MSCST model demonstrates significant potential for accurate and interpretable heart disease risk assessment.
• This AI-driven approach offers a promising advancement in early cardiovascular disease detection.