Predictive Models of Coronary Artery Disease Based on Computational Modeling: The SMARTool System

Area of Science:

• Cardiology
• Medical Imaging
• Computational Biology

Background:

• Coronary artery disease (CAD) poses a significant global health burden.
• Accurate risk stratification and prediction are crucial for effective CAD management.
• Existing methods may not fully integrate diverse data sources for comprehensive risk assessment.

Purpose of the Study:

• To develop and validate a Decision Support System (DSS) for predicting coronary artery disease (CAD).
• To integrate clinical data, computed coronary tomography angiography (CCTA) imaging, and computational modeling for enhanced CAD prediction.
• To improve the accuracy of CAD risk stratification and patient outcomes.

Main Methods:

• Utilized data from 196 patients selected for DSS development from the SMARTool clinical trial (263 recruited).
• Collected traditional risk factors, blood examinations, and CCTA at baseline and follow-up (6.22 ± 1.42 years apart).
• Performed computational modeling of blood flow and low-density lipoprotein (LDL) transport at baseline.

Main Results:

• Developed predictive models integrating clinical, imaging, and computational data for CAD prediction.
• Achieved 83% accuracy in predicting CAD at follow-up.
• Identified low endothelial shear stress (ESS) and high LDL accumulation as key predictors when combined with imaging data.

Conclusions:

• The developed DSS demonstrates high accuracy in predicting CAD, offering a valuable tool for clinical decision-making.
• Integrating multi-modal data (clinical, imaging, computational) significantly enhances CAD prediction capabilities.
• This approach supports improved risk stratification and personalized treatment strategies for CAD patients.