Virtual contrast for coronary vessels based on level set generated subvoxel accurate centerlines

Area of Science:

• Cardiovascular Imaging
• Medical Image Analysis
• Computational Cardiology

Background:

• Coronary artery disease is a leading cause of mortality.
• Accurate visualization of coronary vessels in cardiac MRI is crucial for diagnosis.
• Manual vessel tracing in MRI scans is time-consuming and subjective.

Purpose of the Study:

• To develop and validate an automated tool for coronary vessel tracking in cardiac MRI.
• To improve the efficiency and accuracy of heart disease screening using MRI.

Main Methods:

• A novel tool employing eigenvalue analysis of the Hessian and level set-based extraction for vessel centerline identification.
• User interaction via single clicks for vessel identification and endpoint adjustment.
• Visualization of identified vessels using a virtual contrast agent and spherical curved reformation.

Main Results:

• Successful validation on 40 MRI scans.
• Demonstrated accuracy in coronary vessel tracking.
• Significant time savings compared to manual vessel tracing methods.

Conclusions:

• The developed tool offers an accurate and efficient method for coronary vessel analysis in cardiac MRI.
• This technology has the potential to enhance the screening process for heart diseases.
• Automated vessel tracking can streamline cardiovascular imaging workflows.