Towards quantitative analysis of coronary CTA

Area of Science:

• Cardiovascular Imaging
• Medical Image Analysis
• Radiology

Background:

• Multi-slice computed tomography (MSCT) coronary angiography enables non-invasive 3D imaging of coronary arteries.
• Current visual inspection of stenosis in MSCT angiography achieves high sensitivity and specificity.
• There is a need for increased reproducibility in the quantitative analysis of coronary artery disease.

Purpose of the Study:

• To present a novel method for quantitative analysis of coronary artery disease with limited user interaction.
• To improve the reproducibility of MSCT coronary angiography analysis.
• To enable automated segmentation and quantification of coronary artery morphology.

Main Methods:

• Segmentation of coronary vessels using seed points.
• Extraction of vessel centerlines.
• Reformatting image volumes and contour detection.
• Quantification of vessel morphological parameters.

Main Results:

• The method allows segmentation of entire coronary trees with a single seed point.
• Quantitative analysis of vessel segments is possible with proximal and distal seed points.
• Vessel centerline deviation is minimal, with over 80% path coincidence when seed points are appropriately placed.
• No user variability is expected in quantification if seed points are placed away from lesions.

Conclusions:

• The developed method offers a reproducible approach to quantitative analysis of coronary artery disease from MSCT angiography.
• High image quality is crucial for automated quantitative analysis, a factor expected to improve with technological advancements in MSCT.
• This technique has the potential to enhance clinical practice by providing objective and reproducible assessments of coronary artery morphology.