Accurate quantification of vessel cross-sectional area using CT angiography: a simulation study

Area of Science:

• Medical Imaging
• Cardiovascular Imaging
• Radiology

Background:

• Coronary computed tomography (CT) angiography visualizes coronary atherosclerosis but faces limitations in stenosis severity assessment due to partial volume effects and calcification.
• A need exists for a more quantitative and accurate method to assess coronary stenosis severity.

Purpose of the Study:

• To develop and evaluate a novel semi-automated technique for quantitative assessment of coronary artery stenosis severity.
• To compare the accuracy of the new technique against traditional manual measurements.

Main Methods:

• Polyenergetic fan beam CT simulations were used, modeling contrast-enhanced lumens and lipid backgrounds to represent normal and stenotic coronary vessels.
• Simulations included vessels with varying diameters and area stenoses (10-90%), with occlusions created by lipid placement.
• Cross-sectional areas were measured manually by two readers and using a new semi-automated technique based on integrated Hounsfield units.

Main Results:

• Both manual and semi-automated techniques showed excellent correlation between measured and known cross-sectional areas for normal and stenotic vessels.
• The semi-automated integrated Hounsfield unit technique demonstrated over a twofold improvement in accuracy compared to the manual method.
• The new technique effectively corrects for the impact of coronary calcification on measurements.

Conclusions:

• The semi-automated integrated Hounsfield unit technique offers a significant accuracy improvement for quantitative coronary stenosis assessment.
• This method enhances the reliability of coronary CT angiography, particularly in the presence of calcification.
• The technique holds promise for more precise evaluation of coronary artery disease.