Robust semi-automated path extraction for visualising stenosis of the coronary arteries

Area of Science:

• Medical Imaging
• Cardiovascular Disease
• Computer-Aided Diagnosis

Background:

• Computed tomography angiography (CTA) is crucial for diagnosing and planning treatments for heart disease.
• 3D visualization of coronary arteries is challenging due to contrast agent in surrounding structures like the aorta and left ventricle.

Purpose of the Study:

• To present a composite method for improved 3D visualization of coronary arteries in CTA.
• To develop a novel technique for vessel centerline extraction and lumen measurement to quantify stenosis.

Main Methods:

• A composite method combining segmentation and volume rendering techniques was developed.
• A novel Fast Marching minimal path cost function was employed for accurate vessel centerline extraction.
• Two volume visualization techniques were utilized, incorporating segmented arteries and lumen measurements.

Main Results:

• The proposed method effectively overcomes the difficulty of 3D coronary artery visualization caused by surrounding contrast agents.
• The novel centerline extraction method enables accurate computation of vessel lumen, providing a measure of stenosis.
• The system was successfully evaluated and demonstrated using both synthetic and clinical datasets.

Conclusions:

• The presented composite method significantly enhances 3D visualization of coronary arteries in CTA.
• The developed techniques offer a valuable tool for assessing coronary artery stenosis and aiding in cardiovascular disease management.
• This approach shows promise for improving diagnostic accuracy and treatment planning in cardiology.