Automated ascending aorta delineation from ECG-gated computed tomography images

Area of Science:

• Cardiovascular imaging
• Medical image analysis
• Computational geometry

Background:

• Accurate geometrical quantification is crucial for understanding aortic growth and disease.
• Automated algorithms for landmark detection and measurement are under development.
• Objective methods are needed for precise aortic delineation.

Purpose of the Study:

• To introduce an objective, automated method for delineating the ascending aorta based on its geometrical properties.
• To automatically identify the sinotubular junction (proximal border) and the brachiocephalic artery origin (distal border).
• To evaluate the accuracy of the automated method against manual measurements.

Main Methods:

• Automated delineation of the ascending aorta using geometrical properties of the aortic wall.
• Identification of the sinotubular junction via mean surface curvature transition.
• Definition of the brachiocephalic artery origin using centerline cross-section analysis at the bifurcation.
• Validation against manual border identification by two experienced observers on 44 CT scans.

Main Results:

• The automated method demonstrated high accuracy in detecting proximal and distal borders of the ascending aorta.
• Median distance measurements were 0.89 mm (method) vs. 1.02 mm (observer) for the proximal border.
• Median inclination measurements were 4.66° (method) vs. 5.07° (observer) for the proximal border.
• Median distance measurements were 2.18 mm (method) vs. 2.39 mm (observer) for the distal border.
• Median inclination measurements were 7.13° (method) vs. 4.77° (observer) for the distal border.
• Accuracy was comparable to interobserver variability and previously reported methods.

Conclusions:

• The developed automated method provides accurate and objective delineation of the ascending aorta.
• This technique facilitates precise geometrical quantification for aortic research and clinical applications.
• The automated identification of key aortic landmarks improves the efficiency and reliability of measurements.