Skeleton-based abdominal aorta registration technique.

Area of Science:

• Medical Imaging
• Computational Anatomy
• Vascular Surgery

Background:

• Vascular diseases pose significant health challenges in developed nations.
• Accurate vascular segmentation and registration are crucial for disease assessment and treatment monitoring.
• Abdominal aortic aneurysms require precise evaluation of treatment efficacy.

Purpose of the Study:

• To develop and validate an abdominal aorta registration technique.
• To enable visualization of correspondences between pre- and postoperative Computed Tomography (CT) data for aneurysm treatment assessment.
• To match all voxels belonging to the aorta across different CT series.

Main Methods:

• A hybrid level-set active contour approach for aorta lumen segmentation.
• A path similarity skeleton graph matching procedure for registration.
• Testing on a database of 8 patients with two contrast-enhanced CT series each.

Main Results:

• The developed methodology successfully registered abdominal aorta data.
• All registration results were verified by an expert, demonstrating efficiency.
• The approach enables precise voxel-wise matching of aorta data from different CT scans.

Conclusions:

• The proposed abdominal aorta registration technique is effective for assessing abdominal aortic aneurysm treatment.
• The method's accuracy in matching pre- and postoperative CT data supports its clinical utility.
• Further development of this registration approach is warranted.