Topological Analysis for Arteriovenous Malformations via Computed Tomography Angiography: Part 2: Practical...

Area of Science:

• Radiology
• Medical Imaging
• Computational Anatomy

Background:

• Previous work introduced a technique to count abnormal vascular loops in 3D CT angiography.
• This method requires validation for quantitative assessment of soft-tissue arteriovenous malformations (AVMs).

Purpose of the Study:

• To assess the validity of a vascular loop counting technique for evaluating soft-tissue AVMs.
• To establish a quantitative method for AVM assessment using topological analysis.

Main Methods:

• Utilized CT angiography data from 19 soft-tissue AVMs and 18 controls.
• Extracted enhanced vascular lumens (>120 HU) and created skeletonized graph models.
• Calculated vascular loop structures using graph theory (1 - nodes + edges).

Main Results:

• AVMs showed significantly higher loop density (9.5 loops/ml) than controls (1.3 loops/ml) (P < 0.001).
• Untreated AVMs demonstrated increased loop numbers and density; treated AVMs showed significant decreases (P = 0.0156).
• Loop count reduction was more pronounced with entire lesion treatment compared to partial treatment.

Conclusions:

• Vascular loop count and density effectively indicate the presence, progression, and remission of soft-tissue AVMs.
• Topological analysis shows promise as a quantitative evaluation method for AVMs.