Automatic generation of anatomic characteristics from cerebral aneurysm surface models

Area of Science:

• Medical imaging analysis
• Computational geometry
• Vascular modeling

Background:

• Cerebral aneurysm research relies on surface models of vessel lumens.
• Deriving anatomical descriptors from these models is crucial for advanced analysis.
• Current methods may lack automated, anatomy-aware feature extraction.

Purpose of the Study:

• To develop an automated approach for decomposing adjacent vessels into near- and far-regions.
• To compute the axial plane for cerebral vasculature models.
• To demonstrate applications in automatic vessel ordering and viewpoint selection.

Main Methods:

• Utilized approximation methods for analyzing vessel cross-sections and area profiles.
• Employed transition zones between vessel regions for axial plane computation.
• Defined vessel order and viewpoint via projection and normal vectors related to the axial plane.

Main Results:

• Successfully applied the method to diverse vascular datasets.
• Demonstrated robustness against noise in automatic vessel decomposition.
• Achieved broad agreement with medical experts on defined vessel transition zones.

Conclusions:

• The algorithms are generalizable to various cerebral aneurysm configurations.
• Geometric information aids in correcting automated decomposition failures.
• Derived descriptors enhance visualization, exploration, and analysis of cerebral aneurysms.