ArteryX: A Reliable End-To-End Toolbox for Standardized Intracranial Artery Feature Extraction From 3D TOF-MRA

Area of Science:

• Neuroimaging
• Medical Image Analysis
• Cerebrovascular Research

Background:

• Quantitative analysis of intracranial arteries using time-of-flight magnetic resonance angiography (TOF-MRA) is crucial for cerebrovascular research.
• Existing processing pipelines suffer from inconsistent artery labeling and require significant manual correction.
• Standardized artery classification across proximal and distal vascular territories is needed.

Purpose of the Study:

• To present ArteryX, a novel toolbox for standardized artery feature extraction and classification from TOF-MRA.
• To improve the consistency and reduce the manual burden in analyzing intracranial arteries.
• To enable reproducible and comprehensive reporting of artery-specific features.

Main Methods:

• ArteryX integrates segmentation handling, isotropic geometric processing, vessel-fused graph construction, and constrained landmark-based classification.
• It extracts artery-level morphological, topological, and complexity features (e.g., length, radius, tortuosity, fractal dimensionality).
• Validation was performed using benchmarking datasets (TopBrain-Challenge), synthetic data, and an in vivo cohort with and without cerebral small vessel disease (CSVD).

Main Results:

• ArteryX demonstrated minimal bias and robust quantification performance across different segmentation sources, outperforming the iCafe toolbox.
• Agreement analyses showed minimal bias for radius and good sensitivity for extent-dependent metrics, even with noisy segmentations.
• A human-in-the-loop protocol with ArteryX required less manual intervention than iCafe.
• Exploratory in vivo analysis revealed group-level differences in ArteryX-derived features in CSVD patients not observed with iCafe.

Conclusions:

• ArteryX provides a standardized, reproducible, and efficient workflow for intracranial artery analysis from TOF-MRA.
• The toolbox significantly reduces manual correction burden and improves the accuracy and sensitivity of feature extraction compared to existing methods.
• ArteryX facilitates the discovery of clinically relevant vascular differences, particularly in conditions like CSVD.