Shape and Scale in Quantifying Aortic Morphology Evolution and Chronicity.

Area of Science:

• Cardiovascular Imaging
• Medical Physics
• Computational Anatomy

Background:

• Thoracic aortic disease management relies on diameter thresholds, overlooking complex 3D morphology.
• Existing shape measures' utility is scale-dependent, limiting clinical application.

Purpose of the Study:

• Identify optimal scales for a size-invariant shape descriptor (normalized fluctuation in total integrated Gaussian curvature, δK~) in thoracic aortic disease.
• Relate the tuned δK~ signal to clinical markers of chronicity in aortic dissection.

Main Methods:

• Constructed a scale space of aortic surfaces from CTA images by varying smoothing, mesh density, and partitioning size.
• Computed δK~ at each scale, assessing signal robustness and clinical predictive value.
• Modeled disease progression using Gaussian Process Regression and linked δK~ to aortic chronicity markers.

Main Results:

• A stable zone emerged at centimeter-scale partitioning, maximizing δK~ informativeness and resilience to variability.
• δK~ improved preoperative risk stratification and, with Gaussian Process Regression, indicated progressive remodeling in dissection.
• The findings suggest δK~ captures disease chronicity beyond diameter alone.

Conclusions:

• Optimized scale space yielded a robust, clinically interpretable, size-invariant shape signal (δK~).
• The association of δK~ with chronicity supports its use as a complementary marker to diameter.
• Shape-aware, curvature-based decision tools warrant prospective validation for thoracic aortic disease management.