Comprehensive geometric modeling of human cerebral vasculature for quantitative vascular analysis

Abstract

In this work we developed a comprehensive and structured geometric model of human cerebral vasculature for quantitative anatomical analysis. We first proposed a general and structured geometric representation of the interconnected vascular network as a framework. We then described an image processing pipeline for the segmentation of vascular anatomy from discrete scalar field images, and applied the pipeline to segment the anatomical structures of cerebral vasculatures from whole brain magnetic resonance angiography (MRA) scans of healthy adult subjects. Next, we employed the proposed geometric representation to generate the comprehensive geometric model of human cerebral vasculature from those segmented anatomies. In the end, we performed quantitative anatomical analysis to the anterior cerebral arteries (ACA) and internal carotid arteries (ICA), and characterized their varying size and tortuosity in the cerebral arterial circulation.