Morphology of Ocular Blood Vessels

INTRODUCTION Glaucoma is the 2nd most common cause of vision loss worldwide [1]. Nearly 80 million people are projected to have glaucoma by the year 2020, and there is currently no cure [2]. Glaucoma is characterized by optic neuropathy, which is damage to the optic nerve, and causes gradual loss of sight. Recent evidence suggests that blow flow regulation and perfusion in the eye may play a role in the development and progression of this disease [3]. To properly study blood flow and perfusion of the eye, the morphology of the ocular blood vessels must be quantified. One way that people study the morphology of blood vessels is by analyzing plastic casts of the vessels. Plastic corrosion casts are created by perfusing the vessels of the eye with liquid plastic, and allowing that plastic to harden. The surrounding tissue is then eroded away using strong acids, and what is left is the hardened plastic in the shape of the ocular vessels [4]. Previous attempts at quantifying the vessel morphology using these casts involved taking 2D pictures of them from different angles and stitching those images together to extract physical parameters [5,6]. However, this approach results in a lack of volumetric and 3D spatial data. There is also no representation of vessel networks beneath the surface layers of vasculature. The 3D structure of blood vessels is what determines the amount of blood flowing to certain areas of the eyes, making 3D morphological analysis critical. Changes in various physical parameters of blood vessels affect the resistance to flow within those vessels and which regions of the eye are receiving blood. One such parameter is tortuosity, which measures how curved the path of a blood vessel is. An increased tortuosity of blood vessels has been shown to be associated with hypertension [7]. Additionally, vessel diameter and branch density are determinants of how much blood is being distributed throughout the eye. To better study the morphology of these plastic casts, a digital 3D model of the casts must be generated and a method for extracting parameters from specific vascular regions of the eye must be developed. In our study, digital 3D models of the casts are acquired by imaging the casts with a micro computed tomography (μCT) scanner. This results in cross-section images of the casts, which can be stitched together to form a 3D rendering of the vessels. However, to extract parameters from specific vessels within the complete rendering, a method for identifying singular vessels needs to be developed. Many digital 3D image programs utilize a process called segmentation to identify specific regions of 3D scans. Segmentation involves selecting brightness thresholds within the images to differentiate structures and materials. Once the separate structures are identified, they can be extracted from the overall image as their own 3D rendering and their size and shape measured.