|Glaucoma is a potentially blinding degenerative disease of the optic nerve (REF Glaucoma Quigley Lancet). The only established therapeutic approach is to lower the pressure in the affected eyes. The intraocular pressure is the result of the complex dynamics of aqueous humor production and drainage within the eye. The preferred way to lower the pressure is to increase aqueous drainage. Among the two routes of aqueous drainage, uveoscleral and trabecular (or conventional) route, the latter is the one most frequently addressed by a number of emerging minimal invasive surgical techniques, called MIGS (minimally invasive glaucoma surgery) (REF MIGS). They aim at reducing aqueous flow resistance at the anterior part of the conventional drainage route by opening or bypassing the trabecular meshwork and/or destroying septae in Schlemm’s canal. Thus, access of aqueous humor to the more distal, intrascleral part of the drainage system – collector channels and intrascleral aqueous veins – should be facilitated. However, this concept relies on a functioning intrascleral drainage system at the sector where the procedure was applied (REF WECKER / OTERENDORP TRABEKTOM).
Our knowledge about the anatomy of the intrascleral aqueous drainage system is remarkably poor, largely based on the work of Timothy Ashton in the 1950s (REFs Ashton). The complex network of small diameter vessels is situated deep in the scleral tissue and, thus, difficult to image in vivo. Optical coherence tomography (OCT) has been applied with partial success, limited by the highly scattering scleral tissue surrounding the aqueous veins (REF Kageman, Schuman). Moreover, OCT-imaging does not provide functional data on aqueous flow. Aqeuous angiography using scanning laser ophthalmoscopes and fluorescent tracers, such as fluorescein or indocyanine green, can provide anatomical and functional data but is limited to the two-dimensional space (REF Huang).
We have recently established a novel imaging modality using lipid emulsions as contrast agent for optical coherence tomography angiography (OCT-A; REF Gottschalk). Thus, we obtained time-resolved 3D-datasets of the aqueous vein plexus in ex vivo pork eyes. Potentially, this technique could be applied to other mammalian species, including human. The high-resolution 3D-datasets of the intrascleral aqueous veins may provide valuable information on the properties of the network, such as the mean cross-sectional area, ramification and interindividual variability of these parameters. However, prior to analyse these datasets a segmentation algorithm has to be applied in order to extract the aqueous vessel data from surrounding non-relevant data and artefacts. For this study, we evaluated different segmentation algorithms to extract the signal derived from aqueous veins in 2D- and 3D-datasets of lipid-emulsion based OCT-A from ex vivo porcine eyes.