Automated segmentation and characterization of choroidal vessels in high-penetration optical coherence tomography

Abstract

An automated choroidal vessel segmentation and quantification method for high-penetration optical coherence tomography (OCT) was developed for advanced visualization and evaluation of the choroidal vasculature. This method uses scattering OCT volumes for the segmentation of choroidal vessels by using a multi-scale adaptive threshold. The segmented choroidal vessels are then processed by multi-scale morphological analysis to quantify the vessel diameters. The three-dimensional structure and the diameter distribution of the choroidal vasculature were then obtained. The usefulness of the method was then evaluated by analyzing the OCT volumes of normal subjects.