An approach to the quantitative assessment of retinal layer distortions and subretinal fluid in SD-OCT images

Abstract

A modern tool for age-related macular degeneration (AMD) investigation is Optical Coherence Tomography (OCT), which can produce high resolution cross-sectional images of retinal layers. AMD is one of the most frequent reasons for blindness in economically developed countries. AMD means degeneration of the macula, which is responsible for central vision. Since AMD affects only this specific part of the retina, untreated patients lose their fine shape- and face recognition, reading ability, and central vision. Here, we deal with the automatic localization of subretinal fluid areas and also analyze retinal layers, since layer information can help to localize fluid regions. We present an algorithm that automatically delineates the two extremal retinal layers, successfully localizes subretinal fluid regions, and computes their extent. We present our results using a set of SD-OCT images. The quantitative information can also be visualized in an anatomical context for visual assessment.