A texture-based method for choroid segmentation in retinal EDI-OCT images

Abstract

Retinal layers can be identified by ophthalmologists using OCT images, which is useful for the diagnosis of different diseases. Recent EDI-OCT technique allows to explore the choroid layer, whose segmentation has become one of the hottest topics in the field of retinal imaging. In this sense, and taking into account that the choroid layer has different visual properties than the other retinal layers, a methodology based on textural information is presented in this paper to segment the choroid. From a retinal EDI-OCT image, a region of interest is detected and its low-level features are extracted, generating a feature vector that describes it, to finally segment the choroid. This paper includes several texture analysis methods to calculate the feature vectors. Results provided by the proposed methodology showed that the approach is adequate for the problem at hand, since it allows to segment the choroid layer with promising results. © Springer International Publishing Switzerland 2015.