Development of technology for choroidal state assessment based on pre-processing and quantitative analysis of void images from angio-OCT data

This paper presents a technology of selecting regions of interest in retinal angio-OCT images for conducting a quantitative analysis of choroidal parameters for the eye disease detection. The relevance of the research is evident from an important role of the choroid in the eye, which provides nourishment to the retina, thus ensuring its normal functioning. Abnormalities in choroid functioning lead to various eye diseases, including degenerative retinal diseases and glaucoma. We propose a method of choroidal condition assessment based on finding areas of vascular signal absence in the retinal angio-OCT images. A comparative analysis of the applicability of the proposed features for the classification of normality and pathology is carried out. The results of study obtained under different parameters of the algorithm of feature calculation reveal that the developed technology shows promise for designing a classifier. The results presented can be useful for specialists in ophthalmology and help to improve the diagnosis and treatment of eye diseases.