Обработка и сегментация снимков глазного дна с использованием сверточных нейронных сетей

Abstract

The purpose of the study is the processing and segmentation of fundus images with different stages of diabetic retinopathy using convolutional neural networks. As a result, a trained neural network was obtained that allows segmenting fundus images with some accuracy. The accuracy of the determination is sufficient to verify the pathological processes of the retina with diabetic retinopathy.