Classificação automática de glaucoma utilizando o dataset JustRAIGS

Abstract

Glaucoma, a group of eye diseases that cause damage to the optic nerve, can lead to irreversible vision loss if left undetected. Fundus imaging is crucial for glaucoma diagnosis, but manual analysis is expensive and inconsistent. This paper presents a method that leverages the extensive JustRAIGS challenge dataset to train a Convolutional Neural Network (CNN) for automatic glaucoma detection. The recommendation model identifies images with referenceable glaucoma signals. As a preprocessing step, a pretrained segmentation model isolates the optic disc region for focused analysis. By addressing class imbalance through sampling techniques and employing pretrained CNN architectures, our method demonstrates superior performance when compared to a related work.