Glaucoma Diagnosis over Eye Fundus Image through Deep Features

Glaucoma is an ocular disease that causes damage to the eye's optic nerve and successive narrowing of the visual field in affected patients, causing an increase of intra-ocular pressure, which can lead the patient, in advanced stage, to blindness. This work presents a study on the use of Convolutional Neural Networks (CNNs) for the automatic diagnosis through eye fundus images. Therefore, a comparison was made among the main CNNs architectures for feature extraction. The features extracted were compared using different classifiers and tested on RIM-ONE datasets. The results are promising for the combination of ResNet and Logistic Regression, on the RIM-ONE-r2, with AUC of 0.957 and through InceptionResNet with the same classifier with AUC of 0.860 on the RIM-ONE-r3.