Glaucoma Disease Diagnosis from Retinal Fundus Images using Deep Learning Model

Glaucoma could be an incessant and irreversible eye malady that disables vision and quality of life. The main objective of this paper is to perform automatic disease detection of glaucoma using deep neural network. A complete convolutional neural network (CNN) system has been developed, with the help of which the glaucoma diagnosis can be easily detected. The VGG16 convolutional neural network (CNN) model is proposed for image classification from the Kaggle glaucoma fundus imaging dataset. A total of six layers were required for training this dataset, including five transform layers and one fully connected layer. By trending this model on the dataset we got 99.8% which is much better than the state of art algorithm. This technique can be used for automatic glaucoma detection.