Detection of Glaucoma Eye Disease Using Deep Learning

The motivation behind developing the “Glaucoma Eye Disease System' is that glaucoma disease does not have symptoms. Today, glaucoma is the leading cause of blindness. It damages optic nerve of eye. Glaucoma changes the thickness of the retinal nerve fiber layer. An increase in pressure inside the optic nerve head causes glaucoma. Glaucoma is an eye disease that is incurable. Manual glaucoma diagnosis is a time-consuming process. Also, this process is less accurate. Early-stage glaucoma disease detection may prevent complete eye vision loss. The deep learning process is useful in terms of image segmentation and classification. A convolutional neural network can extract features from an image to classify it as having glaucoma or being healthy. This system reduces manpower and complexity in glaucoma diagnosis. We trained our model using the ORIGA dataset and performed cross-validation on the same dataset. The model was tested on 65 images and achieved an accuracy of 89%.