Glaucoma Detection using Convolution Neural Networks

The World Health Organization (WHO) ranks glaucoma as the secondary main cause of vision loss and irreversible visual impairment. Due to its effects on the optic nerve head, glaucoma is sometimes called the “silent burglary of sight” by medical professionals. Finding and classifying glaucoma early on will help patients get the treatment they need from eye surgeons, which in turn will enhance their daily lives. Blinding glaucoma strikes without warning. It is extremely difficult to diagnose glaucoma in its early stages, and advanced cases of the disease cannot be treated at this time. Multiple automated eye disease detection systems were analyzed thoroughly for the study. A comprehensive literature review was conducted on the topics of which was before, extraction of features, extraction of selection, machine learning (ML) algorithm is proposed, and testing/training datasets. Machine learning, and especially deep learning techniques, have become the standard for analyzing medical images and data. Unfortunately, little has been accomplished in automating the prediction of glaucoma, despite its critical importance. Automated eye disease diagnosis is still in its infancy, but it has advanced to the point where most ML methods can correctly identify 85 percent of patients. Optical coherence tomography is a useful tool for the early detection of glaucoma (OCT). The use of machine learning strategies trained on eye unit of measure and field of vision data for the detection of glaucoma is also discussed.