Early Diagnosis of Types of Glaucoma Using Multi Feature Analysis Based on DBN Classification

Abstract

Glaucoma is one of the leading causes of blindness because it damages the eye's optic nerve and impairs vision. Early glaucoma diagnosis and treatment are critical to reduce the risk of permanent visual loss. The ultimate goal of this study is to spot early glaucoma symptoms by taking a variety of ocular characteristics into consideration. Glaucoma is difficult to diagnose sinceit does not become apparent until it destroys the eye and causes partial or whole vision loss. A remedy is required to discover this issue at an early stage by examining the retinal properties or features obtained using high-resolution imaging method. Since many of these illnesses share characteristics, it can be difficult for clinicians to identify the proper ailment for treatment, which makes the classification of these illnesses complicated. Previously many methods and techniques were implemented to detect the glaucoma, but the main objective is to classify which type of glaucoma a person is suffering from. Consequently, in this study, unsupervised deep belief network (DBN) is used to extract features at the depth level. So, by using DBN which considers multiple features for analysis in hidden layers whereas other algorithms consider one particular feature as input it gives better accuracy than other algorithms. Improved methods for diagnosing glaucoma sooner and with more accuracy willmake it easier to adopt efficient treatment choices quickly.