Deep learning based classification network for diagnosis of glaucoma in two dimensional retinal fundus images

Abstract

Glaucoma is an eye condition which prompts lasting visual deficiency when the illness advances to a propelled stage. It happens because of high intraocular pressure inside the eye, bringing about harm to the optic nerve. Glaucoma doesn't show any side effects in its initial stage and in this manner, it is critical to analyze ahead of schedule to forestall visual impairment. Fundus photography is broadly utilized by ophthalmologists to aid analysis of glaucoma and is cost-effective. The utilization of Computer Aided Diagnosis(CAD) is successful in the conclusion of glaucoma and can help the clinicians to reduce their remaining task at hand altogether. We have likewise talked about the upsides of utilizing condition of-workmanship procedures, including Deep learning (DL), when building up the robotized framework. The DL techniques are powerful in glaucoma diagnosis. Novel DL calculations with large information accessibility are required to build up a dependable CAD framework. Such procedures can be utilized to analyze other eye diseases precisely. This paper examines the helpfulness of CAD frameworks and deep feed forward neural network to fairly analyze patients with glaucoma. Additionally, unique AI strategies utilized by scientists over the past decade are summarized.