Glaucoma segmentation and diagnosis of retinal fundus image using deep learning methods

Abstract

Glaucoma is often termed the quotsilent thief of sightquot because it typically newlineshows no symptoms until significant damage has occurred. If not detected newlineearly, glaucoma can lead to irreversible blindness. The condition is newlinecharacterized by an increase in the Cup to Disc Ratio (CDR), which leads to newlinethe loss of peripheral vision. Glaucoma causes visual impairment through the newlinedegeneration of optic nerve fibers and cupping of the optic disc. This newlineprogressive optic neuropathy damages retinal ganglion cells and their axons, newlineleading to changes in the visual field. As the optic nerve deteriorates, it newlineultimately results in blindness. Diagnosis involves monitoring the visual field, newlineassessing the optic disc appearance, and measuring Intra-Ocular Pressure newline(IOP). Studies reveal that over one million Indians aged 40 and above are newlineaffected by glaucoma. newlineThis research uses deep learning algorithms to identify three distinct newlinemethods for glaucoma illness identification in retinal pictures. The goal of this newlineresearch work is to create an automated computer-aided system including newlinemodules for Modified LeNET (MLNET) classification, Optic Disc (OD) newlinesegmentation, feature computations, and preprocessing. There have been two newlineprocessing phases for the Glaucoma categorization system: training and newlinetesting. Using preprocessing, OD segmentation, and feature calculations from newlinethe segmented OD region, the training processing phase trains retinal pictures newlinefrom the known dataset, including both healthy and glaucoma images. The newlinesuggested MLNET classifier has further trained these OD region features. newlineUsing the sub processing modules for preprocessing, OD region newlinesegmentation, and feature computations, the testing processing phase newlineclassifies the unknown retinal picture into the healthy or glaucoma categories. newline