A Novel Method to Improve the Detection of Glaucoma Disease Using Machine Learning

Abstract

Glaucoma is the subsequent driving reason for long-lasting visual impairment around the world. Early identification of glaucoma can restrict the infection movement. This infection is an asymptomatic neurological sickness. Early glaucomatous eyes and unusual seeming eyes that show no proof of sickness movement after some time (e.g., physiologic measuring). To stay away from obstruction, the veins are divided and prohibited toward the starting through in-painting. The cup division is more troublesome than the circle division because of the presence of high thickness vascular design in the district of the optic cup crossing the cup limit. So, this paper proposes the original technique to upgrade the exhibition of the cup division strategy by remembering a technique for vessel identification and vessel for painting. Likewise, machine learning procedures will be applied to find the reasonable boundaries in a few equations, including edge discovery approach and limit level set approach. Substitute elements of glaucoma from fundus picture and utilization of various classifiers for additional working on the presentation of the strategy. The trial examination has been completed as far as exactness, accuracy, review, F1 score, RMSE and Guide.