International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 09 Issue: 06 | June 2022
p-ISSN: 2395-0072
www.irjet.net
Study on Glaucoma Detection Using CNN Jayraj F1, Aditya K2, Mahit M3, Nihal S4, Rahul K5, Pushpalatha S.Nikkam6 1Department
of Information Science and Engineering, SDMCET, Dharwad, Karnataka, India Department of Information Science and Engineering, SDMCET, Dharwad, Karnataka, India 6 Pushpalatha S. Nikkam , 6Assistant Professor, Department of Information Science and Engineering, SDMCET, Dharwad, Karnataka, India ---------------------------------------------------------------------***--------------------------------------------------------------------glaucoma. By several observational studies Abstract - Glaucoma is a persistent and incurable eye 2,3,4,5Student,
(Rotterdam Eye Study, Blue Mountains Eye Study, Visual Impairment Project, Proyecto VER, and Latino Eye Study), glaucoma is untreated in 50 percent of cases in the Western part of the world, with greater rates in specific ethnic groups, and up to 90 percent in developing countries. On the other side, glaucoma is frequently over treated: many patients are treated even when they have no illness. This strongly advocates for a global increase in illness detection precision.
condition that makes eyesight and life quality to deteriorate. We present a deep learning (DL) architecture using a convolutional neural network for automated glaucoma detection in this study. Deep learning systems, like as CNN models, can infer a hierarchical representation of pictures in order to distinguish between glaucoma and non-glaucoma patterns for diagnostic purposes. Six learned layers are included in the proposed DL architecture: four convolutional layers and two fully-connected layers. In this paper, we suggest a CNN method to glaucoma diagnosis. We create a network using Convolutional Neural Network (CNN) architecture and data augmentation to recognize the subtle elements involved in the classification job, such as microaneurysms, exudate, and haemorrhages on the retina.
Key Words:
1.2 Proposed System Glaucoma is an eye disorder that leads to lifelong blindness. Glaucoma is a chronic condition that can only be prevented if it is recognised properly at an early stage. The proposed method creates an automated glaucoma detection computer-aided system that allows ophthalmologists to accurately diagnose glaucoma patients early. The method uses a pre-processed fundus picture and extracts the optic cup and optic disc before calculating the Cup to disc ratio. To train and evaluate the classifier, intensity and textural information are taken from the picture. The outcomes of disease diagnosis using CDR are combined with characteristics to identify the picture as glaucoma or non-glaucoma suspect.
Convolutional Neural Network, Deep
Learning.
1. INTRODUCTION Glaucoma is a disorder that damages the optic nerve in your eye and worsens over time. It is frequently linked to greater in eye pressure. Glaucoma is usually hereditary and may not manifest it until later in life. The increased pressure, known as optic nerve, which sends images to the brain, can be damaged by intraocular pressure. Glaucoma can cause irreversible vision loss if the damage also isn't treated. Glaucoma, if left untreated, can result in complete and irreversible blindness within several years.
2. DESIGN AND DEVELOPMENT
1.1 Existing System
2.1 Objectives
Glaucoma is frequently detected too late because it is generally asymptomatic for years: half of all cases have mitigate to progressive disease first shows itself, it is in the worse eye, even in countries with high standards, massively increasing the disease's human and economic burden on individuals and society. Measurements of intraocular pressure (IOP) are used during regular eye exams, but they cannot distinguish between healthy and glaucomatous eyes up to half of glaucoma patients may not have an elevated IOP upon examination, and many ocular hypertensive patients do not require treatment and will never develop
© 2022, IRJET
|
Impact Factor value: 7.529
|
To develop an efficient Pre-processing /Data Augmentation technique.
To develop a Novel algorithm to extract features using CNN.
To develop high computational classifier to detect the Glaucomatous images
ISO 9001:2008 Certified Journal
|
Page 3302