Please use this identifier to cite or link to this item:
Title: An Automatic Grading System for Detection of Diabetic Retinopathy Severity Levels
Authors: Grewal, Reaya
Supervisor: Malhi, Avleen Kaur
Chopra, Palika
Keywords: Diabetic Retinopathy;Micro Aneurysms;CLAHE;Gray Level Co-occurrence Matrix;Support Vector Machine;K-Nearest Neighbor Algorithm;Receiver Operating Curve
Issue Date: 5-Aug-2019
Abstract: Diabetic retinopathy has become one of the most leading and recurrent cases of blindness among children and adults who have been suffering from diabetes for an extremely long period of time. The increase in the level of the glucose in the blood causes lack of oxygen in the veins which further fails to supply nutrition to the retina. The signals are sent to brain for formation of newer blood vessels that are weak and causes their breakdown, hence releasing blood spots and other lesions that hinder the vision capability of retina. Exudates, micro aneurysms, cotton wool spots and hemorrhages are the features of diabetic retinopathy. The existing researches have focused either on the detection of the features of diabetic retinopathy or on the use of image processing for grading in diabetic retinopathy. In our work, grading has been done to know the severity of diabetic retinopathy i.e. whether it is mild, moderate or severe using exudates and micro aneurysms in 1361 fundus images. An automated approach that uses image processing, features extraction and machine learning models to predict accurately the presence of the exudates and micro aneurysms and grade accordingly has been proposed. The research is carried out in two segments; one for exudates and another for micro aneurysms. The grading is done via exudates based upon their distance from macula whereas grading via micro aneurysms is done by calculating their count. For grading using exudates, Support Vector machine and K-Nearest neighbor show the highest accuracy of 92.1% and for grading using micro aneurysms, decision tree shows the highest accuracy of 99.9% in prediction of severity levels of the disease.
Description: Master of Engineering - CS
Appears in Collections:Masters Theses@CSED

Files in This Item:
File Description SizeFormat 
801732041_ReayaGrewal_ME Thesis.pdf2.4 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.