Dissertation/Thesis Abstract

An Automated Diabetic Retinopathy Detection System
by Bhoghoju, Aditya, M.S., California State University, Long Beach, 2017, 37; 10263619
Abstract (Summary)

Diabetic retinopathy (DR) is a complication seen in individuals with diabetes and is a condition marked by high levels of blood glucose. The current project presents an efficient automated diabetic retinopathy detection system, which identifies blood vessel leaks (known as exudates) in digital images of the fundus. This system detects and classifies exudates into two types: soft and hard. The project also presents and compares three techniques used for enhancing a digital fundus image: The Histogram Equalization (HE), the Contrast Limited Adaptive Histogram Equalization (CLAHE), and the Mahalanobis Distance (MD). Experimental results show that the MD-based technique is the best method for accurate and efficient detection and classification of exudates in a digital fundus image. The proposed system can assist ophthalmologists in monitoring the evolution of diabetic retinopathy, and implement interventions and treatment to preserve vision in patients with diabetes.

Indexing (document details)
Advisor: Chassiakos, Anastasios
Commitee: Khoo, I-Hung, Yeh, Hen-Geul
School: California State University, Long Beach
Department: Electrical Engineering
School Location: United States -- California
Source: MAI 56/04M(E), Masters Abstracts International
Subjects: Electrical engineering
Publication Number: 10263619
ISBN: 978-1-369-80508-6
Copyright © 2020 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy