Thumbnail Image
Publication

ENHANCING GLAUCOMA DIAGNOSIS: FUNDUS IMAGE CLASSIFICATION/ANALYSIS APPROACH

Alharbi, Wafa
; 
Mateen, Kulsoom
; 
Sarfaraz, Safiya
Citations
Altmetric:
Type
Capstone
Supervisor
Salem, Nema
Subject
Glaucoma detection, fundus imaging, machine learning, segmentation, feature extraction, classification, AI-assisted diagnosis.
Date
Collections
ECE - Undergraduate works
Show all metadata
Research Projects
Organizational Units
Journal Issue
Abstract
This study presents an automated diagnostic system for early glaucoma detection using fundus imaging. The methodology integrates wavelet-based noise reduction, vessel removal, and watershed segmentation to enhance diagnostic accuracy. Validated on G1020 and HRF datasets, the system achieved an 85.75% accuracy, 89% precision, 92% recall, and 90.5% F1-score using SVM with Linear Discriminant Analysis (LDA). Designed for scalability, the system leverages open-source tools, making it cost-effective and applicable in resource-limited settings. Potential applications include telemedicine platforms and portable diagnostic kits, enabling early glaucoma screening and supporting Sustainable Development Goal 3. While challenges such as dataset variability remain, this work lays a foundation for advancing AIassisted ophthalmological diagnostics.
Department
Electrical and Computer Engineering
Publisher
Effat University
Sponsor
Copyright
Book title
Journal title
Open Access
http://hdl.handle.net/20.500.14131/2039
DOI
Embedded videos