IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Earlier Glaucoma Detection Using Optimised Blood Vessel Segmentation With CNN

Earlier Glaucoma Detection Using Optimised Blood Vessel Segmentation With CNN
View Sample PDF
Author(s): A. Sindhu (JCT College of Engineering and Technology, India)
Copyright: 2026
Pages: 22
Source title: Sustainable Environment Monitoring Systems for Medical Care
Source Author(s)/Editor(s): Calin Ciufudean (Stefan cel Mare University of Suceava, Romania)
DOI: 10.4018/979-8-3373-5636-5.ch008

Purchase

View Earlier Glaucoma Detection Using Optimised Blood Vessel Segmentation With CNN on the publisher's website for pricing and purchasing information.

Abstract

Glaucoma, being one of the major causes of irreversible blindness, is largely asymptomatic in its early stages, and therefore, early detection is the hour of need. Traditional diagnostic techniques are time-consuming and are prone to variability. In response to this problem, we introduce a deep learning-based approach that efficiently segments the glaucoma-sensitive optic disc (OD) and optic cup (OC) using an ensemble of convolutional neural networks (CNNs). To ensure increased accuracy of predictions, the system uses majority voting and state-of-the-art Attention U-Net models with pre-trained ResNet34, Inceptionv3, and DenseNet121 backbones. Following training of the ensemble over a heterogeneous dataset of REFUGE, Drishti-GS, and RIM-ONE images, the ensemble outperforms in glaucoma classification and fundus image segmentation. The novelty lies in the complementary use of ensemble learning and backbone-specific Attention U-Nets, which improve segmentation accuracy and reduce model overfitting. Traditional preprocessing methods, Gaussian noise removal, and a CNN-based classification pipeline further enhance the accuracy of predictions. The system finally offers an efficient and automated way to enable early glaucoma screening and diagnosis.

Related Content

M. Vinayaka, B. Lokeshappa, Shanmukha N. T.. © 2026. 16 pages.
Kholis Kholis Ernawati. © 2026. 32 pages.
Cristina Elena Turcu, Corneliu Octavian Turcu. © 2026. 40 pages.
Muhammad Usman Tariq. © 2026. 28 pages.
Gaganpreet Kaur, Amandeep Kaur, Ramandeep Sandhu. © 2026. 22 pages.
C. V. Suresh Babu, V. Karunya Lydia, M. Jeevananthan, S. Sidharth. © 2026. 26 pages.
Renu Mishra, Adarsh Tiwari, Sneha Sinha, Anmol Kr. Sah, Mamta Narwaria. © 2026. 24 pages.
Body Bottom