Citation :

Malla Sireesha, Meka James Stephen, P.V.G.D. Prasad Reddy, "Enhancing Automated Glaucoma Detection: A Lightweight Hybrid Model with EfficientNetB3, CBAM, and Vision Transformers," International Journal of Engineering Trends and Technology (IJETT), vol. 73, no. 12, pp. 174-184, 2025. Crossref, https://doi.org/10.14445/22315381/IJETT-V73I12P114

Abstract

Glaucoma is one of the leading causes of permanent blindness globally, and has continued over the years as a dangerous eye condition that slowly damages the optic nerve. It is usually associated with high pressure in the eyes and can be affected by several factors, including age, family history, diabetes, and high blood pressure. Common tests such as tonometry, perimetry, and optical imaging are used to check for glaucoma, but they often find it difficult to detect the disease in its early stages, making early treatment harder. Recent advances in Deep Learning have resulted in several automated methods for detecting diseases from fundus images. There are many existing models that suffer from limitations such as a lack of clarity, generalization issues across datasets, and poor accuracy during times of confusion. The proposed EfficientNetB0 and Transformer architecture performed very well in automatically detecting glaucoma using normal fundus images. This architecture included an Explainable AI (XAI) method, which allows the decision-making process to be visually represented and enhances model accessibility by expanding the proposed model with EfficientNetB3 in place of EfficientNetB0, Vision Transformer (ViT), along with including the Convolutional Block Attention Model (CBAM), while working with a lightweight fundus image dataset. The Hybrid Deep Learning Model achieved the best AUC of 0.98 compared to all existing models.

Keywords

Lightweight Fundus Images, XAI, Grad-CAM, EfficientNetB3, CBAM, ViT.

References

[1] Xiao Chun Ling et al., “Deep Learning in Glaucoma Detection and Progression Prediction: A Systematic Review and Meta-Analysis,” Biomedicines, vol. 13, no. 2, pp. 1-26, 2025.
[CrossRef] [Google Scholar] [Publisher Link]

[2] Allen Beck, and Ta Chen Peter Chang, “Glaucoma: Definitions and Classification,” American Academy of Ophthalmology, 2015. [Online]. Available: https://www.aao.org/education/disease-review/glaucoma-definitions-classification

[3] Muhammad Naseer Bajwa et al., “Two-Stage Framework for Optic Disc Localization and Glaucoma Classification in Retinal Fundus Images using Deep Learning,” BMC Medical Informatics and Decision Making, vol. 19, no. 1, pp. 1-16, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[4] Diagnosing Glaucoma, The Glaucoma Foundation, 2018. [Online]. Available: https://glaucomafoundation.org/about-glaucoma-2/diagnosing-glaucoma/

[5] Glaucoma - Diagnosis and Treatment, Mayo Clinic, 2025. [Online]. Available: https://www.mayoclinic.org/diseases-conditions/glaucoma/diagnosis-treatment/drc-20372846

[6] Tareek Pattewar et al., “Glaucoma Detection using Machine Learning with Fundus Images,” Communications on Applied Nonllinear Analysis, vol. 32, no. 9s, pp. 1171-1184, 2025.
[CrossRef] [Publisher Link]

[7] Venkatesh Guntreddi, and Sivakumar V, “Deep Learning based Glaucoma Detection Using Majority Voting Ensemble of ResNet50, VGG16, and Swin Transformer,” Results in Engineering, vol. 28, pp. 1-13, 2025.
[CrossRef] [Google Scholar] [Publisher Link]

[8] Sajib Saha, Janardhan Vignarajan, and Shaun Frost, “A Fast and Fully Automated System for Glaucoma Detection using Color Fundus Photographs,” Scientific Reports, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[9] Guangzhou An et al., “Comparison of Machine-Learning Classification Models for Glaucoma Management,” Journal of Health care Engineering, vol. 2018, pp. 1-8, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[10] Sumaiya Pathan et al., “Automated Segmentation and Classifcation of Retinal Features for Glaucoma Diagnosis,” Biomedical Signal Processing and Control, vol. 63, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[11] Stefan Maetschke et al., “A Feature Agnostic Approach for Glaucoma Detection in Oct Volumes,” Plos One, vol. 14, no. 7, pp. 1-12, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[12] Gavin D’Souza, P.C. Siddalingaswamy, and Mayur Anand Pandya, “AlterNet-K: A Small and Compact Model for the Detection of Glaucoma,” Biomedical Engineering Letters, vol. 14, pp. 23-33, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[13] Fairouz Alsulami, et al., “HiGAN-CNN: A Hybrid Generative Adversarial Network and Convolutional Neural Network for Glaucoma Detection,” International Journal of Computer Science and Network Security, vol. 22, no. 9, pp. 23-30, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[14] Kaggle, EyePACS-AIROGS-light-V2 Leaderboard- ConvNeXtTiny Benchmark, 2023. [Online]. Available: https://github.com/TheBeastCoding/glaucoma-dataset-metadata/blob/main/benchmark-eyepacs-airogs-light-v2.md

[15] Oscar Perdomo et al., “Glaucoma Diagnosis from Eye Fundus Images based on Deep Morphometric Feature Extraction,” Computational Pathology and Ophthalmic Medical Image Analysis, Lecture Notes in Computer Science, pp. 319-327, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[16] S. Ajitha et al., “Identification of Glaucoma from Fundus Images using Deep Learning,” Indian Journal of Ophthalmology, vol. 69, no. 10, pp. 2702-2709, 2021. 
[CrossRef] [Google Scholar] [Publisher Link]

[17] Sajib Saha, Janardhan Vignarajan, and Shaun Frost, “A Fast and Fully Automated System for Glaucoma Detection using Color Fundus Photographs,” Scientific Reports, vol. 13, no. 1, pp. 1-11, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[18] Oluwatobi Joshua Afolabi et al., “The Use of U-Net Lite and Extreme Gradient Boost (XGB) for Glaucoma Detection,” IEEE Access, vol. 9, pp. 47411-47424, 2021. 
[CrossRef] [Google Scholar] [Publisher Link]

[19] Abeer Aljohani, and Rua Y. Aburasain, “A Hybrid Framework for Glaucoma Detection through Federated Machine Learning and Deep Learning Models,” BMC Medical Informatics and Decision Making, vol. 24, no. 1, pp. 1-16, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[20] Mir Tanvir Islam et al., “Deep Learning-Based Glaucoma Detection with Cropped Optic Cup and Disc and Blood Vessel Segmentation,” IEEE Access, vol. 10, pp. 2828-2841, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[21] Mamta Juneja et al., “Automated Detection of Glaucoma using Deep Learning Convolutional Network (G-net),” Multimedia Tools and Applications, vol. 79, no. 21-22, pp. 15531-15553, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[22] Ayesha Shoukat et al., “Automatic Diagnosis of Glaucoma from Retinal Images using Deep Learning Approach,” Diagnostics, vol. 13, no. 10, pp. 1-17, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[23] Ahmed Al-Mahrooqi et al., “GARDNet: Robust Multi-View Network for Glaucoma Classification in Color Fundus Images,” arXiv Preprint, pp. 1-11, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[24] Gavin D’Souza, P. C. Siddalingaswamy, and Mayur Anand Pandya, “AlterNet-K: A Small and Compact Model for the Detection of Glaucoma,” Biomedical Engineering Letters, vol. 14, no. 1, pp. 23-33, 2023. 
[CrossRef] [Google Scholar] [Publisher Link]

[25] Riley Kiefer, Glaucoma Dataset: EyePACS-AIROGS-light-V2, Kaggle. [Online]. Available: https://www.kaggle.com/datasets/deathtrooper/glaucoma-dataset-eyepacs-airogs-light-v2

[26] Chandra Nugraha, and Sri Hadianti, “Glaucoma Detection in Fundus Eye Images using Convolutional Neural Network Method with Visual Geometric Group 16 and Residual Network 50 Architecture,” Medical Informatics Technology, vol. 1, no. 2, pp. 36‑41, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[27] Yu Feng, Cong Wu & Yuan Zhou, “DG2Net: A MLP-Based Dynamixing Gate and Depthwise Group Norm Network for Glaucoma Detection,” Pattern Recognition, pp. 295-308, 2024.
[CrossRef] [Google Scholar] [Publisher Link]

[28] DiagIA, MobileNetV3 Model for Glaucoma Detection Kaggle Pipeline, 2025. [Online]. Available: https://github.com/DiagIA/retina-datasets.