Automated Glaucoma Detection In Retinal Fundus Images Using Deep Learning And Hybrid Feature Models
DOI:
https://doi.org/10.29284/ijasis.11.6s.2025.23-42Keywords:
Machine Learning, Deep learning, Glaucoma, Logistic Regression, Random Forest, Vgg16, Resnet50, MobileNet.Abstract
- Glaucoma is an eye disease that gets worse over time and can ruin your sight if it is not found early. Expert analysis is needed for traditional diagnostic methods, which can be subjective and take a lot of time. This study uses two standard datasets, ACRIMA and ORIGA, to test how deep learning techniques can be used to find glaucoma on its own. Within this study, deep learning models such as RestNet50, DenseNet, and Vgg16 are utilized to extract features and organize photos of individuals with and without glaucoma. Different views of the eyes can be found in the ORIGA dataset and the ACRIMA dataset. That way, they can see how well they teach machine learning models. At 84% of the times it was tried, DenseNet121 with SVM (RBF) was the most accurate model. With 81%, ResNet50 with SVM (RBF) came in second, and VGG16 with Random Forest got 80%. Each of these types worked well and did a good job. Tests show that models learned on ACRIMA are more sensitive, while models learned on ORIGA are more general because of its variety. Combining traits from both datasets improves classification performance, which suggests that a strong glaucoma detection system might be possible. This study shows how important diverse datasets and advances in machine learning are for finding glaucoma earlier and helping eye doctors treat and intervene in time.
References
[1] M. H. Wang et al., “AI-Based Advanced Approaches and Dry Eye Disease Detection Based on Multi-Source Evidence: Cases, Applications, Issues, and Future Directions,” Big Data Min. Anal., vol. 7, no. 2, pp. 445–484, 2024, doi: 10.26599/BDMA.2023.9020024.
[2] Y. Luo et al., “Harvard Glaucoma Fairness: A Retinal Nerve Disease Dataset for Fairness Learning and Fair Identity Normalization,” IEEE Trans. Med. Imaging, vol. 43, no. 7, pp. 2623–2633, 2024, doi: 10.1109/TMI.2024.3377552.
[3] A. Albelaihi and D. M. Ibrahim, “DeepDiabetic: An Identification System of Diabetic Eye Diseases Using Deep Neural Networks,” IEEE Access, vol. 12, pp. 10769–10789, 2024, doi: 10.1109/ACCESS.2024.3354854.
[4] N. Kalita and S. K. Borgohain, “An Ocular Feature-Based Novel Biomarker Determination for Glaucoma Diagnosis Using Supervised Machine Learning and Fundus Imaging,” IEEE Sensors Lett., vol. 8, no. 11, pp. 1–4, 2024, doi: 10.1109/LSENS.2024.3483990.
[5] R. F. Mansour and A. Al-Marghilnai, “Glaucoma detection using novel perceptron based convolutional multi-layer neural network classification,” Multidimens. Syst. Signal Process. vol. 32, no. 4, pp. 1217–1235, 2021, doi: 10.1007/s11045-021-00781-0.
[6] M. Melo Ferreira, G. Braz Junior, J. D. de Almeida, and A. Cardoso Paiva, “Glaucoma Grading Using Multimodal Imaging and Multilevel CNN,” IEEE Lat. Am. Trans., vol. 21, no. 10, pp. 1095–1102, 2023, doi: 10.1109/TLA.2023.10255449
[7] A. Neto, J. Camera, S. Oliveira, A. Cláudia, and A. Cunha, “Optic disc and cup segmentations for glaucoma assessment using cup-to-disc ratio,” Procedia Comput. Sci., vol. 196, pp. 485–492, 2022, doi: https://doi.org/10.1016/j.procs.2021.12.040.
[8] E. P. Ong et al., “Automatic Glaucoma Detection from Stereo Fundus Images.,” Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. IEEE Eng. Med. Biol. Soc. Annu. Int. Conf., vol. 2020, pp. 1540–1543, Jul. 2020, doi: 10.1109/EMBC44109.2020.9175923.
[9] A. R. Prananda, E. L. Frannita, A. H. T. Hutami, M. R. Maarif, N. L. Fitriyani, and M. Syafrudin, “Retinal Nerve Fiber Layer Analysis Using Deep Learning to Improve Glaucoma Detection in Eye Disease Assessment,” Appl. Sci., vol. 13, no. 1, 2023, doi: 10.3390/app13010037.
[10] J. Pruthi, “Review of Image Processing Technique for Glaucoma Detection,” vol. 2, pp. 99–105, 2013.
[11] A. Sarhan, J. Rokne, and R. Alhajj, “Glaucoma detection using image processing techniques: A literature review,” Comput. Med. Imaging Graph., vol. 78, p. 101657, 2019, doi: https://doi.org/10.1016/j.compmedimag.2019.101657.
[12] M. Shi et al., “Artifact-Tolerant Clustering-Guided Contrastive Embedding Learning for Ophthalmic Images in Glaucoma,” IEEE J. Biomed. Heal. Informatics, vol. 27, no. 9, pp. 4329–4340, 2023, doi: 10.1109/JBHI.2023.3288830.
[13] A. Singh, M. K. Dutta, M. ParthaSarathi, V. Uher, and R. Burget, “Image processing based automatic diagnosis of glaucoma using wavelet features of segmented optic disc from fundus image,” Comput. Methods Programs Biomed., vol. 124, no. c, pp. 108–120, 2016, doi: 10.1016/j.cmpb.2015.10.010.
[14] Y. Xu et al., “A hierarchical deep learning approach with transparency and interpretability based on small samples for glaucoma diagnosis,” nature.comY Xu, M Hu, H Liu, H Yang, H Wang, S Lu, T Liang, X Li, M Xu, L Li, H Li, X Ji, Z Wang, L LiNPJ Digit. Med. 2021•nature.com, Accessed: Nov. 26, 2023. [Online]. Available: https://www.nature.com/articles/s41746-021-00417-4
[15] D. A. Yunitasari, R. Sigit, and T. Harsono, “Glaucoma Detection Based on Cup-to-Disc Ratio in Retinal Fundus Image Using Support Vector Machine,” in 2021 International Electronics Symposium (IES), 2021, pp. 368–373. doi: 10.1109/IES53407.2021.9593943.
[16] F. Z. Zulfira, S. Suyanto, and A. Septiarini, “Segmentation technique and dynamic ensemble selection to enhance glaucoma severity detection,” Comput. Biol. Med., vol. 139, p. 104951, 2021, doi: https://doi.org/10.1016/j.compbiomed.2021.104951.
[17] R. Shinde, “Glaucoma detection in retinal fundus images using U-Net and supervised machine learning algorithms,” Intell. Med., vol. 5, p. 100038, 2021, doi: https://doi.org/10.1016/j.ibmed.2021.100038.
[18] N. Thakur and M. Juneja, “Classification of glaucoma using hybrid features with machine learning approaches,” Biomed. Signal Process. Control, vol. 62, p. 102137, 2020, doi: https://doi.org/10.1016/j.bspc.2020.102137.
[19] J. Pruthi, “Review of Image Processing Technique for Eye disesase Detection,” vol. 2, pp. 99–105, 2013.
[20] A. R. Prananda, E. L. Frannita, A. H. T. Hutami, M. R. Maarif, N. L. Fitriyani, and M. Syafrudin, “Retinal Nerve Fiber Layer Analysis Using Deep Learning to Improve Eye disesase Detection in Eye Disease Assessment,” Appl. Sci., vol. 13, no. 1, 2023, doi: 10.3390/app13010037.
[21] D. Božić-Štulić, M. Braović, and D. Stipaničev, “Deep learning based approach for optic disc and optic cup semantic segmentation for Eye disesase analysis in retinal fundus images,” Int. J. Electr. Comput. Eng. Syst., vol. 11, no. 2, pp. 111–120, 2020, doi: 10.32985/IJECES.11.2.6.
[22] E. P. Ong et al., “Automatic Eye disesase Detection from Stereo Fundus Images.,” Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. IEEE Eng. Med. Biol. Soc. Annu. Int. Conf., vol. 2020, pp. 1540–1543, Jul. 2020, doi: 10.1109/EMBC44109.2020.9175923.
[23] A. Neto, J. Camera, S. Oliveira, A. Cláudia, and A. Cunha, “Optic disc and cup segmentations for Eye disesase assessment using cup-to-disc ratio,” Procedia Comput. Sci., vol. 196, pp. 485–492, 2022, doi: https://doi.org/10.1016/j.procs.2021.12.040.
[24] F. Z. Zulfira, S. Suyanto, and A. Septiarini, “Segmentation technique and dynamic ensemble selection to enhance Eye disesase severity detection,” Comput. Biol. Med., vol. 139, p. 104951, 2021, doi: https://doi.org/10.1016/j.compbiomed.2021.104951.
[25] Y. Xu et al., “A hierarchical deep learning approach with transparency and interpretability based on small samples for Eye disesase diagnosis,” nature.comY Xu, M Hu, H Liu, H Yang, H Wang, S Lu, T Liang, X Li, M Xu, L Li, H Li, X Ji, Z Wang, L LiNPJ Digit. Med. 2021•nature.com, Accessed: Nov. 26, 2023. [Online]. Available: https://www.nature.com/articles/s41746-021-00417-4
[26] D. A. Yunitasari, R. Sigit, and T. Harsono, “Eye disesase Detection Based on Cup-to-Disc Ratio in Retinal Fundus Image Using Support Vector Machine,” in 2021 International Electronics Symposium (IES), 2021, pp. 368–373. doi: 10.1109/IES53407.2021.9593943.
[27] R. F. Mansour and A. Al-Marghilnai, “Eye disesase detection using novel perceptron based convolutional multi-layer neural network classification,” Multidimens. Syst. Signal Process., vol. 32, no. 4, pp. 1217–1235, 2021, doi: 10.1007/s11045-021- 00781-0.
[28] M. Ashtari-majlan, M. Mahdi, D. Member, D. Masip, and S. Member, “Spatial-aware Transformer-GRU Framework for Enhanced Glaucoma Diagnosis from 3D OCT Imaging,” pp. 1–9.
[29] R. Swaminathan, “An Attention Infused Deep Learning System with Grad-CAM Visualization for Early Screening of Glaucoma”.
[30] N. A. Alkhaldi and R. E. Alabdulathim, “Optimizing Glaucoma Diagnosis with Deep Learning-Based Segmentation and Classification of Retinal Images,” Appl. Sci., vol. 14, no. 17, 2024, doi: 10.3390/app14177795.
[31] I. Govindharaj, R. Rampriya, G. Michael, S. Yazhinian, K. Dinesh Kumar, and R. Anandh, “Capsule network-based deep learning for early and accurate diabetic retinopathy detection,” Int. Ophthalmol., vol. 45, no. 1, p. 78, 2025, doi: 10.1007/s10792-024-03391-4.
[32] S. Chakraborty, A. Roy, P. Pramanik, V. Daria, and R. Sarkar, “A Dual Attention-aided DenseNet-121 for Classification of Glaucoma from Fundus Images.” 2024. doi: 10.48550/arXiv.2406.15113.
[33] R. Hemelings et al., “A generalizable deep learning regression model for automated glaucoma screening from fundus images,” npj Digit. Med., vol. 6, no. 1, p. 112, 2023, doi: 10.1038/s41746-023-00857-0.
[34] R. Hemelings et al., “A generalizable deep learning regression model for automated glaucoma screening from fundus images,” npj Digit. Med., vol. 6, no. 1, p. 112, 2023, doi: 10.1038/s41746-023-00857-0.
[35] Y. Shi, Harnessing Deep Learning for Early Glaucoma Detection : A Review of CNN-Based Diagnostic Models, no. Icaic. Atlantis Press International BV, 2024. doi: 10.2991/978-94-6463-512-6.
[36] B. Al-Bander, W. Al-Nuaimy, M. A. Al-Taee, and Y. Zheng, “Automated glaucoma diagnosis using deep learning approach,” in 2017 14th International Multi-Conference on Systems, Signals & Devices (SSD), 2017, pp. 207–210. doi: 10.1109/SSD.2017.8166974.
[37] D. Muduli et al., “Retinal imaging based glaucoma detection using modified pelican optimization based extreme learning machine,” Sci. Rep., vol. 14, no. 1, p. 29660, 2024, doi: 10.1038/s41598-024-79710-7.
[38] D. Meedeniya, T. Shyamalee, G. Lim, and P. Yogarajah, “Glaucoma identification with retinal fundus images using deep learning: Systematic review,” Informatics Med. Unlocked, vol. 56, p. 101644, 2025, doi: https://doi.org/10.1016/j.imu.2025.101644.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
