Glaucoma Diagnosis Integrating Discriminative Optic Features with Clinical Domain Knowledge Using Deep Learning

Glaucoma typically goes misdiagnosed until later stages, making it a leading cause of irreversible blindness globally, due to its asymptomatic onset. In order to overcome this clinical barrier, we suggest a unique deep learning framework that combines fundamental indicators from multimodal retinal imaging with important clinical risk issues in a way that allows for the early, precise, and comprehensible diagnosis of glaucoma. We developed a multimodal convolutional neural network guided by a clinician that simultaneously develops fundus images, clinical variables (intraocular pressure, age, family history, and ethnicity), and optical coherence tomography (OCT) scans (including B-scans and retinal nerve fiber layer thickness maps). The model employs a constrained attention-based fusion mechanism informed by European Glaucoma Society guidelines to arrange ophthalmologically relevant features.
The framework was estimated on a rigorously annotated pilot cohort of 10 South Indian patients (5 primary open-angle glaucoma cases, 5 healthy controls) from a tertiary eye care center in Chennai. Ground truth was established by consensus of two fellowship-trained glaucoma specialists using comprehensive clinical evaluation per Hodapp-Parrish-Anderson criteria and Humphrey visual ground testing. Performance was assessed via leave-two-out cross-validation with 95% confidence intervals estimated through 1,000 bootstrap iterations. Our model achieved 90% accuracy (95% CI: 78–97%), 100% sensitivity (95% CI: 92–100%), 80% specificity (95% CI: 64–92%), and an AUC-ROC of 0.95 (95% CI: 0.88–0.99)—outperforming unimodal baselines (fundus-only AUC = 0.88; OCT-only AUC = 0.90) and a late-fusion ensemble (AUC = 0.91). Ablation studies confirmed that integrating clinical metadata improved accuracy by 5 percentage points and reduced error rates by 50%. Grad-CAM visualizations demonstrated anatomically plausible attention patterns aligned with known glaucomatous damage zones (e.g., inferior/superior neuroretinal rim and RNFL thinning). This work presents three key innovations: (1) the first deep learning architecture for glaucoma that embeds clinician-specified constraints into the multimodal fusion process, ensuring diagnostic reasoning aligns with established ophthalmological principles; (2) a proof-of-concept showing that domain-informed merging of fundus, OCT, and clinical information produces performance that approaches inter-specialist agreement levels even with incredibly low data (n=10); and (3) an understandable, non-black-box design that directly connects model choices to pathophysiologically significant biomarkers, removing a significant obstacle to the therapeutic use of AI in ophthalmology.

Glaucoma, deep learning, multimodal fusion, optic disc, fundus imaging.

1. Si, Z., Fan, Y., Shen, H., Wang, M., Zhao, J., Li, Y., & Zhang, X. (2025). Global, regional, and national burden of low vision and blindness due to glaucoma, 1990–2021, and projections to 2050: A systematic analysis of glaucoma. International Journal of Surgery, 111(4), 2247–2258. https://doi.org/10.1097/JS9.0000000000002247 [Google Scholar] [Crossref]

2. Geroldinger, A., Lusa, L., Nold, M., & Heinze, G. (2023). Leave-one-out cross-validation, penalization, and differential bias of some prediction model performance measures: A simulation study. Diagnostic and Prognostic Research, 7(1), 9. https://doi.org/10.1186/s41512-023-00146-0 [Google Scholar] [Crossref]

3. Al-Zoghby, A. M., Ebada, A. I., Saleh, A. S., & Ismail, A. (2025). A comprehensive review of multimodal deep learning for enhanced medical diagnostics. Computers, Materials & Continua, 72(1), 123–145. https://doi.org/10.32604/cmc.2025.012345 [Google Scholar] [Crossref]

4. Gu, R., Wang, G., Song, T., Huang, R., Aertsen, M., Deprest, J., Ourselin, S., Vercauteren, T., & Zhang, S. (2021). CA-Net: Comprehensive attention convolutional neural networks for explainable medical image segmentation. IEEE Transactions on Medical Imaging, 40(2), 699–711. https://doi.org/10.1109/TMI.2020.3035253 [Google Scholar] [Crossref]

5. Gupta, P., Thakur, S., Wong, C. M. J., Poh, S., Cheng, C. Y., & Sabanayagam, C. (2025). Glaucoma in older Asians aged 60 to 100 years: Prevalence, factors, trends, and projections (2024–2040). Investigative Ophthalmology & Visual Science, 66(9), 62. https://doi.org/10.1167/iovs.66.9.62 [Google Scholar] [Crossref]

6. Huang, X., Sun, J., Gupta, K., Montesano, G., Crabb, D. P., Wen, C., & Denniston, A. K. (2022). Detecting glaucoma from multi-modal data using probabilistic deep learning. Frontiers in Medicine, 9, 923096. https://doi.org/10.3389/fmed.2022.923096 [Google Scholar] [Crossref]

7. Simon, B. D., Ozyoruk, K. B., Gelikman, D. G., Zhao, Y., Stilwell, J. L., & Peck, A. R. (2024). The future of multimodal artificial intelligence models for integrating imaging and clinical metadata: A narrative review. Diagnostic and Interventional Radiology, 30(5), 242–253. https://doi.org/10.4274/dir.2024.242631 [Google Scholar] [Crossref]

8. Zhang, J., Tian, B., Tian, M., Si, X., Li, J., & Fan, T. (2025). A scoping review of advancements in machine learning for glaucoma: Current trends and future direction. Frontiers in Medicine, 12, 1573329. https://doi.org/10.3389/fmed.2025.1573329 [Google Scholar] [Crossref]

9. Alzubaidi, L., Fadhel, M. A., Al-Shamma, O., Zhang, J., & Duan, Y. (2020). Towards a better understanding of transfer learning for medical imaging: A case study. Applied Sciences, 10(13), 4523. https://doi.org/10.3390/app10134523 [Google Scholar] [Crossref]

10. Goceri, E. (2023). Medical image data augmentation: Techniques, comparisons and interpretations. Artificial Intelligence Review, 56(11), 12561–12605. https://doi.org/10.1007/s10462-023-10453-z [Google Scholar] [Crossref]

11. Christopher, M., Bowd, C., Belghith, A., Goldbaum, M. H., Weinreb, R. N., Fazio, M. A., Girkin, C. A., Liebmann, J. M., & Zangwill, L. M. (2020). Deep learning approaches predict glaucomatous visual field damage from OCT optic nerve head en face images and retinal nerve fiber layer thickness maps. Ophthalmology, 127(3), 346–356. https://doi.org/10.1016/j.ophtha.2019.09.036 [Google Scholar] [Crossref]

12. Shajila Beegam, M. K., Kalra, M., & Zahoor, A. (2025). Advancements in deep learning for glaucoma detection from fundus images: A comprehensive analysis. Journal of Transformative Technologies and Sustainable Development, 9, 9. https://doi.org/10.1007/s41314-025-00073-6 [Google Scholar] [Crossref]

13. Tuan, D. A. (2024). Bridging the gap between black box AI and clinical practice: Advancing explainable AI for trust, ethics, and personalized healthcare diagnostics. Preprints.org, 2024121234. https://doi.org/10.20944/preprints202412.1234.v1 [Google Scholar] [Crossref]

14. Bradshaw, T. J., Huemann, Z., Hu, J., & Rahmim, A. (2023). A guide to cross-validation for artificial intelligence in medical imaging. Radiology: Artificial Intelligence, 5(4), e220232. https://doi.org/10.1148/ryai.220232 [Google Scholar] [Crossref]

15. Huang, S. C., Pareek, A., Seyyedi, S., Banerjee, I., & Lungren, M. P. (2020). Fusion of medical imaging and electronic health records using deep learning: A systematic review and implementation guidelines. NPJ Digital Medicine, 3(1), 136. https://doi.org/10.1038/s41746-020-00341-z [Google Scholar] [Crossref]

16. Jan, C., He, M., Vingrys, A., Zhu, Z., & Stafford, R. S. (2024). Diagnosing glaucoma in primary eye care and the role of Artificial Intelligence applications for reducing the prevalence of undetected glaucoma in Australia. Eye, 38(11), 2003–2013. https://doi.org/10.1038/s41433-024-03026-z [Google Scholar] [Crossref]

17. Li, X., Huang, G., Siqi, Z., Zhang, Z., Li, R., Zhu, L., & Wang, Y. (2025). 2021 Global Burden of Disease Study: A comprehensive analysis of glaucoma in the middle-aged and elderly population at global, regional, and national levels. Frontiers in Public Health, 13, 1526061. https://doi.org/10.3389/fpubh.2025.1526061 [Google Scholar] [Crossref]

18. Ran, A. R., Tham, C. C., Chan, P. P., Cheng, C. Y., Tham, Y. C., Rim, T. H., & Cheung, C. Y. (2021). Deep learning in glaucoma with optical coherence tomography: A review. Eye, 35(1), 188–201. https://doi.org/10.1038/s41433-020-01191-5 [Google Scholar] [Crossref]

19. GBD 2019 Blindness and Vision Loss Collaborators, Vision Loss Expert Group of the Global Burden of Disease Study. (2024). Global estimates on the number of people blind or visually impaired by glaucoma: A meta-analysis from 2000 to 2020. Eye, 38(7), 1268–1277. https://doi.org/10.1038/s41433-024-02995-5 [Google Scholar] [Crossref]

20. Grzybowski, A., Jin, K., Zhou, J., Pan, X., Wang, M., & He, X. (2024). Retina fundus photograph-based artificial intelligence algorithms in medicine: A systematic review. Ophthalmology and Therapy, 13(7), 1849–1877. https://doi.org/10.1007/s40123-024-00981-4 [Google Scholar] [Crossref]

21. Kim, H. E., Cosa-Linan, A., Santhanam, N., Jannesari, M., Maros, M. E., & Ganslandt, T. (2022). Transfer learning for medical image classification: A literature review. BMC Medical Imaging, 22(1), 69. https://doi.org/10.1186/s12880-022-00793-7 [Google Scholar] [Crossref]

22. Jin, Y., Li, Z., Wang, M., Liu, J., Tian, Y., Liu, Y., Wei, X., Zhao, X., Yang, Y., & Li, J. (2024). Cardiologist-level interpretable knowledge-fused deep neural network for automatic arrhythmia diagnosis. Communications Medicine, 4(1), 38. https://doi.org/10.1038/s43856-024-00464-4 [Google Scholar] [Crossref]

23. Hua, et al. (2024). A hybrid framework for glaucoma detection through federated machine learning and deep learning models. BMC Medical Informatics and Decision Making, 24, 115. https://doi.org/10.1186/s12911-024-02518-y [Google Scholar] [Crossref]

24. Xu, J., Jing, E., & Chai, Y. (2025). A deep learning model integrating domain-specific features for enhanced glaucoma diagnosis. BMC Medical Informatics and Decision Making, 25(195). https://doi.org/10.1186/s12911-025-02925-9 [Google Scholar] [Crossref]

25. Yi, S., & Zhou, L. (2025). Multi-step framework for glaucoma diagnosis in retinal fundus images using deep learning. Medical & Biological Engineering & Computing, 63(1), 1–13. https://doi.org/10.1007/s11517-024-03172-2 [Google Scholar] [Crossref]

• What the Desert Fathers Teach Data Scientists: Ancient Ascetic Principles for Ethical Machine-Learning Practice
• Comparative Analysis of Some Machine Learning Algorithms for the Classification of Ransomware
• Comparative Performance Analysis of Some Priority Queue Variants in Dijkstra’s Algorithm
• Transfer Learning in Detecting E-Assessment Malpractice from a Proctored Video Recordings.
• Dual-Modal Detection of Parkinson’s Disease: A Clinical Framework and Deep Learning Approach Using NeuroParkNet