Automated Detection and Classification of Ocular Diseases Using Deep Neural Networks
International Journal of Medical Toxicology and Forensic Medicine,
Vol. 16 (2026),
1 January 2026
,
Page 1-11
https://doi.org/10.22037/ijmtfm.v16.53486
Abstract
Background: To prevent irreparable vision loss, it is critical to identify eye diseases in the earliest stages and start therapy as soon as feasible. Color fundus photography for patient screening is a practical and cost-effective technique. There is a dire need for automatic, trustworthy diagnostic algorithms that rely on color fundus images, since early-stage disease often shows few evident symptoms. Disorders such as diabetic retinopathy, glaucoma, age-related macular degeneration (AMD), myopia, hypertension, and cataract are indicated by changes in surrounding anatomical components, such as blood vessels, the optic disc, the fovea, and the macula.
Method: We propose three models to classify multi-label fundus images of ophthalmological diseases using deep learning and convolutional neural network (CNN) techniques. The Ocular Disease Intelligent Recognition (ODIR) database is used for this investigation. This database contains fundus photographs of patients' left and right eyes, broken down into eight categories.
Results: Compared with the other two models, ResNet and AlexNet, the U-Net classifier achieved an accuracy of approximately 99.3%.
Conclusion: Within the scope of this work, we propose an automatic multi-class learning framework for detecting eye diseases using a convolutional neural network (CNN). The current approach has several shortcomings that future research should address.
- Fundus image, Ocular disease, CNN, ODIR, ResNet, U-Net, AlexNet
How to Cite
References
[1] Bourne RR, Stevens GA, White RA, Smith JL, Flaxman SR, Price H, et al. Causes of vision loss worldwide, 1990–2010: a systematic analysis. Lancet Glob Health. 2013;1(6):e339–49. [DOI: 10.1016/S2214-109X(13)70113-X]
[2] WebMD Editorial Contributors. Top causes of eye problems [Internet]. WebMD; 2024 Nov 6 [cited 2026 Oct 13]. [Link]
[3] Paudel P, Ramson P, Naduvilath T, Wilson D, Phuong HT, Ho SM, et al. Prevalence of vision impairment and refractive error in school children in Ba Ria–Vung Tau province, Vietnam. Clin Exp Ophthalmol. 2014;42(3):217–26. [DOI: 10.1111/ceo.12273]
[4] World Health Organization. Blindness and vision impairment [Internet]. Geneva: World Health Organization; 2026 Feb 10 [cited 2026 Oct 13]. [Link]
[5] He J, Li C, Ye J, Qiao Y, Gu L. Self-speculation of clinical features based on knowledge distillation for accurate ocular disease classification. Biomed Signal Process Control. 2021;67:102491. [DOI: 10.1016/j.bspc.2021.102491]
[6] Alam KN, Khan MM. CNN based COVID-19 prediction from chest X-ray images. In: 2021 IEEE 12th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON). IEEE; 2021. p. 486–92. [DOI: 10.1109/UEMCON53757.2021.9666508]
[7] Roy AG, Conjeti S, Karri SPK, Sheet D, Katouzian A, Wachinger C, et al. ReLayNet: retinal layer and fluid segmentation of macular optical coherence tomography using fully convolutional networks. Biomed Opt Express. 2017;8(8):3627–42. [DOI: 10.1364/BOE.8.003627]
[8] Litjens G, Kooi T, Bejnordi BE, Setio AAA, Ciompi F, Ghafoorian M, et al. A survey on deep learning in medical image analysis. Med Image Anal. 2017;42:60–88. [DOI: 10.1016/j.media.2017.07.005]
[9] Zhao W, Yang J, Sun Y, Li C, Wu W, Jin L, et al. 3D deep learning from CT scans predicts tumor invasiveness of subcentimeter pulmonary adenocarcinomas. Cancer Res. 2018;78(24):6881–9. [DOI: 10.1158/0008-5472.CAN-18-0696]
[10] Yang J, Deng H, Huang X, Ni B, Xu Y. Relational learning between multiple pulmonary nodules via deep set attention transformers. In: 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI). IEEE; 2020. p. 1875–8. [DOI: 10.1109/ISBI45749.2020.9098722]
[11] Liefers B, Venhuizen FG, Schreur V, van Ginneken B, Hoyng C, Fauser S, et al. Automatic detection of the foveal center in optical coherence tomography. Biomed Opt Express. 2017;8(11):5160–78. [DOI: 10.1364/BOE.8.005160]
[12] Meng X, Xi X, Yang L, Zhang G, Yin Y, Chen X. Fast and effective optic disk localization based on convolutional neural network. Neurocomputing. 2018;312:285–95. [DOI: 10.1016/j.neucom.2018.05.114]
[13] Katibeh M, Pakravan M, Yaseri M, Pakbin M, Soleimanizad R. Prevalence and causes of visual impairment and blindness in central Iran; the Yazd eye study. J Ophthalmic Vis Res. 2015;10(3):279–85. [DOI: 10.4103/2008-322X.170362]
[14] Wang W. Further results on mean-square exponential input-to-state stability of stochastic delayed Cohen-Grossberg neural networks. Neural Process Lett. 2023;55:3953–65. [DOI: 10.1007/s11063-022-10974-8]
[15] Huang C, Liu B, Yang H, Cao j. Positive almost periodicity on SICNNs incorporating mixed delays and D operator. Nonlinear Anal Model Control. 2022;27(4):719–39. [DOI: 10.15388/namc.2022.27.27417]
[16] Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition. arXiv [Preprint]. 2014. arXiv:1409.1556. [DOI: 10.48550/arXiv.1409.1556]
[17] Ashir AM, Ibrahim S, Abdulghani M, Ibrahim AA, Anwar MS. Diabetic retinopathy detection using local extrema quantized Haralick features with long short-term memory network. Int J Biomed Imaging. 2021;2021:6618666. [DOI: 10.1155/2021/6618666]
[18] Bibi I, Mir J, Raja G. Automated detection of diabetic retinopathy in fundus images using fused features. Phys Eng Sci Med. 2020;43(4):1253–64. [DOI: 10.1007/s13246-020-00929-5]
[19] Devarajan D, Ramesh SM, Gomathy B. A metaheuristic segmentation framework for detection of retinal disorders from fundus images using a hybrid ant colony optimization [retracted article]. Soft Comput. 2020;24(17):13347–56. [DOI: 10.1007/s00500-020-04753-7]
[20] Playout C, Duval R, Cheriet F. A novel weakly supervised multitask architecture for retinal lesions segmentation on fundus images. IEEE Trans Med Imaging. 2019;38(10):2434–44. [DOI: 10.1109/TMI.2019.2906319]
- Abstract Viewed: 35 times
- 53486 Downloaded: 17 times