Volume 10 • Issue 2 • PP: 01–08 • 2025
Hybrid CNN–LSTM Architecture for OCT Retinal Disease Classification
Open Access & Copyright
© 2025 The Author(s). Published by ASPG. This article is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0).
Editorial responsibility:
This article was handled by an independent editor. The author serving on the Editorial Board had no involvement in the peer-review process or the final editorial decision.
View PDF Abstract
Optical coherence tomography (OCT) provides detailed retinal images, but their interpretation can be difficult across clinical settings and patient groups. This paper presents a hybrid Convolutional Neural Network–Long Short-Term Memory (CNN–LSTM) architecture for classifying retinal OCT images into eight disease categories. Two parallel convolutional branches extract spatial features; their outputs are pooled, reshaped into a sequence, and processed by an LSTM layer to model inter-feature relationships before dense classification. The model is evaluated on 24,000 images and compared with ResNet50, MobileNetV2, InceptionV3, and Xception. The proposed architecture achieves 93.00% accuracy, 94% precision, 93% recall, and a 0.93 F1-score, outperforming the compared models. The results indicate that combining convolutional feature extraction with sequence modeling is a promising approach to automated retinal disease classification.
Keywords
References
[1] S. Pang, B. Zou, X. Xiao, Q. Peng, J. Yan, W. Zhang, and K. Yue, “A novel approach for automatic classifi-cation of macular degeneration oct images,” Scientific Reports, vol. 14, p. 19285, Aug. 2024.
[2] Z. Tan, Q. Zhang, G. Lan, J. Xu, C. Ou, L. An, J. Qin, and Y. Huang, “Oct retinopathy classification via a semi-supervised pseudo-label sub-domain adaptation and fine-tuning method,” Mathematics, vol. 12, no. 2, p. 347, Jan. 2024.
[3] K. Du, S. Shah, S. C. Bollepalli, M. N. Ibrahim, A. Gadari, S. Sutharahan, J.-A. Sahel, J. Chhablani, and K. K. Vupparaboina, “Inter-rater reliability in labeling quality and pathological features of retinal oct scans: A customized annotation software approach,” PLOS ONE, vol. 19, no. 12, p. e0314707, Dec. 2024.
[4] C. Ou et al., “A deep learning network for accurate retinal multidisease diagnosis using multiview fusion of en face and b-scan images: A multicenter study,” Translational Vision Science & Technology, 2024.
[5] M. Guo, D. Gong, and W. Yang, “In-depth analysis of research hotspots and emerging trends in ai for retinal diseases over the past decade,” Frontiers in Medicine, 2024.
[6] A. M. Alenezi et al., “Multiscale attention-over-attention network for retinal disease recognition in oct radiology images,” Frontiers in Medicine, 2024.
[7] J. Bell et al., “Validation of inter-reader agree-ment/consistency for quantification of ellipsoid zone integrity and sub-rpe compartmental features across reti-nal diseases,” Diagnostics, 2024.
[8] Q. Zuo et al., “Multi-resolution visual mamba with multi-directional selective mechanism for retinal dis-ease detection,” Frontiers in Cell and Developmental Biology, 2024.
[9] J. Yang et al., “Dry age-related macular degeneration classification from optical coherence tomography im-ages based on ensemble deep learning architecture,” Frontiers in Medicine, 2024.
[10] A. Domalpally et al., “Data harmonization, standard-ization, and collaboration for diabetic retinal disease research: Report from the 2024 mary tyler moore vi-sion initiative workshop on data,” Translational Vision Science & Technology, 2024.
[11] M. Kulyabin et al., “Segment anything in optical coher-ence tomography: Sam 2 for volumetric segmentation of retinal biomarkers,” Bioengineering, 2024.
[12] J. Zhang, P. Zhao, Y. Zhao, C. Li, and D. Hu, “Few-shot class-incremental learning for retinal disease recogni-tion,” IEEE Journal of Biomedical and Health Informat-ics, 2024.
[13] D. J. Williamson et al., “Artificial intelligence to fa-cilitate clinical trial recruitment in age-related macular degeneration,” Ophthalmology Science, 2024.
[14] F. H. Rizk, M. E. Mohamed, B. Sameh, A. M. Zaki, M. M. Eid, and E.-S. M. El-Kenawy, “Predictive mod-eling of portuguese student performance: Compara-tive machine learning analysis,” in 2024 International Telecommunications Conference (ITC-Egypt), pp. 26–31, 2024.
[15] J. R. Taylor, J. Drinkwater, D. C. Sousa, V. Shah, and A. W. Turner, “Real-world evaluation of retcad deep-learning system for the detection of referable diabetic retinopathy and age-related macular degeneration,” Clin-ical and Experimental Optometry, 2024.
[16] A. A. A. El-Khalek et al., “A comprehensive review of ai diagnosis strategies for age-related macular degenera-tion,” Bioengineering, 2024.
[17] A. Laouarem, C. Kara-Mohamed, E.-B. Bourennane, and A. Hamdi-Cherif, “Htc-retina: A hybrid reti-nal diseases classification model using transformer-convolutional neural network from optical coherence to-mography images,” Computers in Biology and Medicine, 2024.
[18] J. Zhao, J. Zhu, J. He, G. Cao, and C. Dai, “Multi-label classification of retinal diseases based on fundus images using resnet and transformer,” Medical and Biological Engineering and Computing, 2024.
[19] S. V. Priyan, R. V. Kumar, C. Moorthy, and V. S. Nishok, “A fusion of deep neural networks and game theory for retinal disease diagnosis with oct images,” Journal of X- ray Science and Technology, 2024.
[20] J. Wang, H. Peng, S. Chen, and S. Ren, “Ensemble learning for retinal disease recognition under limited resources,” Medical and Biological Engineering and Computing, 2024.
[21] J. Carmichael, E. Costanza, A. Blandford, R. Struyven, P. A. Keane, and K. Balaskas, “Diagnostic decisions of specialist optometrists exposed to ambiguous deep-learning outputs,” Scientific Reports, 2024.
[22] A. J. Prabha, C. Venkatesan, M. S. Fathimal, K. K. Nithiyanantham, and S. P. A. Kirubha, “Rd-oct net: Hy-brid learning system for automated diagnosis of macular diseases from oct retinal images,” Biomedical Physics and Engineering Express, 2024.
[23] P. D. Angeli et al., “Splicing defects and crispr-cas9 cor-rection in isogenic homozygous photoreceptor precur-sors harboring clustered deep-intronic abca4 variants,” Molecular Therapy – Nucleic Acids, 2024.
[24] D. Wang, J. Lian, and W. Jiao, “Multi-label classifica-tion of retinal disease via a novel vision transformer model,” Frontiers in Neuroscience, 2024.
[25] J. Peng, J. Lu, J. Zhuo, and P. Li, “Multi-scale-denoising residual convolutional network for retinal disease clas-sification using oct,” Sensors, 2024.
[26] A. Torimura et al., “Profiling mirnas in tear extracellular vesicles: A pilot study with implications for diagnosis of ocular diseases,” Japanese Journal of Ophthalmology, 2023.
[27] M. Bhandari, T. B. Shahi, and A. Neupane, “Evalu-ating retinal disease diagnosis with an interpretable lightweight cnn model resistant to adversarial attacks,” Journal of Imaging, 2023.
[28] Y. Yan et al., “Clinical evaluation of deep learning sys-tems for assisting in the diagnosis of the epiretinal mem-brane grade in general ophthalmologists,” Eye, 2023.
[29] B. A. Hammou, F. Antaki, M.-C. Boucher, and R. Duval, “Mbt: Model-based transformer for retinal optical coher-ence tomography image and video multi-classification,” International Journal of Medical Informatics, 2023.
[30] P. Dutta, K. A. Sathi, M. A. Hossain, and M. A. A. De-wan, “Conv-vit: A convolution and vision transformer-based hybrid feature extraction method for retinal dis-ease detection,” Journal of Imaging, 2023.
[31] C.-T. Wu, T.-Y. Lin, C.-J. Lin, and D.-K. Hwang, “The future application of artificial intelligence and telemedicine in the retina: A perspective,” Taiwan Jour-nal of Ophthalmology, 2023.
[32] M. J. M. Zedan, M. A. Zulkifley, A. A. Ibrahim, A. M. Moubark, N. A. M. Kamari, and S. R. Abdani, “Au-tomated glaucoma screening and diagnosis based on retinal fundus images using deep learning approaches: A comprehensive review,” Diagnostics, 2023.
[33] S. A. Gocuk, J. K. Jolly, T. L. Edwards, and L. N. Ayton, “Female carriers of x-linked inherited retinal diseases: Genetics, diagnosis, and potential therapies,” Progress in Retinal and Eye Research, 2023.
[34] X. Huang et al., “Gabnet: Global attention block for retinal oct disease classification,” Frontiers in Neuro-science, 2023.
[35] N. N. El-Den et al., “Scale-adaptive model for detection and grading of age-related macular degeneration from color retinal fundus images,” Scientific Reports, 2023.
[36] A. R. Chłopowiec, K. Karanowski, T. Skrzypczak, M. Grzesiuk, A. B. Chłopowiec, and M. Tabakov, “Counteracting data bias and class imbalance – towards a useful and reliable retinal disease recognition system,” Diagnostics, 2023.
[37] M. A. Manan, F. Jinchao, T. M. Khan, M. Yaqub, S. Ahmed, and I. S. Chuhan, “Semantic segmentation of retinal exudates using a residual encoder-decoder archi-tecture in diabetic retinopathy,” Microscopy Research and Technique, 2023.
[38] M. D. Varela et al., “Artificial intelligence in retinal dis-ease: Clinical application, challenges, and future direc-tions,” Graefe’s Archive for Clinical and Experimental Ophthalmology, 2023.
[39] R. Greeshma and P. Simon, “Deep learning approach based plant seedlings classification with xception model,” in International Conference on IoT Based Con-trol Networks and Intelligent Systems. Springer, pp. 65–76, 2023.
[40] A. G. Howard et al., “Mobilenets: Efficient convolu-tional neural networks for mobile vision applications,” arXiv preprint, 2017.
[41] C. Szegedy et al., “Going deeper with convolutions,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1–9, 2015.
[42] C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, and Z. Wojna, “Rethinking the inception architecture for computer vision,” in Proceedings of the IEEE Confer-ence on Computer Vision and Pattern Recognition, pp. 2818–2826, 2016.
[43] F. Chollet, “Xception: Deep learning with depthwise separable convolutions,” in Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni-tion, pp. 1251–1258, 2017.
Cite This Article
Choose your preferred format
Publisher's Note
The statements, opinions, and data presented in this article are solely those of the author(s) and do not necessarily represent those of ASPG, the journal, or its editors. ASPG and the editors disclaim responsibility for any harm arising from the use of any ideas, methods, instructions, or products described in this article, to the fullest extent permitted by applicable law.