Meta-Learning based Few-Shot Classification of Retinal Diseases
DOI:
https://doi.org/10.5753/jbcs.2026.5898Keywords:
Few-shot learning, meta-learning, Reptile, eye disease, retina image, image classificationAbstract
To address sample scarcity, a common challenge in medical imaging, we investigate the Reptile meta-learning algorithm for few-shot disease classification in fundus images. In a setup with eight training classes and four testing classes, we investigate different architectures, training strategies, varying N-way K-shot configurations, and different data augmentation techniques. Our results show that Reptile outperforms standard transfer learning approaches in several settings and that, when combined with data augmentation, especially during evaluation, correct predictions tend to receive higher confidence scores. The quantitative results are further supported by an analysis of prediction confidence levels and activation-map visualizations.
Downloads
References
Cai, A., Chen, L., Chen, Y., Fang, J., Sun, M., and Chuan, Z. (2022). Pre-mocodiagnosis: Few-shot ophthalmic diseases recognition using contrastive learning. In IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pages 2059-2066. DOI: 10.1109/BIBM55620.2022.9994890.
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., Uszkoreit, J., and Houlsby, N. (2021). An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. In 9th International Conference on Learning Representations (ICLR). DOI: 10.48550/arxiv.2010.11929.
Finn, C., Abbeel, P., and Levine, S. (2017). Model-agnostic meta-learning for fast adaptation of deep networks. In 34th International Conference on Machine Learning, pages 1126-1135. DOI: 10.48550/arxiv.1703.03400.
Gildenblat, J. and contributors (2021). PyTorch library for CAM methods. Available at:[link].
He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770-778. DOI: 10.1109/CVPR.2016.90.
Hospedales, T., Antoniou, A., Micaelli, P., and Storkey, A. (2022). Meta-learning in neural networks: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(9):5149-5169. DOI: 10.1109/TPAMI.2021.3079209.
Huang, W., Huang, Y., and Tang, X. (2022). LesionPaste: One-Shot Anomaly Detection for Medical Images. arXiv:2203.06354. DOI: 10.48550/arxiv.2203.06354.
Karthik, Maggie, and Dane, S. (2019). APTOS 2019 Blindness Detection. Available at:[link]. Kaggle.
Kawasaki, R. and Grauslund, J. (2019). Chapter 2 - Clinical motivation and the needs for RIA in healthcare. In Trucco, E., MacGillivray, T., and Xu, Y., editors, Computational Retinal Image Analysis, The Elsevier and MICCAI Society Book Series, pages 5-17. Academic Press. DOI: 10.1016/B978-0-08-102816-2.00002-2.
Koch, G., Zemel, R., and Salakhutdinov, R. (2015). Siamese neural networks for one-shot image recognition. In ICML Deep Learning Workshop. Avaialble at:[link].
Li, N., Li, T., Hu, C., Wang, K., and Kang, H. (2021). A Benchmark of Ocular Disease Intelligent Recognition: One Shot for Multi-disease Detection. In Benchmarking, Measuring, and Optimizing, volume 12614, pages 177-193. Cham. Series Title: Lecture Notes in Computer Science. DOI: 10.1007/978-3-030-71058-3_11.
Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., and Guo, B. (2021). Swin Transformer: Hierarchical Vision Transformer using Shifted Windows. In International Conference on Computer Vision (ICCV), pages 9992-10002. DOI: 10.1109/ICCV48922.2021.00986.
Mahapatra, D., Ge, Z., and Reyes, M. (2022). Self-supervised generalized zero shot learning for medical image classification using novel interpretable saliency maps. IEEE Transactions on Medical Imaging, 41(9):2443-2456. DOI: 10.1109/TMI.2022.3163232.
Men, Y., Fhima, J., Celi, L. A., Ribeiro, L. Z., Nakayama, L. F., and Behar, J. A. (2023). DRStageNet: Deep Learning for Diabetic Retinopathy Staging from Fundus Images. arXiv:2312.14891. DOI: 10.48550/arxiv.2312.14891.
Nakayama, L. F., Goncalves, M., Zago Ribeiro, L., Santos, H., Ferraz, D., Malerbi, F., Celi, L. A., and Regatieri, C. (2024a). A Brazilian Multilabel Ophthalmological Dataset (BRSET). PhysioNet. Version 1.0.1. DOI: 10.13026/1pht-2b69.
Nakayama, L. F., Restrepo, D., Matos, J., Ribeiro, L. Z., Malerbi, F. K., Celi, L. A., and Regatieri, C. S. (2024b). BRSET: A Brazilian Multilabel Ophthalmological Dataset of Retina Fundus Photos. PLOS Digital Health, 3(7):1-16. DOI: 10.1371/journal.pdig.0000454.
Nakayama, L. F., Zago Ribeiro, L., Restrepo, D., Santos Barboza, N., Dias Fiterman, R., Vieira Sousa, M. l., Pereira, A. D. A., Regatieri, C., Malerbi, F. K., and Andrade, R. (2024c). mBRSET, a Mobile Brazilian Retinal Dataset. PhysioNet. Version 1.0. DOI: 10.13026/qxpd-1y65.
Ni, R., Goldblum, M., Sharaf, A., Kong, K., and Goldstein, T. (2021). Data augmentation for meta-learning. In 38th International Conference on Machine Learning, volume 139, pages 8152-8161. DOI: 10.48550/arxiv.2010.07092.
Nichol, A., Achiam, J., and Schulman, J. (2018). On first-order meta-learning algorithms. arXiv:1803.02999. DOI: 10.48550/arxiv.1803.02999.
Nurgazin, M. and Tu, N. (2023). A comparative study of vision transformer encoders and few-shot learning for medical image classification. In International Conference on Computer Vision Workshops (ICCVW), pages 2505-2513. DOI: 10.1109/ICCVW60793.2023.00265.
Pachetti, E. and Colantonio, S. (2023). A systematic review of few-shot learning in medical imaging. arXiv:2309.11433. DOI: 10.1016/j.artmed.2024.102949.
Perin, G. J. and Hirata, N. S. T. (2024). Few-shot Retinal Disease Classification on the Brazilian Multilabel Ophtalmological Dataset. In 2024 37th SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), pages 01-06, Manaus, Brazil. IEEE. DOI: 10.1109/SIBGRAPI62404.2024.10716320.
Rajeswaran, A., Finn, C., Kakade, S. M., and Levine, S. (2019). Meta-learning with implicit gradients. In Advances in Neural Information Processing Systems, volume 32. DOI: 10.48550/arxiv.1909.04630.
Rajpoot, A. and Seeja, K. R. (2023). Enhancing rare retinal disease classification: a few-shot meta-learning framework utilizing fundus images. Multimedia Tools and Applications, 83(18):55731–55749. DOI: 10.1007/s11042-023-17691-x.
Roychowdhury, S. (2021). Few shot learning framework to reduce inter-observer variability in medical images. In 25th International Conference on Pattern Recognition (ICPR), pages 4581-4588. DOI: 10.1109/ICPR48806.2021.9412620.
Singh, R., Bharti, V., Purohit, V., Kumar, A., Singh, A. K., and Singh, S. K. (2021). MetaMed: Few-shot medical image classification using gradient-based meta-learning. Pattern Recognition, 120:108111. DOI: 10.1016/j.patcog.2021.108111.
Snell, J., Swersky, K., and Zemel, R. (2017). Prototypical networks for few-shot learning. In 31st International Conference on Neural Information Processing Systems, pages 4080-4090. DOI: 10.48550/arxiv.1703.05175.
Song, Y., Wang, T., Cai, P., Mondal, S. K., and Sahoo, J. P. (2023). A comprehensive survey of few-shot learning: Evolution, applications, challenges, and opportunities. ACM Comput. Surv., 55(13). DOI: 10.1145/3582688.
Tazoar, M. S., Jyoti, O., Goni, M. O. F., and Rahman, M. M. (2024). Multiple Major Ocular Disease Classification Using Deep Learning Methods: A Comprehensive Review. In 2024 3rd International Conference on Advancement in Electrical and Electronic Engineering (ICAEEE), pages 1-6, Gazipur, Bangladesh. IEEE. DOI: 10.1109/ICAEEE62219.2024.10561679.
TorchVision Maintainers and contributors (2016). Torchvision: Pytorch's computer vision library. Available at:[link].
Vinyals, O., Blundell, C., Lillicrap, T., Kavukcuoglu, K., and Wierstra, D. (2016). Matching networks for one shot learning. In 30th International Conference on Neural Information Processing Systems, pages 3637-3645. DOI: 10.48550/arxiv.1606.04080.
Wang, Y., Yao, Q., Kwok, J. T., and Ni, L. M. (2020). Generalizing from a few examples: A survey on few-shot learning. ACM Comput. Surv., 53(3). DOI: 10.1145/3386252.
Wightman, R. (2019). PyTorch Image Models. [link].
Yun, S., Han, D., Chun, S., Oh, S. J., Yoo, Y., and Choe, J. (2019). CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features. In International Conference on Computer Vision (ICCV), pages 6022-6031. DOI: 10.1109/ICCV.2019.00612.
Zhang, H., Cisse, M., Dauphin, Y. N., and Lopez-Paz, D. (2018). mixup: Beyond empirical risk minimization. arXiv:1710.09412. DOI: 10.48550/arXiv.1710.09412.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Gabriel J. Perin, Diogo J. C. Alves, Lucas M. S. Sousa, Erik M. de Elias, Nina S. T. Hirata
This work is licensed under a Creative Commons Attribution 4.0 International License.
