{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T05:06:17Z","timestamp":1779771977164,"version":"3.53.1"},"reference-count":45,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T00:00:00Z","timestamp":1779753600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>Hybrid attention few-shot models are promising for automated retinal disease classification in data-scarce settings, yet two issues persist: static fusion of global and local features, and weak performance on minority classes. We propose Adaptive Gating and Focal Debiasing (AGFD), which adds a Dynamic Attention Gating (DAG) module that learns input-specific weights for global and local attention branches, and replaces cross-entropy with focal loss to shift learning toward hard, underrepresented cases. On ODIR-5K under episodic evaluation, AGFD improves over a strong hybrid-attention baseline across all shot settings, reaching 78.7% accuracy and 76.9% macro-F1 in the 5-shot setting, and 83.2% accuracy at 10-shot. Minority classes benefit most, with +11\u201315 percentage-point F1 gains for Glaucoma, Cataract, Hypertension, and Other. Gate analysis shows higher global weighting for widespread conditions and higher local weighting for lesion-driven diseases. Coupling adaptive fusion with a debiased objective improves overall accuracy and reliability on underrepresented classes, a step toward more clinically useful screening.<\/jats:p>","DOI":"10.3389\/frai.2026.1825132","type":"journal-article","created":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T04:50:26Z","timestamp":1779771026000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Adaptive Gating and Focal Debiasing for robust few-shot retinal disease classification"],"prefix":"10.3389","volume":"9","author":[{"given":"Stewart","family":"Muchuchuti","sequence":"first","affiliation":[{"name":"Discipline of Computer Science, School of Agriculture and Science, University of KwaZulu-Natal","place":["Durban, South Africa"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Serestina","family":"Viriri","sequence":"additional","affiliation":[{"name":"Discipline of Computer Science, School of Agriculture and Science, University of KwaZulu-Natal","place":["Durban, South Africa"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,5,26]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"e200267","DOI":"10.1148\/ryai.2021200267","article-title":"Assessing the trustworthiness of saliency maps for localizing abnormalities in medical imaging","volume":"3","author":"Arun","year":"2021","journal-title":"Radiol. Artif. Intell."},{"key":"B2","doi-asserted-by":"publisher","first-page":"14493","DOI":"10.1109\/CVPR42600.2020.01450","article-title":"\u201cImproved few-shot visual classification,\u201d","author":"Bateni","year":"2020","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition (CVPR)"},{"key":"B3","doi-asserted-by":"publisher","first-page":"1070","DOI":"10.1001\/jamaophthalmol.2020.3269","article-title":"Low-shot deep learning of diabetic retinopathy with potential applications to address artificial intelligence bias in retinal diagnostics and rare ophthalmic diseases","volume":"138","author":"Burlina","year":"2020","journal-title":"JAMA Ophthalmol"},{"key":"B4","first-page":"1567","article-title":"Learning imbalanced datasets with label-distribution-aware margin loss","volume":"32","author":"Cao","year":"2019","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B5","doi-asserted-by":"publisher","first-page":"321","DOI":"10.1613\/jair.953","article-title":"SMOTE: synthetic minority over-sampling technique","volume":"16","author":"Chawla","year":"2002","journal-title":"J. Artif. Intell. Res"},{"key":"B6","doi-asserted-by":"publisher","first-page":"9268","DOI":"10.1109\/CVPR.2019.00949","article-title":"\u201cClass-balanced loss based on effective number of samples,\u201d","author":"Cui","year":"2019","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition"},{"key":"B7","doi-asserted-by":"publisher","first-page":"231","DOI":"10.5566\/ias.1155","article-title":"Feedback on a publicly distributed image database: the messidor database","volume":"33","author":"Decenciere","year":"2014","journal-title":"Image Anal. Stereol"},{"key":"B8","doi-asserted-by":"crossref","DOI":"10.1109\/DSInS64146.2024.10992169","article-title":"\u201cEnhancing few-shot learning with integrated data and GAN model approaches,\u201d","volume-title":"2024 4th International Conference on Digital Society and Intelligent Systems (DSInS)","author":"Feng","year":"2024"},{"key":"B9","first-page":"1126","article-title":"\u201cModel-agnostic meta-learning for fast adaptation of deep networks,\u201d","volume-title":"Proceedings of the 34th 6 international conference on machine learning, volume 70 of proceedings of machine learning research (PMLR)","author":"Finn","year":"2017"},{"key":"B10","doi-asserted-by":"publisher","first-page":"33543","DOI":"10.1038\/s41598-025-18329-8","article-title":"Multi-stage fusion of local and global features for few-shot image classification","volume":"15","author":"Gu","year":"2025","journal-title":"Sci. Rep"},{"key":"B11","first-page":"1321","article-title":"\u201cOn calibration of modern neural networks,\u201d","volume-title":"Proceedings of the 34th international conference on machine learning (ICML), volume 70 of proceedings of machine learning research (PMLR)","author":"Guo","year":"2017"},{"key":"B12","doi-asserted-by":"publisher","first-page":"12175","DOI":"10.1109\/CVPR52688.2022.01186","article-title":"\u201cCMT: convolutional neural networks meet vision transformers,\u201d","author":"Guo","year":"2022","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition"},{"key":"B13","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TIM.2021.3122172","article-title":"Incremental cross-domain adaptation for robust retinopathy screening via bayesian deep learning","volume":"70","author":"Hassan","year":"2021","journal-title":"IEEE Trans. Instrum. Meas"},{"key":"B14","doi-asserted-by":"publisher","first-page":"770","DOI":"10.1109\/CVPR.2016.90","article-title":"\u201cDeep residual learning for image recognition,\u201d","author":"He","year":"2016","journal-title":"Proceedings of the IEEE conference on computer vision and pattern recognition"},{"key":"B15","doi-asserted-by":"publisher","first-page":"7132","DOI":"10.1109\/CVPR.2018.00745","article-title":"\u201cSqueeze-and-excitation networks,\u201d","author":"Hu","year":"2018","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)"},{"key":"B16","doi-asserted-by":"publisher","first-page":"79","DOI":"10.1162\/neco.1991.3.1.79","article-title":"Adaptive mixtures of local experts","volume":"3","author":"Jacobs","year":"1991","journal-title":"Neural Comput"},{"key":"B17","doi-asserted-by":"publisher","first-page":"123885","DOI":"10.1016\/j.eswa.2024.123885","article-title":"Hmrm: hierarchy-aware misclassification risk minimization for few-shot learning","volume":"251","author":"Jin","year":"2024","journal-title":"Expert Syst. Appl"},{"key":"B18","doi-asserted-by":"publisher","first-page":"3248","DOI":"10.1007\/s10278-024-01322-4","article-title":"Systematic review of hybrid vision transformer architectures for radiological image analysis","volume":"38","author":"Kim","year":"2025","journal-title":"J. Imag. Inform. Med"},{"key":"B19","doi-asserted-by":"publisher","first-page":"6405","DOI":"10.5555\/3295222.3295387","article-title":"Simple and scalable predictive uncertainty estimation using deep ensembles","volume":"30","author":"Lakshminarayanan","year":"2017","journal-title":"Adv. Neural Inform. Process. Syst."},{"key":"B20","doi-asserted-by":"publisher","first-page":"10657","DOI":"10.1109\/CVPR.2019.01091","article-title":"\u201cMeta-learning with differentiable convex optimization,\u201d","author":"Lee","year":"2019","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition"},{"key":"B21","doi-asserted-by":"publisher","first-page":"177","DOI":"10.1007\/978-3-030-71058-3_11","article-title":"\u201cA benchmark of ocular disease intelligent recognition: one shot for multi-disease detection,\u201d","author":"Li","year":"2021","journal-title":"Benchmarking, Measuring, and Optimizing. Bench 2020, volume 12614 of Lecture Notes in Computer Science"},{"key":"B22","doi-asserted-by":"publisher","first-page":"1489","DOI":"10.1016\/j.ins.2019.06.011","article-title":"\u201cDiagnostic assessment of deep learning algorithms for diabetic retinopathy screening,\u201d","author":"Li","year":"2019","journal-title":"2019 IEEE international conference on bioinformatics and biomedicine (BIBM)"},{"key":"B23","doi-asserted-by":"publisher","first-page":"492","DOI":"10.1007\/978-981-95-7075-1_32","article-title":"\u201cGeneralised few-shot node classification via training set refinement,\u201d","author":"Li","year":"2026","journal-title":"PRICAI 2025: trends in artificial intelligence"},{"key":"B24","doi-asserted-by":"publisher","first-page":"2980","DOI":"10.1109\/ICCV.2017.324","article-title":"\u201cFocal loss for dense object detection,\u201d","author":"Lin","year":"2017","journal-title":"Proceedings of the IEEE international conference on computer vision"},{"key":"B25","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1016\/j.media.2017.07.005","article-title":"A survey on deep learning in medical image analysis","volume":"42","author":"Litjens","year":"2017","journal-title":"Med. Image Anal."},{"key":"B26","first-page":"21464","article-title":"Energy-based out-of-distribution detection","volume":"33","author":"Liu","year":"2020","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B27","doi-asserted-by":"publisher","first-page":"84","DOI":"10.3390\/jimaging9040084","article-title":"Retinal disease detection using deep learning techniques: a comprehensive review","volume":"9","author":"Muchuchuti","year":"2023","journal-title":"J. Imaging"},{"key":"B28","doi-asserted-by":"publisher","first-page":"e11474472","DOI":"10.1109\/ACCESS.2026.3680450","article-title":"Hybrid few-shot learning with attention for multi-disease and out-of-distribution detection in retinal imaging","volume":"14","author":"Muchuchuti","year":"2026","journal-title":"IEEE Access"},{"key":"B29","article-title":"\u201cObtaining well calibrated probabilities using Bayesian binning,\u201d","volume-title":"Proceedings of the AAAI conference on artificial intelligence","author":"Naeini","year":"2015"},{"key":"B30","first-page":"23751","article-title":"Balanced meta-softmax for long-tailed visual recognition","volume":"33","author":"Ren","year":"2020","journal-title":"Adv. Neural Inform. Process. Syst"},{"key":"B31","doi-asserted-by":"publisher","first-page":"234","DOI":"10.1007\/978-3-319-24574-4_28","article-title":"\u201cU-Net: convolutional networks for biomedical image segmentation,\u201d","author":"Ronneberger","year":"2015","journal-title":"Medical image computing and computer-assisted intervention (MICCAI)"},{"key":"B32","doi-asserted-by":"publisher","first-page":"109240","DOI":"10.1016\/j.patcog.2022.109240","article-title":"A proxy learning curve for the bayes classifier","volume":"132","author":"Salazar","year":"2022","journal-title":"Pattern Recognit"},{"key":"B33","doi-asserted-by":"publisher","first-page":"113819","DOI":"10.1016\/j.eswa.2020.113819","article-title":"Generative adversarial networks and markov random fields for oversampling very small training sets","volume":"163","author":"Salazar","year":"2021","journal-title":"Expert Syst. Appl"},{"key":"B34","doi-asserted-by":"publisher","first-page":"197","DOI":"10.1016\/j.media.2019.01.012","article-title":"Attention gated networks: learning to leverage salient regions in medical images","volume":"53","author":"Schlemper","year":"2019","journal-title":"Med. Image Anal"},{"key":"B35","doi-asserted-by":"publisher","first-page":"618","DOI":"10.1109\/ICCV.2017.74","article-title":"\u201cGrad-CAM: visual explanations from deep networks via gradient-based localization,\u201d","author":"Selvaraju","year":"2017","journal-title":"Proceedings of the IEEE international conference on computer vision"},{"key":"B36","article-title":"\u201cOutrageously large neural networks: the sparsely-gated mixture-of-experts layer,\u201d","volume-title":"Proceedings of the 5th international conference on learning representations (ICLR)","author":"Shazeer","year":"2017"},{"key":"B37","article-title":"\u201cPrototypical networks for few-shot learning,\u201d","volume-title":"Advances in neural information processing systems, Vol. 30","author":"Snell","year":"2017"},{"key":"B38","doi-asserted-by":"publisher","first-page":"1199","DOI":"10.1109\/CVPR.2018.00131","article-title":"\u201cLearning to compare: relation network for few-shot learning,\u201d","author":"Sung","year":"2018","journal-title":"Proceedings of the IEEE conference on computer vision and pattern recognition"},{"key":"B39","doi-asserted-by":"publisher","first-page":"11664","DOI":"10.1109\/CVPR46437.2021.00173","article-title":"\u201cEqualization loss v2: a new gradient balance approach for long-tailed object recognition,\u201d","author":"Tan","year":"2021","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition"},{"key":"B40","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3386252","article-title":"Generalizing from a few examples: a survey on few-shot learning","volume":"53","author":"Wang","year":"2020","journal-title":"ACM Comput. Surv"},{"key":"B41","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1007\/978-3-030-01234-2_1","article-title":"\u201cCBAM: convolutional block attention module,\u201d","author":"Woo","year":"2018","journal-title":"Proceedings of the European conference on computer vision (ECCV)"},{"key":"B42","doi-asserted-by":"publisher","first-page":"100033","DOI":"10.59717\/j.xinn-inform.2026.100033","article-title":"Informative missingness and its implications in semi-supervised learning","volume":"2","author":"Wu","year":"2026","journal-title":"Innov. Inform."},{"key":"B43","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1007\/978-3-030-01261-8_1","article-title":"\u201cGroup normalization,\u201d","author":"Wu","year":"2018","journal-title":"Proceedings of the European Conference on Computer Vision (ECCV)"},{"key":"B44","doi-asserted-by":"publisher","first-page":"10819","DOI":"10.1109\/CVPR52688.2022.01055","article-title":"\u201cMetaformer is actually what you need for vision,\u201d","author":"Yu","year":"2022","journal-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition (CVPR)"},{"key":"B45","doi-asserted-by":"publisher","first-page":"6023","DOI":"10.1109\/ICCV.2019.00612","article-title":"\u201cCutmix: regularization strategy to train strong classifiers with localizable features,\u201d","author":"Yun","year":"2019","journal-title":"Proceedings of the IEEE\/CVF international conference on computer vision"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1825132\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T04:50:52Z","timestamp":1779771052000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1825132\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,26]]},"references-count":45,"alternative-id":["10.3389\/frai.2026.1825132"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1825132","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,26]]},"article-number":"1825132"}}