{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,5]],"date-time":"2026-05-05T04:10:00Z","timestamp":1777954200569,"version":"3.51.4"},"reference-count":16,"publisher":"Public Library of Science (PLoS)","issue":"5","license":[{"start":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T00:00:00Z","timestamp":1777852800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100004731","name":"Natural Science Foundation of Zhejiang Province","doi-asserted-by":"publisher","award":["Y1110023"],"award-info":[{"award-number":["Y1110023"]}],"id":[{"id":"10.13039\/501100004731","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>We propose ADGNET, a semi-supervised framework for Alzheimer\u2019s disease (AD) diagnosis that jointly optimizes image reconstruction and classification through shared feature representations. The architecture integrates a residual backbone with attention modulation for dynamic feature selection, an encoder-decoder reconstruction branch for unsupervised representation learning, and a classification branch with focal loss to address class imbalance. This dual-task design enables effective feature learning from limited annotations. On two public MRI datasets\u2014KACD (2D, 6,400 images) and ROAD (3D, 532 scans)\u2014ADGNET achieves average performance improvements of 4.1% and 7.2% over state-of-the-art methods (ResNeXt WSL, SimCLR) across six metrics. Interpretability analysis using Grad-CAM and attention visualization confirms that the model focuses on clinically relevant neuroanatomical structures, particularly the hippocampus and temporal lobes, with strong correlation to established AD pathology (r\u2009=\u20090.67, p\u2009&lt;\u20090.001). These results validate the model\u2019s exceptional generalization capability and feature representation effectiveness across multi-modal medical imaging data, offering an efficient solution for few-shot medical image analysis.<\/jats:p>","DOI":"10.1371\/journal.pone.0348596","type":"journal-article","created":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T17:32:00Z","timestamp":1777915920000},"page":"e0348596","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["Reducing annotation burden in medical imaging with ADGNET: A semi-supervised deep learning strategy"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6164-4131","authenticated-orcid":true,"given":"Xiaobo","family":"Yang","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"340","published-online":{"date-parts":[[2026,5,4]]},"reference":[{"issue":"1","key":"pone.0348596.ref001","doi-asserted-by":"crossref","first-page":"29","DOI":"10.1016\/j.nic.2009.08.016","article-title":"Brain magnetic resonance imaging techniques in the diagnosis of parkinsonian syndromes","volume":"20","author":"K Seppi","year":"2010","journal-title":"Neuroimaging Clin N Am"},{"issue":"7","key":"pone.0348596.ref002","doi-asserted-by":"crossref","first-page":"2146","DOI":"10.3390\/jcm9072146","article-title":"A Comprehensive Machine-Learning Model Applied to Magnetic Resonance Imaging (MRI) to Predict Alzheimer\u2019s Disease (AD) in Older Subjects","volume":"9","author":"G Battineni","year":"2020","journal-title":"J Clin Med"},{"issue":"2","key":"pone.0348596.ref003","doi-asserted-by":"crossref","first-page":"373","DOI":"10.1007\/s10994-019-05855-6","article-title":"A survey on semi-supervised learning","volume":"109","author":"JE van Engelen","year":"2019","journal-title":"Mach Learn"},{"key":"pone.0348596.ref004","first-page":"181","volume-title":"Exploring the limits of weakly supervised pretraining\/\/Proceedings of the European Conference on Computer Vision, September 8-14, 2018, Munich, Germany","author":"D Mahajan","year":"2018"},{"key":"pone.0348596.ref005","unstructured":"Chen T, Kornblith S, Hinton G, et al. A simple framework for contrastive learning of visual representations. Proceedings of the 37th International Conference on Machine Learning, 2020. 1597\u2013607."},{"issue":"3","key":"pone.0348596.ref006","first-page":"735","article-title":"Multimodal image registration with deep context reinforcement learning","volume":"42","author":"X Hu 6","year":"2023","journal-title":"IEEE Transactions on Medical Imaging"},{"key":"pone.0348596.ref007","first-page":"102895","article-title":"DTC: A dual-task consistency framework for semi-supervised medical image segmentation","volume":"89","author":"L Wang","year":"2023","journal-title":"Medical Image Analysis"},{"key":"pone.0348596.ref008","unstructured":"Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need. Advances in Neural Information Processing Systems, 2017. 5998\u20136008."},{"key":"pone.0348596.ref009","unstructured":"Klein PN, Can SG, Berahas AS. An empirical evaluation of activation functions in deep learning. Proceedings of the 2021 SIAM International Conference on Data Mining (SDM 2021), 2021. 288\u201396."},{"key":"pone.0348596.ref010","doi-asserted-by":"crossref","first-page":"14","DOI":"10.1016\/j.neunet.2021.01.026","article-title":"A survey on modern trainable activation functions","volume":"138","author":"A Apicella","year":"2021","journal-title":"Neural Netw"},{"key":"pone.0348596.ref011","unstructured":"ROAD: A Challenge on Alzheimer\u2019s Disease Diagnosis. China Computer Federation (CCF), 2018. https:\/\/github.com\/ChinaCVD\/ADRI-2018"},{"key":"pone.0348596.ref012","unstructured":"Kaggle. Alzheimer\u2019s dataset (4 class of images). https:\/\/www.kaggle.com\/datasets\/tourist55\/alzheimers-dataset-4-class-of-images. 2020."},{"issue":"12","key":"pone.0348596.ref013","first-page":"9454","article-title":"Exploring the limits of weakly supervised Pretraining","volume":"44","author":"D Mahajan","year":"2022","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"pone.0348596.ref014","unstructured":"Chen T, Kornblith S, Swersky K, et al. Big Self-Supervised Models are Strong Semi-Supervised Learners. Advances in Neural Information Processing Systems, 2020. 22243\u201355."},{"key":"pone.0348596.ref015","first-page":"185","article-title":"Exploring the limits of weakly supervised pretraining","volume-title":"Lecture Notes in Computer Science","author":"D Mahajan","year":"2018"},{"key":"pone.0348596.ref016","unstructured":"Zbontar J, Jing L, Misra I, et al. Barlow Twins: Self-Supervised Learning via Redundancy Reduction. Proceedings of the 38th International Conference on Machine Learning (ICML 2021), 2021. 12310\u201320."}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0348596","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T17:32:07Z","timestamp":1777915927000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0348596"}},"subtitle":[],"editor":[{"given":"Vince","family":"Grolmusz","sequence":"first","affiliation":[],"role":[{"role":"editor","vocabulary":"crossref"}]}],"short-title":[],"issued":{"date-parts":[[2026,5,4]]},"references-count":16,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2026,5,4]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0348596","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,4]]}}}