{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,7,30]],"date-time":"2025-07-30T16:15:59Z","timestamp":1753892159327,"version":"3.41.2"},"reference-count":41,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,6,11]],"date-time":"2025-06-11T00:00:00Z","timestamp":1749600000000},"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. Physiol."],"abstract":"<jats:sec><jats:title>Introduction<\/jats:title><jats:p>Recognizing a tooth-marked tongue has important clinical diagnostic value in traditional Chinese medicine. Current deep learning methods for tooth mark detection require extensive manual labeling and tongue segmentation, which is labor-intensive. Therefore, we propose a weakly supervised multipleinstance active learning model for tooth-marked tongue recognition, aiming to eliminate preprocessing segmentation and reduce the annotation workload while maintaining diagnostic accuracy.<\/jats:p><\/jats:sec><jats:sec><jats:title>Method<\/jats:title><jats:p>We propose a one-stage method tongenerate tooth mark instances that eliminates the need for pre-segmentation of the tongue. To make full use of unlabeled data, we introduce a semisupervised learning paradigm to pseudo-label unlabeled tongue images with high model confidence in active learning and incorporate them into the training set to improve the training efficiency of the active learning model. In addition, we propose an instance-level hybrid query method considering the diversity of tooth marks.<\/jats:p><\/jats:sec><jats:sec><jats:title>Result<\/jats:title><jats:p>Experimental results on clinical tongue images verify the effectiveness of the proposed method, which achieves an accuracy of 93.88% for tooth-marked tongue recognition, outperforming the recently introduced weakly supervised approaches.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusion<\/jats:title><jats:p>The proposed method is effective with only a small amount of image-level annotation, and its performance is comparable to that of image-level annotation, instance-level annotation and pixel-level annotation, which require a large number of tooth markers. Our method significantly reduces the annotation cost of the binary classification task of traditional Chinese medicine tooth mark recognition.<\/jats:p><\/jats:sec>","DOI":"10.3389\/fphys.2025.1598850","type":"journal-article","created":{"date-parts":[[2025,6,11]],"date-time":"2025-06-11T04:10:20Z","timestamp":1749615020000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Weakly supervised multiple-instance active learning for tooth-marked tongue recognition"],"prefix":"10.3389","volume":"16","author":[{"given":"Feilin","family":"Deng","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Shangxuan","family":"Li","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zizhu","family":"Yang","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Wu","family":"Zhou","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2025,6,11]]},"reference":[{"key":"B1","article-title":"Support vector machines for multiple-instance learning","volume-title":"Advances in Neural Information Processing Systems","author":"Andrews","year":"2002"},{"key":"B2","doi-asserted-by":"publisher","first-page":"03671","DOI":"10.48550\/arXiv.1906.03671","article-title":"Deep batch active learning by diverse, uncertain gradient lower bounds","author":"Ash","year":"2019","journal-title":"CoRR abs\/1906"},{"key":"B3","doi-asserted-by":"publisher","first-page":"1277121","DOI":"10.3389\/fmicb.2023.1277121","article-title":"Oral_voting_transfer: classification of oral microorganisms\u2019 function proteins with voting transfer model","volume":"14","author":"Bao","year":"2024","journal-title":"Front. Microbiol."},{"key":"B4","first-page":"2846","article-title":"Weakly supervised deep detection networks","author":"Bilen","year":"2016"},{"key":"B5","doi-asserted-by":"publisher","first-page":"179","DOI":"10.2174\/0115748936330499240909082529","article-title":"Clpr_in_ml: cleft lip and palate reconstructed features with machine learning","volume":"20","author":"Chen","year":"2024","journal-title":"Curr. Bioinforma."},{"key":"B6","doi-asserted-by":"publisher","DOI":"10.1007\/978-981-10-2167-1","article-title":"Tongue image analysis","author":"David Zhang","year":"2017"},{"key":"B7","first-page":"4476","article-title":"Look closer to see better: recurrent attention convolutional neural network for fine-grained image recognition","author":"Fu","year":"2017"},{"key":"B8","first-page":"1440","volume-title":"Fast r-cnn","author":"Girshick","year":"2015"},{"key":"B9","doi-asserted-by":"publisher","first-page":"00028","DOI":"10.48550\/arXiv.1704.00028","article-title":"Improved training of wasserstein gans","author":"Gulrajani","year":"2017","journal-title":"CoRR abs\/1704"},{"key":"B10","first-page":"1321","article-title":"On calibration of modern neural networks","volume-title":"Proceedings of the 34th international conference on machine learning","author":"Guo","year":"2017"},{"key":"B11","first-page":"770","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"B12","article-title":"Automatic tongue feature extraction","volume-title":"2010 international computer symposium (ICS2010)","author":"Hsu","year":"2011"},{"article-title":"Simple and scalable predictive uncertainty estimation using deep ensembles","year":"2017","author":"Lakshminarayanan","key":"B13"},{"key":"B14","doi-asserted-by":"publisher","first-page":"380","DOI":"10.1109\/TCYB.2017.2772289","article-title":"Tooth-marked tongue recognition using multiple instance learning and cnn features","volume":"49","author":"Li","year":"2018","journal-title":"IEEE Trans. Cybern."},{"key":"B15","doi-asserted-by":"publisher","first-page":"418","DOI":"10.1016\/j.ins.2021.08.019","article-title":"Deep active learning for object detection","volume":"579","author":"Li","year":"2021","journal-title":"Inf. Sci."},{"key":"B16","doi-asserted-by":"publisher","first-page":"505063","DOI":"10.1155\/2012\/505063","article-title":"The study on the agreement between automatic tongue diagnosis system and traditional Chinese medicine practitioners","volume":"2012","author":"Lo","year":"2012","journal-title":"Evidence-Based Complementray Altern. Med."},{"key":"B17","first-page":"1","article-title":"Prtmtm: a priori regularization method for tooth-marked tongue classification","author":"Lu","year":"2023"},{"key":"B18","doi-asserted-by":"crossref","first-page":"779","DOI":"10.1109\/CVPR.2016.91","article-title":"You only look once: unified, real-time object detection","volume-title":"2016 IEEE conference on computer vision and pattern recognition (CVPR)","author":"Redmon","year":"2016"},{"key":"B19","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1804.02767","article-title":"Yolov3: an incremental improvement","author":"Redmon","year":"2018","journal-title":"CoRR abs\/1804.02767"},{"key":"B20","first-page":"1609","volume-title":"Multi-instance multi-label learning with application to scene classification","author":"Sch\u00f6lkopf","year":"2007"},{"key":"B21","first-page":"618","article-title":"Grad-cam: visual explanations from deep networks via gradient-based localization","author":"Selvaraju","year":"2017"},{"key":"B22","first-page":"968","article-title":"Recognition of teeth-marked tongue based on gradient of concave region","author":"Shao","year":"2014"},{"key":"B23","doi-asserted-by":"publisher","first-page":"00908","DOI":"10.48550\/arXiv.2106.00908","article-title":"Transmil: transformer based correlated multiple instance learning for whole slide image classication","author":"Shao","year":"2021","journal-title":"CoRR abs\/2106"},{"key":"B24","doi-asserted-by":"publisher","first-page":"45","DOI":"10.3390\/fi11020045","article-title":"Tooth-marked tongue recognition using gradient-weighted class activation maps","volume":"11","author":"Sun","year":"2019","journal-title":"Future Internet"},{"key":"B25","first-page":"1274","article-title":"Tooth-marked tongue recognition based on wavelet transform and feature fusion","author":"Tan","year":"2023"},{"key":"B26","first-page":"1879","article-title":"Segmentation as selective search for object recognition","author":"Van de Sande","year":"2011"},{"key":"B27","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-540-71050-9","volume-title":"Optimal transport: Old and new. grundlehren der mathema-tischen wissenschaften","author":"Villani","year":"2009"},{"key":"B28","doi-asserted-by":"publisher","first-page":"2591","DOI":"10.1109\/TCSVT.2016.2589879","article-title":"Cost-effective active learning for deep image classification","volume":"27","author":"Wang","year":"2017","journal-title":"IEEE Trans. Circuits Syst. Video Technol."},{"key":"B29","first-page":"1","article-title":"Active learning through two-stage clustering","author":"Wang","year":"2018"},{"key":"B30","doi-asserted-by":"publisher","first-page":"973","DOI":"10.1016\/j.csbj.2020.04.002","article-title":"Artificial intelligence in tongue diagnosis: using deep convolutional neural network for recognizing unhealthy tongue with tooth-mark","volume":"18","author":"Wang","year":"2020","journal-title":"Comput. Struct. Biotechnol. J."},{"key":"B31","doi-asserted-by":"crossref","first-page":"10852","DOI":"10.1109\/CVPR46437.2021.01071","article-title":"Crest: a class-rebalancing self-training framework for imbalanced semi-supervised learning","volume-title":"2021 IEEE\/CVF conference on computer vision and pattern recognition (CVPR)","author":"Wei","year":"2021"},{"key":"B32","doi-asserted-by":"publisher","DOI":"10.1002\/cpe.6262","article-title":"A weakly supervised tooth-mark and crack detection method in tongue image","volume":"33","author":"Weng","year":"2021","journal-title":"Concurrency Comput. Pract. Exp."},{"key":"B33","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1906.08113","article-title":"Wasserstein adversarial imitation learning","author":"Xiao","year":"2019","journal-title":"CoRR abs\/1906.08113"},{"key":"B34","doi-asserted-by":"crossref","first-page":"5326","DOI":"10.1109\/CVPR46437.2021.00529","article-title":"Multiple instance active learning for object detection","volume-title":"2021 IEEE\/CVF conference on computer vision and pattern recognition (CVPR)","author":"Yuan","year":"2021"},{"key":"B35","first-page":"1165","article-title":"Wasserstein distance guided adversarial imitation learning with reward shape exploration","author":"Zhang","year":"2020"},{"key":"B36","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR52688.2022.02001","article-title":"Boostmis: boosting medical image semi-supervised learning with adaptive pseudo labeling and informative active annotation","volume-title":"Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition","author":"Zhang","year":"2022"},{"key":"B37","doi-asserted-by":"publisher","first-page":"847267","DOI":"10.3389\/fphys.2022.847267","article-title":"Weakly supervised deep learning for tooth-marked tongue recognition","volume":"13","author":"Zhou","year":"2022","journal-title":"Front. Physiology"},{"key":"B38","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1093\/nsr\/nwx106","article-title":"A brief introduction to weakly supervised learning","volume":"5","author":"Zhou","year":"2017","journal-title":"Natl. Sci. Rev."},{"key":"B39","doi-asserted-by":"publisher","first-page":"06567","DOI":"10.48550\/arXiv.2012.06567","article-title":"A comprehensive study of deep video action recognition","author":"Zhu","year":"2020","journal-title":"CoRR abs\/2012"},{"key":"B40","doi-asserted-by":"crossref","first-page":"391","DOI":"10.1007\/978-3-319-10602-1_26","article-title":"Edge boxes: locating object proposals from edges","volume-title":"Computer vision \u2013 eccv 2014","author":"Zitnick","year":"2014"},{"key":"B41","doi-asserted-by":"crossref","first-page":"5981","DOI":"10.1109\/ICCV.2019.00608","article-title":"Confidence regularized self-training","volume-title":"2019 IEEE\/CVF international conference on computer vision (ICCV)","author":"Zou","year":"2019"}],"container-title":["Frontiers in Physiology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fphys.2025.1598850\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,11]],"date-time":"2025-06-11T04:10:25Z","timestamp":1749615025000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fphys.2025.1598850\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,6,11]]},"references-count":41,"alternative-id":["10.3389\/fphys.2025.1598850"],"URL":"https:\/\/doi.org\/10.3389\/fphys.2025.1598850","relation":{},"ISSN":["1664-042X"],"issn-type":[{"type":"electronic","value":"1664-042X"}],"subject":[],"published":{"date-parts":[[2025,6,11]]},"article-number":"1598850"}}