{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T18:16:47Z","timestamp":1788286607784,"version":"build-2803163510"},"reference-count":57,"publisher":"Public Library of Science (PLoS)","issue":"9","license":[{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100021124","name":"Tr\u01b0\u1eddng \u0110\u1ea1i h\u1ecdc C\u00f4ng nghi\u1ec7p th\u00e0nh ph\u1ed1 H\u1ed3 Ch\u00ed Minh","doi-asserted-by":"publisher","award":["This research is supported by Industrial University of Ho Chi Minh City (IUH) under grant number 104\/H\u00d0-\u00d0HCN"],"award-info":[{"award-number":["This research is supported by Industrial University of Ho Chi Minh City (IUH) under grant number 104\/H\u00d0-\u00d0HCN"]}],"id":[{"id":"10.13039\/100021124","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:sec id=\"sec001\">\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Artificial Intelligence (AI) models for mammography classification is prone to shortcut learning because diagnostically relevant evidence is typically sparse, localized, and easily dominated by non-lesion background context. This study aimed to develop a mammography-specific framework that integrates lesion-focused evidence with global image representations to improve classification performance and provide more clinically interpretable decision support.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec002\">\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We propose LENS, a hybrid CNN-Transformer family designed specifically for mammography. LENS combines a lightweight multi-scale convolutional neural network (CNN) backbone with an alternating local-global Transformer encoder. In addition to global image representations, a weakly supervised lesion-aware branch identifies and aggregates suspicious regional evidence to support image-level prediction. LENS was evaluated on the large-scale VinDrMammo dataset for the three-class mammography task and compared to advanced architectures, including ConvNeXt, DINOv2, GMIC, and Swin Transformer under a unified experimental protocol.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec003\">\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>LENS-Base achieved the highest overall Accuracy of 85.5%, Macro F1-score of 79.6%, and Matthews correlation coefficient of 0.65. Quantitative localization evaluation results indicated that the selected regional evidence frequently overlapped with annotated abnormalities. Furthermore, these promising results come with fewer parameters and lower FLOPs.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec id=\"sec004\">\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>LENS integrates local lesion-related features with global anatomical context to achieve improved class-balanced mammography classification while providing quantitatively supported lesion-focused evidence. These findings suggest that LENS is a promising framework for AI-assisted mammography screening.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1371\/journal.pone.0350720","type":"journal-article","created":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T17:49:03Z","timestamp":1788284943000},"page":"e0350720","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["LENS: A mammography-specific hybrid CNN-Transformer with lesion-aware evidence modeling"],"prefix":"10.1371","volume":"21","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5666-5728","authenticated-orcid":true,"given":"Duc Quy","family":"Hoang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Van Kien","family":"Cao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tan Nhu","family":"Nguyen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1414-6831","authenticated-orcid":true,"given":"Ngoc Son","family":"Nguyen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,9,1]]},"reference":[{"issue":"4","key":"pone.0350720.ref001","doi-asserted-by":"crossref","first-page":"1154","DOI":"10.1038\/s41591-025-03502-3","article-title":"Global patterns and trends in breast cancer incidence and mortality across 185 countries","volume":"31","author":"J Kim","year":"2025","journal-title":"Nat Med"},{"issue":"3","key":"pone.0350720.ref002","doi-asserted-by":"crossref","first-page":"233","DOI":"10.1001\/jama.2023.25881","article-title":"Analysis of breast cancer mortality in the US\u20141975 to 2019","volume":"331","author":"JL Caswell-Jin","year":"2024","journal-title":"JAMA"},{"issue":"22","key":"pone.0350720.ref003","doi-asserted-by":"crossref","first-page":"1918","DOI":"10.1001\/jama.2024.5534","article-title":"Screening for breast cancer: US preventive services task force recommendation statement","volume":"331","author":"Force US Preventive Services Task","year":"2024","journal-title":"JAMA"},{"issue":"1","key":"pone.0350720.ref004","doi-asserted-by":"crossref","DOI":"10.1001\/jamanetworkopen.2024.52821","article-title":"Cost-effectiveness of breast cancer screening using digital mammography in Canada","volume":"8","author":"AN Wilkinson","year":"2025","journal-title":"JAMA Netw Open"},{"issue":"2","key":"pone.0350720.ref005","doi-asserted-by":"crossref","first-page":"218","DOI":"10.3348\/kjr.2018.0193","article-title":"Interpretive performance and inter-observer agreement on digital mammography test sets","volume":"20","author":"SH Kim","year":"2019","journal-title":"Korean J Radiol"},{"issue":"3","key":"pone.0350720.ref006","doi-asserted-by":"crossref","first-page":"332","DOI":"10.1093\/jbi\/wbab012","article-title":"What causes the most stress in breast radiology practice? A survey of members of the society of breast imaging","volume":"3","author":"JR Parikh","year":"2021","journal-title":"J Breast Imaging"},{"issue":"7788","key":"pone.0350720.ref007","doi-asserted-by":"crossref","first-page":"89","DOI":"10.1038\/s41586-019-1799-6","article-title":"International evaluation of an AI system for breast cancer screening","volume":"577","author":"SM McKinney","year":"2020","journal-title":"Nature"},{"issue":"9","key":"pone.0350720.ref008","article-title":"Effect of artificial intelligence-based triaging of breast cancer screening mammograms on cancer detection and radiologist workload: a retrospective simulation study","volume":"2","author":"K Dembrower","year":"2020","journal-title":"Lancet Digit Health"},{"issue":"3","key":"pone.0350720.ref009","doi-asserted-by":"crossref","DOI":"10.1148\/radiol.232479","article-title":"Early indicators of the impact of using ai in mammography screening for breast cancer","volume":"311","author":"AD Lauritzen","year":"2024","journal-title":"Radiology"},{"issue":"5","key":"pone.0350720.ref010","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0322925","article-title":"Comparative analysis of diagnostic performance in mammography: a reader study on the impact of AI assistance","volume":"20","author":"MT Ramli Hamid","year":"2025","journal-title":"PLoS One"},{"issue":"3","key":"pone.0350720.ref011","article-title":"AI for mammography: making double screen-reading history","volume":"7","author":"N Houssami","year":"2025","journal-title":"Lancet Digit Health"},{"issue":"3","key":"pone.0350720.ref012","doi-asserted-by":"crossref","first-page":"917","DOI":"10.1038\/s41591-024-03408-6","article-title":"Nationwide real-world implementation of AI for cancer detection in population-based mammography screening","volume":"31","author":"N Eisemann","year":"2025","journal-title":"Nat Med"},{"issue":"2","key":"pone.0350720.ref013","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pdig.0001231","article-title":"A novel statistical framework for quantifying risks and benefits of AI automation in screening mammography","volume":"5","author":"MH Bernstein","year":"2026","journal-title":"PLOS Digit Health"},{"issue":"3","key":"pone.0350720.ref014","doi-asserted-by":"crossref","first-page":"189","DOI":"10.1049\/iet-cvi.2016.0244","article-title":"Mammographic mass classification according to Bi\u2010RADS lexicon","volume":"11","author":"F Chokri","year":"2016","journal-title":"IET Computer Vision"},{"key":"pone.0350720.ref015","doi-asserted-by":"crossref","unstructured":"Zhu W, Lou Q, Vang YS, Xie X. Deep multi-instance networks with sparse label assignment for whole mammogram classification. In: Medical image computing and computer assisted intervention - MICCAI 2017, 2017. 603\u201311.","DOI":"10.1007\/978-3-319-66179-7_69"},{"issue":"1","key":"pone.0350720.ref016","doi-asserted-by":"crossref","first-page":"12495","DOI":"10.1038\/s41598-019-48995-4","article-title":"Deep learning to improve breast cancer detection on screening mammography","volume":"9","author":"L Shen","year":"2019","journal-title":"Sci Rep"},{"key":"pone.0350720.ref017","doi-asserted-by":"crossref","first-page":"168","DOI":"10.1016\/j.neucom.2019.01.112","article-title":"Multi-label transfer learning for the early diagnosis of breast cancer","volume":"392","author":"H Chougrad","year":"2020","journal-title":"Neurocomputing"},{"key":"pone.0350720.ref018","doi-asserted-by":"crossref","unstructured":"Duc Quy H, Van Kien C, Huy Anh HP, Son NN. Multi-view digital mammography mass classification: a convolutional neural network model approach. In: 2021 International symposium on electrical and electronics engineering (ISEE), 2021. 133\u20138.","DOI":"10.1109\/ISEE51682.2021.9418797"},{"issue":"4","key":"pone.0350720.ref019","doi-asserted-by":"crossref","first-page":"2431","DOI":"10.1007\/s11831-023-10052-9","article-title":"Breast mammograms diagnosis using deep learning: state of art tutorial review","volume":"31","author":"OB Naeem","year":"2024","journal-title":"Arch Computat Methods Eng"},{"key":"pone.0350720.ref020","doi-asserted-by":"crossref","first-page":"113372","DOI":"10.1016\/j.knosys.2025.113372","article-title":"An interpretable CNN-based model for mass classification in mammography","volume":"316","author":"G Li","year":"2025","journal-title":"Knowledge-Based Systems"},{"key":"pone.0350720.ref021","unstructured":"Dosovitskiy A, Beyer L, Kolesnikov A, Weissenborn D, Zhai X, Unterthiner T. An image is worth 16x16 words: transformers for image recognition at scale. In: International Conference on Learning Representations. 2021. https:\/\/openreview.net\/forum?id=YicbFdNTTy"},{"key":"pone.0350720.ref022","doi-asserted-by":"crossref","first-page":"103399","DOI":"10.1016\/j.media.2024.103399","article-title":"Multi-scale region selection network in deep features for full-field mammogram classification","volume":"100","author":"L Sun","year":"2025","journal-title":"Med Image Anal"},{"key":"pone.0350720.ref023","doi-asserted-by":"crossref","unstructured":"Liu Z, Lin Y, Cao Y, Hu H, Wei Y, Zhang Z, et al. Swin transformer: hierarchical vision transformer using shifted windows. In: 2021 IEEE\/CVF International Conference on Computer Vision (ICCV), 2021. 9992\u201310002.","DOI":"10.1109\/ICCV48922.2021.00986"},{"key":"pone.0350720.ref024","doi-asserted-by":"crossref","unstructured":"Liu Z, Hu H, Lin Y, Yao Z, Xie Z, Wei Y, et al. Swin transformer V2: scaling up capacity and resolution. In: 2022 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022. 11999\u20132009.","DOI":"10.1109\/CVPR52688.2022.01170"},{"key":"pone.0350720.ref025","unstructured":"Chu X, Tian Z, Wang Y, Zhang B, Ren H, Wei X. Twins: revisiting the design of spatial attention in vision transformers. In: Advances in neural information processing systems. 2021. 9355\u201366."},{"key":"pone.0350720.ref026","unstructured":"Chu X, Tian Z, Zhang B, Wang X, Shen C. Conditional positional encodings for vision transformers. In: International conference on learning representations (ICLR). 2023."},{"issue":"1","key":"pone.0350720.ref027","doi-asserted-by":"crossref","first-page":"87","DOI":"10.1109\/TPAMI.2022.3152247","article-title":"A survey on vision transformer","volume":"45","author":"K Han","year":"2023","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"pone.0350720.ref028","doi-asserted-by":"crossref","first-page":"95496","DOI":"10.1109\/ACCESS.2025.3571735","article-title":"Transformers for vision: a survey on innovative methods for computer vision","volume":"13","author":"B Palanisamy","year":"2025","journal-title":"IEEE Access"},{"key":"pone.0350720.ref029","doi-asserted-by":"crossref","first-page":"110057","DOI":"10.1016\/j.engappai.2025.110057","article-title":"Exploring the synergies of hybrid convolutional neural network and vision transformer architectures for computer vision: a survey","volume":"144","author":"Y Haruna","year":"2025","journal-title":"Eng Appl Artificial Intelligence"},{"issue":"5","key":"pone.0350720.ref030","doi-asserted-by":"crossref","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":"JW Kim","year":"2025","journal-title":"J Imaging Inform Med"},{"issue":"11","key":"pone.0350720.ref031","doi-asserted-by":"crossref","first-page":"2971","DOI":"10.3390\/biomedicines10112971","article-title":"A hybrid workflow of residual convolutional transformer encoder for breast cancer classification using digital X-ray mammograms","volume":"10","author":"RM Al-Tam","year":"2022","journal-title":"Biomedicines"},{"issue":"1","key":"pone.0350720.ref032","doi-asserted-by":"crossref","first-page":"89","DOI":"10.3390\/diagnostics13010089","article-title":"ETECADx: ensemble self-attention transformer encoder for breast cancer diagnosis using full-field digital X-ray breast images","volume":"13","author":"AM Al-Hejri","year":"2022","journal-title":"Diagnostics (Basel)"},{"key":"pone.0350720.ref033","doi-asserted-by":"crossref","first-page":"77187","DOI":"10.1109\/ACCESS.2025.3563218","article-title":"A robust hybrid CNN+ViT framework for breast cancer classification using mammogram images","volume":"13","author":"VR Patheda","year":"2025","journal-title":"IEEE Access"},{"issue":"11","key":"pone.0350720.ref034","doi-asserted-by":"crossref","first-page":"1361","DOI":"10.3390\/diagnostics15111361","article-title":"Multi-scale vision transformer with optimized feature fusion for mammographic breast cancer classification","volume":"15","author":"S Ahmed","year":"2025","journal-title":"Diagnostics (Basel)"},{"key":"pone.0350720.ref035","doi-asserted-by":"crossref","first-page":"108116","DOI":"10.1016\/j.bspc.2025.108116","article-title":"InceptionNeXt-transformer: a novel multi-scale deep feature learning architecture for multimodal breast cancer diagnosis","volume":"110","author":"I Pacal","year":"2025","journal-title":"Biomed Signal Process Control"},{"key":"pone.0350720.ref036","doi-asserted-by":"crossref","first-page":"105799","DOI":"10.1016\/j.chemolab.2026.105799","article-title":"DeformNeXt-Swin: a hybrid CNN-transformer framework for breast lesion classification in ultrasound and mammography","volume":"276","author":"I Pacal","year":"2026","journal-title":"Chemomet Intelligent Laboratory Systems"},{"issue":"11","key":"pone.0350720.ref037","doi-asserted-by":"crossref","first-page":"665","DOI":"10.1038\/s42256-020-00257-z","article-title":"Shortcut learning in deep neural networks","volume":"2","author":"R Geirhos","year":"2020","journal-title":"Nat Mach Intell"},{"key":"pone.0350720.ref038","doi-asserted-by":"crossref","first-page":"101908","DOI":"10.1016\/j.media.2020.101908","article-title":"An interpretable classifier for high-resolution breast cancer screening images utilizing weakly supervised localization","volume":"68","author":"Y Shen","year":"2021","journal-title":"Med Image Anal"},{"issue":"9","key":"pone.0350720.ref039","doi-asserted-by":"crossref","first-page":"842","DOI":"10.1016\/j.jacr.2023.06.025","article-title":"Shortcuts causing bias in radiology artificial intelligence: causes, evaluation, and mitigation","volume":"20","author":"I Banerjee","year":"2023","journal-title":"J American College of Radiol"},{"key":"pone.0350720.ref040","doi-asserted-by":"crossref","first-page":"320","DOI":"10.1016\/j.neucom.2021.02.070","article-title":"A multi-instance networks with multiple views for classification of mammograms","volume":"443","author":"T Hu","year":"2021","journal-title":"Neurocomputing"},{"key":"pone.0350720.ref041","first-page":"364","volume-title":"Multi-scale attention-based multiple instance learning for breast cancer diagnosis. In: Medical image computing and computer assisted intervention \u2013 MICCAI 2025","author":"M Mour\u00e3o","year":"2026"},{"issue":"1","key":"pone.0350720.ref042","doi-asserted-by":"crossref","first-page":"277","DOI":"10.1038\/s41597-023-02100-7","article-title":"VinDr-Mammo: a large-scale benchmark dataset for computer-aided diagnosis in full-field digital mammography","volume":"10","author":"HT Nguyen","year":"2023","journal-title":"Sci Data"},{"key":"pone.0350720.ref043","volume-title":"Breast imaging reporting and data system (BI-RADS\u00ae)","author":"American College of Radiology","year":"2013","edition":"5"},{"issue":"1","key":"pone.0350720.ref044","doi-asserted-by":"crossref","DOI":"10.1148\/rg.240169","article-title":"Understanding BI-RADS category 3","volume":"45","author":"S Fazeli","year":"2025","journal-title":"Radiographics"},{"key":"pone.0350720.ref045","doi-asserted-by":"crossref","first-page":"108973","DOI":"10.1016\/j.bspc.2025.108973","article-title":"Impact of magnification on deep learning approaches through comprehensive comparative study of histopathological breast cancer classification","volume":"113","author":"O Attallah","year":"2026","journal-title":"Biomedical Signal Processing and Control"},{"key":"pone.0350720.ref046","doi-asserted-by":"crossref","unstructured":"Nguyen T-H, Kha QH, Ngoc Toan Truong T, Lam BT, Hung Ngo B, Dinh QV, et al. Towards robust natural-looking mammography lesion synthesis on ipsilateral dual-views breast cancer analysis. In: 2023 IEEE\/CVF International Conference on Computer Vision Workshops (ICCVW), 2023. 2556\u201365.","DOI":"10.1109\/ICCVW60793.2023.00270"},{"key":"pone.0350720.ref047","doi-asserted-by":"crossref","first-page":"110509","DOI":"10.1016\/j.bspc.2026.110509","article-title":"MV-TRAMS: An efficient tumor region-adapted mammography synthesis under multi-view diagnosis","volume":"123","author":"T-H Nguyen","year":"2026","journal-title":"Biomedical Signal Processing and Control"},{"key":"pone.0350720.ref048","first-page":"325","article-title":"DeYOLOv3: an optimal mass detector for advanced breast cancer diagnostics","volume-title":"Lecture notes in networks and systems","author":"HD Quy","year":"2022"},{"key":"pone.0350720.ref049","unstructured":"Tan M, Le QV. EfficientNet: rethinking model scaling for convolutional neural networks. In: IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 2019. 6105-14."},{"key":"pone.0350720.ref050","doi-asserted-by":"crossref","unstructured":"Tan M, Pang R, Le QV. EfficientDet: scalable and efficient object detection. In: 2020 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020. 10778\u201387.","DOI":"10.1109\/CVPR42600.2020.01079"},{"key":"pone.0350720.ref051","doi-asserted-by":"crossref","unstructured":"Lin T-Y, Goyal P, Girshick R, He K, Dollar P. Focal loss for dense object detection. In: 2017 IEEE International Conference on Computer Vision (ICCV), 2017. 2999\u20133007.","DOI":"10.1109\/ICCV.2017.324"},{"key":"pone.0350720.ref052","unstructured":"Loshchilov I, Hutter F. Decoupled weight decay regularization. in: international conference on learning representations. 2019. https:\/\/openreview.net\/forum?id=Bkg6RiCqY7"},{"key":"pone.0350720.ref053","article-title":"Pytorch: an imperative style, high-performance deep learning library","volume":"32","author":"A Paszke","year":"2019","journal-title":"Adv Neural Information Process Syst"},{"issue":"1","key":"pone.0350720.ref054","doi-asserted-by":"crossref","first-page":"189","DOI":"10.1109\/TPAMI.2016.2535231","article-title":"Weakly supervised object localization with multi-fold multiple instance learning","volume":"39","author":"RG Cinbis","year":"2017","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"pone.0350720.ref055","doi-asserted-by":"crossref","unstructured":"Liu Z, Mao H, Wu C-Y, Feichtenhofer C, Darrell T, Xie S. A ConvNet for the 2020s. In: 2022 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022. 11966\u201376.","DOI":"10.1109\/CVPR52688.2022.01167"},{"key":"pone.0350720.ref056","unstructured":"Oquab M, Darcet T, Moutakanni T, Vo H, Szafraniec M, Khalidov V. Dinov2: Learning robust visual features without supervision. arXiv preprint. 2023."},{"key":"pone.0350720.ref057","doi-asserted-by":"crossref","first-page":"109370","DOI":"10.1016\/j.compeleceng.2024.109370","article-title":"A review of explainable artificial intelligence in healthcare","volume":"118","author":"Z Sadeghi","year":"2024","journal-title":"Computers and Electrical Engineering"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0350720","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T17:49:15Z","timestamp":1788284955000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0350720"}},"subtitle":[],"editor":[{"given":"Fahad Farhan","family":"Almutairi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,9,1]]},"references-count":57,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2026,9,1]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0350720","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,1]]}}}