{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,27]],"date-time":"2026-10-27T15:57:17Z","timestamp":1793116637953,"version":"4.5.0"},"reference-count":44,"publisher":"Wiley","issue":"9","license":[{"start":{"date-parts":[[2022,7,30]],"date-time":"2022-07-30T00:00:00Z","timestamp":1659139200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"}],"content-domain":{"domain":["aapm.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Medical Physics"],"published-print":{"date-parts":[[2022,9]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec><jats:title>Purpose<\/jats:title><jats:p>Breast cancer is the most commonly occurring cancer worldwide. The ultrasound reflectivity imaging technique can be used to obtain breast ultrasound (BUS) images, which can be used to classify benign and malignant tumors. However, the classification is subjective and dependent on the experience and skill of operators and doctors. The automatic classification method can assist doctors and improve the objectivity, but current convolution neural network (CNN) is not good at learning global features and vision transformer (ViT) is not good at extraction local features. In this study, we proposed a visual geometry group attention ViT (VGGA\u2010ViT) network to overcome their disadvantages.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>In the proposed method, we used a CNN module to extract the local features and employed a ViT module to learn the global relationship among different regions and enhance the relevant local features. The CNN module was named the VGGA module. It was composed of a VGG backbone, a feature extraction fully connected layer, and a squeeze\u2010and\u2010excitation block. Both the VGG backbone and the ViT module were pretrained on the ImageNet dataset and retrained using BUS samples in this study. Two BUS datasets were employed for validation.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Cross\u2010validation was conducted on two BUS datasets. For the Dataset A, the proposed VGGA\u2010ViT network achieved high accuracy (88.711.55%), recall (90.731.57%), specificity (85.583.35%), precision (90.771.98%), <jats:italic>F<\/jats:italic>1 score (90.731.24%), and Matthews correlation coefficient (MCC) (76.343.29%), which were better than those of all compared previous networks in this study. The Dataset B was used as a separate test set, the test results showed that the VGGA\u2010ViT had highest accuracy (81.722.99%), recall (64.452.96%), specificity (90.283.51%), precision (77.087.21%), <jats:italic>F<\/jats:italic>1 score (70.114.25%), and MCC (57.646.88%).<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusions<\/jats:title><jats:p>In this study, we proposed the VGGA\u2010ViT for the BUS classification, which was good at learning both local and global features. The proposed network achieved higher accuracy than the compared previous methods.<\/jats:p><\/jats:sec>","DOI":"10.1002\/mp.15852","type":"journal-article","created":{"date-parts":[[2022,7,22]],"date-time":"2022-07-22T09:54:44Z","timestamp":1658483684000},"page":"5787-5798","update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":62,"title":["A VGG attention vision transformer network for benign and malignant classification of breast ultrasound images"],"prefix":"10.1002","volume":"49","author":[{"given":"Xiaolei","family":"Qu","sequence":"first","affiliation":[{"name":"School of Instrumentation and Optoelectronics Engineering Beihang University  Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hongyan","family":"Lu","sequence":"additional","affiliation":[{"name":"School of Instrumentation and Optoelectronics Engineering Beihang University  Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wenzhong","family":"Tang","sequence":"additional","affiliation":[{"name":"School of computer Science and Engineering Beihang University  Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shuai","family":"Wang","sequence":"additional","affiliation":[{"name":"Research Institute for Frontier Science Beihang University  Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dezhi","family":"Zheng","sequence":"additional","affiliation":[{"name":"Research Institute for Frontier Science Beihang University  Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yaxin","family":"Hou","sequence":"additional","affiliation":[{"name":"Department of Diagnostic Ultrasound Beijing Tongren Hospital Capital Medical University  Beijing China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jue","family":"Jiang","sequence":"additional","affiliation":[{"name":"Department of Medical Physics Memorial Sloan Kettering Cancer Center  New York New York USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2022,7,30]]},"reference":[{"key":"e_1_2_9_2_1","doi-asserted-by":"publisher","DOI":"10.3322\/caac.21660"},{"key":"e_1_2_9_3_1","doi-asserted-by":"publisher","DOI":"10.5306\/wjco.v2.i4.171"},{"key":"e_1_2_9_4_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2009.05.012"},{"key":"e_1_2_9_5_1","doi-asserted-by":"publisher","DOI":"10.1118\/1.1429239"},{"key":"e_1_2_9_6_1","doi-asserted-by":"publisher","DOI":"10.7863\/jum.2004.23.9.1201"},{"key":"e_1_2_9_7_1","doi-asserted-by":"publisher","DOI":"10.1016\/S0301-5629(02)00788-3"},{"key":"e_1_2_9_8_1","doi-asserted-by":"publisher","DOI":"10.1023\/A:1010676701382"},{"key":"e_1_2_9_9_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2005.07.006"},{"key":"e_1_2_9_10_1","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.2262011843"},{"key":"e_1_2_9_11_1","doi-asserted-by":"publisher","DOI":"10.1016\/S0301-5629(99)00156-8"},{"key":"e_1_2_9_12_1","doi-asserted-by":"publisher","DOI":"10.1109\/42.952727"},{"key":"e_1_2_9_13_1","doi-asserted-by":"publisher","DOI":"10.1118\/1.1485995"},{"key":"e_1_2_9_14_1","unstructured":"KrizhevskyA SutskeverI HintonGE.ImageNet classification with deep convolutional neural networks.Paper presented at: Advances in Neural Information Processing Systems. Vol25;2012:1106\u20101114."},{"key":"e_1_2_9_15_1","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-015-0816-y"},{"key":"e_1_2_9_16_1","unstructured":"SimonyanK ZissermanA.Very deep convolutional networks for large\u2010scale image recognition.International Conference on Learning Representations. Vol1409;2015:1556. arXiv"},{"key":"e_1_2_9_17_1","doi-asserted-by":"crossref","unstructured":"HeK ZhangX RenS SunJ.Deep residual learning for image recognition.Paper presented at: Computer Vision and Pattern Recognition.2015:770\u2010778https:\/\/doi.org\/10.1109\/CVPR.2016.90","DOI":"10.1109\/CVPR.2016.90"},{"key":"e_1_2_9_18_1","doi-asserted-by":"crossref","unstructured":"SzegedyC LiuW JiaY SermanetP RabinovichA.Going deeper with convolutions.Paper presented at: Computer Vision and Pattern Recognition.2015:1\u20109.https:\/\/doi.org\/10.1109\/CVPR.2015.7298594","DOI":"10.1109\/CVPR.2015.7298594"},{"key":"e_1_2_9_19_1","unstructured":"IoffeS SzegedyC.Batch normalization: accelerating deep network training by reducing internal covariate shift.International Conference on Machine Learning. Vol1502;2015:03167. 448\u2010456. arXiv"},{"key":"e_1_2_9_20_1","doi-asserted-by":"crossref","unstructured":"SzegedyC VanhouckeV IoffeS ShlensJ WojnaZ.Rethinking the inception architecture for computer vision.Paper presented at: Computer Vision and Pattern Recognition.2016:2818\u20102826.https:\/\/doi.org\/10.1109\/CVPR.2016.308","DOI":"10.1109\/CVPR.2016.308"},{"key":"e_1_2_9_21_1","doi-asserted-by":"crossref","unstructured":"SzegedyC IoffeS VanhouckeV AlemiA.Inception\u2010v4 inception\u2010ResNet and the impact of residual connections on learning.Paper presented at: Conference on Artificial Intelligence.2017:4278\u20104284.","DOI":"10.1609\/aaai.v31i1.11231"},{"key":"e_1_2_9_22_1","doi-asserted-by":"crossref","unstructured":"CholletF.Xception: deep learning with depthwise separable convolutions.Paper presented at: Computer Vision and Pattern Recognition.2017:1800\u20101807.https:\/\/doi.org\/10.1109\/CVPR.2017.195","DOI":"10.1109\/CVPR.2017.195"},{"key":"e_1_2_9_23_1","doi-asserted-by":"crossref","unstructured":"HuangG LiuZ LaurensV WeinbergerKQ.Densely connected convolutional networks.Paper presented at: Computer Vision and Pattern Recognition.2017:2261\u20102269.https:\/\/doi.org\/10.1109\/CVPR.2017.243","DOI":"10.1109\/CVPR.2017.243"},{"key":"e_1_2_9_24_1","article-title":"SqueezeNet: AlexNet\u2010level accuracy with 50x fewer parameters and <0.5MB model size","volume":"1602","author":"Iandola FN","year":"2016","journal-title":"CoRR"},{"key":"e_1_2_9_25_1","doi-asserted-by":"publisher","DOI":"10.3390\/diagnostics10121055"},{"key":"e_1_2_9_26_1","doi-asserted-by":"publisher","DOI":"10.1118\/1.4957255"},{"key":"e_1_2_9_27_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.12453"},{"key":"e_1_2_9_28_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.13361"},{"key":"e_1_2_9_29_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.14510"},{"key":"e_1_2_9_30_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.15341"},{"key":"e_1_2_9_31_1","unstructured":"BaiY MeiJ YuilleA XieC.Are transformers more robust than CNNs?Paper presented at: Advances in Neural Information Processing Systems.2021. 34 pre\u2010proceedings"},{"key":"e_1_2_9_32_1","unstructured":"DosovitskiyA BeyerL KolesnikovA et\u00a0al.An image is worth 16\u00d716 words: transformers for image recognition at scale.International Conference on Learning Representations.2021;2010:11929. arXiv"},{"key":"e_1_2_9_33_1","article-title":"Vision transformer for classification of breast ultrasound images","volume":"2110","author":"Gheflati Behnaz","year":"2021","journal-title":"CoRR"},{"key":"e_1_2_9_34_1","doi-asserted-by":"crossref","unstructured":"HuJ ShenL SunG.Squeeze\u2010and\u2010excitation networks.Paper presented at: Computer Vision and Pattern Recognition.2018:7132\u20107141.https:\/\/doi.org\/10.1109\/CVPR.2018.00745","DOI":"10.1109\/CVPR.2018.00745"},{"key":"e_1_2_9_35_1","doi-asserted-by":"crossref","unstructured":"WooS ParkJ LeeJY KweonIS.CBAM: convolutional block attention module.Paper presented at: European Conference on Computer Vision.2018:3\u201019.https:\/\/doi.org\/10.1007\/978\u20103\u2010030\u201001234\u20102_1","DOI":"10.1007\/978-3-030-01234-2_1"},{"key":"e_1_2_9_36_1","unstructured":"PaszkeA GrossS MassaF et\u00a0al.PyTorch: an imperative style high\u2010performance deep learning library.Paper presented at: Advances in Neural Information Processing Systems. Vol32;2019:8024\u20108035."},{"key":"e_1_2_9_37_1","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2019.2891622"},{"key":"e_1_2_9_38_1","doi-asserted-by":"publisher","DOI":"10.1002\/mp.14470"},{"key":"e_1_2_9_39_1","doi-asserted-by":"publisher","DOI":"10.1109\/JBHI.2017.2731873"},{"key":"e_1_2_9_40_1","article-title":"MobileNets: efficient convolutional neural networks for mobile vision applications","volume":"1704","author":"Howard AG","year":"2017","journal-title":"CoRR"},{"key":"e_1_2_9_41_1","doi-asserted-by":"crossref","unstructured":"SandlerM HowardA ZhuM ZhmoginovA ChenLC.MobileNetV2: inverted residuals and linear bottlenecks.Paper presented at: Computer Vision and Pattern Recognition.2018:4510\u20104520.https:\/\/doi.org\/10.1109\/CVPR.2018.00474","DOI":"10.1109\/CVPR.2018.00474"},{"key":"e_1_2_9_42_1","doi-asserted-by":"publisher","DOI":"10.1378\/chest.114.3.946-a"},{"key":"e_1_2_9_43_1","doi-asserted-by":"publisher","DOI":"10.1109\/LSP.2014.2337313"},{"key":"e_1_2_9_44_1","first-page":"2579","article-title":"Visualizing data using t\u2010SNE","volume":"9","author":"Laurens VDM","year":"2008","journal-title":"J Mach Learn Res"},{"key":"e_1_2_9_45_1","doi-asserted-by":"crossref","unstructured":"AbnarS ZuidemaW Assoc ComputatL.Quantifying attention flow in transformers.Paper presented at: Annual Meeting of the Association for Computational Linguistics.2020:4190\u20104197.https:\/\/doi.org\/10.18653\/v1\/2020.acl\u2010main.385","DOI":"10.18653\/v1\/2020.acl-main.385"}],"container-title":["Medical Physics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.15852","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/full-xml\/10.1002\/mp.15852","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/mp.15852","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,10,19]],"date-time":"2023-10-19T16:12:10Z","timestamp":1697731930000},"score":1,"resource":{"primary":{"URL":"https:\/\/aapm.onlinelibrary.wiley.com\/doi\/10.1002\/mp.15852"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,7,30]]},"references-count":44,"journal-issue":{"issue":"9","published-print":{"date-parts":[[2022,9]]}},"alternative-id":["10.1002\/mp.15852"],"URL":"https:\/\/doi.org\/10.1002\/mp.15852","archive":["Portico"],"relation":{},"ISSN":["0094-2405","2473-4209"],"issn-type":[{"value":"0094-2405","type":"print"},{"value":"2473-4209","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,7,30]]},"assertion":[{"value":"2022-01-09","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-06-27","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-07-30","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}