{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,14]],"date-time":"2026-10-14T18:44:27Z","timestamp":1792003467265,"version":"4.4.0"},"reference-count":34,"publisher":"Wiley","issue":"4","license":[{"start":{"date-parts":[[2022,12,16]],"date-time":"2022-12-16T00:00:00Z","timestamp":1671148800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"content-domain":{"domain":["acsjournals.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Cancer Cytopathology"],"published-print":{"date-parts":[[2023,4]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Several studies have used artificial intelligence (AI) to analyze cytology images, but AI has yet to be adopted in clinical practice. The objective of this study was to demonstrate the accuracy of AI\u2010based image analysis for thyroid fine\u2010needle aspiration cytology (FNAC) and to propose its application in clinical practice.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>In total, 148,395 microscopic images of FNAC were obtained from 393 thyroid nodules to train and validate the data, and EfficientNetV2\u2010L was used as the image\u2010classification model. The 35 nodules that were classified as atypia of undetermined significance (AUS) were predicted using AI training.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The precision\u2010recall area under the curve (PR AUC) was &gt;0.95, except for poorly differentiated thyroid carcinoma (PR AUC\u00a0=\u00a00.49) and medullary thyroid carcinoma (PR AUC\u00a0=\u00a00.91). Poorly differentiated thyroid carcinoma had the lowest recall (35.4%) and was difficult to distinguish from papillary thyroid carcinoma, medullary thyroid carcinoma, and follicular thyroid carcinoma. Follicular adenomas and follicular thyroid carcinomas were distinguished from each other by 86.7% and 93.9% recall, respectively. For two\u2010dimensional mapping of the data using t\u2010distributed stochastic neighbor embedding, the lymphomas, follicular adenomas, and anaplastic thyroid carcinomas were divided into three, two, and two groups, respectively. Analysis of the AUS nodules showed 94.7% sensitivity, 14.4% specificity, 56.3% positive predictive value, and 66.7% negative predictive value.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>The authors developed an AI\u2010based approach to analyze thyroid FNAC cases encountered in routine practice. This analysis could be useful for the clinical management of AUS and follicular neoplasm nodules (e.g., an online AI platform for thyroid cytology consultations).<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1002\/cncy.22669","type":"journal-article","created":{"date-parts":[[2022,12,16]],"date-time":"2022-12-16T10:51:00Z","timestamp":1671187860000},"page":"217-225","update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":42,"title":["Application of deep learning as an ancillary diagnostic tool for thyroid FNA cytology"],"prefix":"10.1002","volume":"131","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-5558-7142","authenticated-orcid":false,"given":"Mitsuyoshi","family":"Hirokawa","sequence":"first","affiliation":[{"name":"Department of Diagnostic Pathology and Cytology Kuma Hospital  Kobe Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4542-2883","authenticated-orcid":false,"given":"Hirohiko","family":"Niioka","sequence":"additional","affiliation":[{"name":"Institute for Datability Science Osaka University  Suita Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0827-1580","authenticated-orcid":false,"given":"Ayana","family":"Suzuki","sequence":"additional","affiliation":[{"name":"Department of Diagnostic Pathology and Cytology Kuma Hospital  Kobe Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4597-6109","authenticated-orcid":false,"given":"Masatoshi","family":"Abe","sequence":"additional","affiliation":[{"name":"Institute for Datability Science Osaka University  Suita Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3991-6633","authenticated-orcid":false,"given":"Yusuke","family":"Arai","sequence":"additional","affiliation":[{"name":"Institute for Datability Science Osaka University  Suita Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1579-8767","authenticated-orcid":false,"given":"Hajime","family":"Nagahara","sequence":"additional","affiliation":[{"name":"Institute for Datability Science Osaka University  Suita Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0041-197X","authenticated-orcid":false,"given":"Akira","family":"Miyauchi","sequence":"additional","affiliation":[{"name":"Department of Surgery Kuma Hospital  Kobe Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4960-8271","authenticated-orcid":false,"given":"Takashi","family":"Akamizu","sequence":"additional","affiliation":[{"name":"Department of Internal Medicine Kuma Hospital  Kobe Japan"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2022,12,16]]},"reference":[{"key":"e_1_2_10_2_1","doi-asserted-by":"publisher","DOI":"10.1089\/thy.2015.0020"},{"key":"e_1_2_10_3_1","doi-asserted-by":"publisher","DOI":"10.1089\/thy.2012.0211"},{"key":"e_1_2_10_4_1","doi-asserted-by":"publisher","DOI":"10.1089\/ve.2018.0119"},{"key":"e_1_2_10_5_1","doi-asserted-by":"publisher","DOI":"10.1507\/endocrj.ej18\u20100348"},{"key":"e_1_2_10_6_1","doi-asserted-by":"publisher","DOI":"10.1089\/thy.2017.0500"},{"key":"e_1_2_10_7_1","doi-asserted-by":"publisher","DOI":"10.1097\/pap.0000000000000211"},{"key":"e_1_2_10_8_1","doi-asserted-by":"publisher","DOI":"10.5858\/arpa.2016\u20100100\u2010ra"},{"key":"e_1_2_10_9_1","doi-asserted-by":"publisher","DOI":"10.5858\/arpa.2017\u20100174\u2010ra"},{"key":"e_1_2_10_10_1","doi-asserted-by":"publisher","DOI":"10.1001\/jama.2016.17216"},{"key":"e_1_2_10_11_1","doi-asserted-by":"publisher","DOI":"10.1038\/nature21056"},{"key":"e_1_2_10_12_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41591\u2010018\u20100268\u20103"},{"issue":"6","key":"e_1_2_10_13_1","first-page":"494","article-title":"Potential of the back propagation neural network in the morphologic examination of thyroid lesions","volume":"18","author":"Karakitsos P","year":"1996","journal-title":"Anal Quant Cytol Histol"},{"key":"e_1_2_10_14_1","doi-asserted-by":"publisher","DOI":"10.4103\/jpi.jpi_43_18"},{"key":"e_1_2_10_15_1","doi-asserted-by":"publisher","DOI":"10.7150\/jca.28769"},{"key":"e_1_2_10_16_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2020.101814"},{"key":"e_1_2_10_17_1","doi-asserted-by":"publisher","DOI":"10.1002\/dc.23880"},{"key":"e_1_2_10_18_1","first-page":"10096","article-title":"EfficientNetV2: smaller models and faster training","volume":"139","author":"Tan M","year":"2021","journal-title":"Proceedings of the 38th International Conference on Machine Learning. PMLR"},{"key":"e_1_2_10_19_1","volume-title":"EfficientNet: rethinking model scaling for convolutional neural networks","author":"Tan M","year":"2019"},{"key":"e_1_2_10_20_1","unstructured":"WightmanR.PyTorch Image Models.GitHub;2021. Accessed 2021.https:\/\/github.com\/rwightman\/pytorch\u2010image\u2010models"},{"key":"e_1_2_10_21_1","volume-title":"CutMix: regularization strategy to train strong classifiers with localizable features","author":"Yun S","year":"2019"},{"key":"e_1_2_10_22_1","unstructured":"LoshchilovI HutterF.Decoupled weight decay regularization. arXiv preprint.2017;arXiv:1711.05101."},{"key":"e_1_2_10_23_1","unstructured":"GuY LiC XieJ.Attention\u2010aware generalized mean pooling for image retrieval. arXiv preprint.2018;arXiv:1811.00202."},{"key":"e_1_2_10_24_1","unstructured":"M\u00fcllerR KornblithS HintonG.When does label smoothing help?arXiv preprint.2019;arXiv:1906.02629."},{"key":"e_1_2_10_25_1","doi-asserted-by":"publisher","DOI":"10.1109\/tpami.2018.2858826"},{"key":"e_1_2_10_26_1","unstructured":"HeZ XieL ChenX ZhangY WangY TianQ.Data augmentation revisited: rethinking the distribution gap between clean and augmented data. arXiv preprint.2019;arXiv:1909.09148."},{"key":"e_1_2_10_27_1","first-page":"2579","article-title":"Visualizing data using t\u2010SNE","volume":"9","author":"Maaten L","year":"2008","journal-title":"J Mach Learn Res"},{"key":"e_1_2_10_28_1","doi-asserted-by":"crossref","unstructured":"SelvarajuRR CogswellM DasA VedantamR ParikhD BatraD.Grad\u2010CAM: visual explanations from deep networks via gradient\u2010based localization. In:Proceedings of the IEEE International Conference on Computer Vision (ICCV); Computer Vision Foundation;2017:618\u2010626.","DOI":"10.1109\/ICCV.2017.74"},{"key":"e_1_2_10_29_1","doi-asserted-by":"publisher","DOI":"10.1002\/cncy.22238"},{"key":"e_1_2_10_30_1","doi-asserted-by":"publisher","DOI":"10.1002\/dc.23191"},{"key":"e_1_2_10_31_1","doi-asserted-by":"publisher","DOI":"10.1159\/000487502"},{"key":"e_1_2_10_32_1","doi-asserted-by":"publisher","DOI":"10.1002\/cncy.21500"},{"key":"e_1_2_10_33_1","doi-asserted-by":"publisher","DOI":"10.1507\/endocrj.ej17\u20100238"},{"key":"e_1_2_10_34_1","doi-asserted-by":"publisher","DOI":"10.1002\/dc.24937"},{"key":"e_1_2_10_35_1","doi-asserted-by":"publisher","DOI":"10.1159\/000496502"}],"container-title":["Cancer Cytopathology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/cncy.22669","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/full-xml\/10.1002\/cncy.22669","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/acsjournals.onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/cncy.22669","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,12]],"date-time":"2025-10-12T09:24:43Z","timestamp":1760261083000},"score":1,"resource":{"primary":{"URL":"https:\/\/acsjournals.onlinelibrary.wiley.com\/doi\/10.1002\/cncy.22669"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,12,16]]},"references-count":34,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2023,4]]}},"alternative-id":["10.1002\/cncy.22669"],"URL":"https:\/\/doi.org\/10.1002\/cncy.22669","archive":["Portico"],"relation":{},"ISSN":["1934-662X","1934-6638"],"issn-type":[{"value":"1934-662X","type":"print"},{"value":"1934-6638","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,12,16]]},"assertion":[{"value":"2022-06-29","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-11-07","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-12-16","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}