{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,24]],"date-time":"2026-01-24T01:59:27Z","timestamp":1769219967763,"version":"3.49.0"},"reference-count":31,"publisher":"Wiley","issue":"4","license":[{"start":{"date-parts":[[2024,12,29]],"date-time":"2024-12-29T00:00:00Z","timestamp":1735430400000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["World j. surg."],"published-print":{"date-parts":[[2025,4]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec><jats:title>Background<\/jats:title><jats:p>Despite widespread use of standardized classification systems, risk stratification of thyroid nodules is nuanced and often requires diagnostic surgery. Genomic sequencing is available for this dilemma however, costs and access restricts global applicability. Artificial intelligence (AI) has the potential to overcome this issue nevertheless, the need for black\u2010box interpretability is pertinent. We aimed to create an ultrasonographic segmentation and classification model that offers explainability and risk accountability.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methodology<\/jats:title><jats:p>Four hundred and fourteen ultrasonography images were collected from 105 patients undergoing thyroidectomy, divided into training and testing groups. Classification ground truth used is exclusively surgical histopathology. Relevant nodules were manually annotated by a dedicated study radiologist and surgeon. Three AI architectures with and without block attention modules were trained to identify the relevant nodule and the best performing was selected for the subsequent task in classifying identified nodules into benign or malignant. Gradient\u2010Weighted Class Activation Map is used to provide saliency mapping for visual interpretability.<\/jats:p><\/jats:sec><jats:sec><jats:title>Findings<\/jats:title><jats:p>Superior performance was recorded by the block attention model which stratified thyroid nodules into benign versus malignant with an accuracy of 93% versus 90%, F\u2010score 90% versus 89%, sensitivity 93% versus 91% and specificity 92% versus 91% on a training dataset versus a testing dataset respectively.<\/jats:p><\/jats:sec><jats:sec><jats:title>GradCAM<\/jats:title><jats:p>Visual interpretability maps demonstrate salient areas for a benign nodule diagnosis overlaps spongiform areas and malignant diagnosis salient areas overlap solid components of a partially cystic\u2010solid nodule and microcalcifications within nodules. These findings are consistent with established diagnostic criteria for benign and malignant nodules.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusion<\/jats:title><jats:p>We developed an image segmentation and classification model for the risk stratification of thyroid nodules benchmarking surgical histopathology as ground truth and providing visual interpretability.<\/jats:p><\/jats:sec>","DOI":"10.1002\/wjs.12464","type":"journal-article","created":{"date-parts":[[2024,12,29]],"date-time":"2024-12-29T21:49:37Z","timestamp":1735508977000},"page":"1024-1033","source":"Crossref","is-referenced-by-count":2,"title":["Attention\u2010based image segmentation and classification model for the preoperative risk stratification of thyroid nodules"],"prefix":"10.1002","volume":"49","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-0379-6476","authenticated-orcid":false,"given":"Karishma","family":"Jassal","sequence":"first","affiliation":[{"name":"Monash University Endocrine Surgery Unit Department of General Surgery Alfred Hospital  Melbourne Victoria Australia"},{"name":"Department of Surgery Central Clinical School Monash University  Melbourne Victoria Australia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Bruno","family":"Di Muzio","sequence":"additional","affiliation":[{"name":"Department of Radiology Alfred Hospital  Melbourne Victoria Australia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Melissa","family":"Edwards","sequence":"additional","affiliation":[{"name":"Department of Surgery Peter MacCallum Cancer Centre  Melbourne Victoria Australia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Wendy","family":"Brown","sequence":"additional","affiliation":[{"name":"Department of Surgery Central Clinical School Monash University  Melbourne Victoria Australia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jonathan","family":"Serpell","sequence":"additional","affiliation":[{"name":"Monash University Endocrine Surgery Unit Department of General Surgery Alfred Hospital  Melbourne Victoria Australia"},{"name":"Department of Surgery Central Clinical School Monash University  Melbourne Victoria Australia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Afsaneh","family":"Koohestani","sequence":"additional","affiliation":[{"name":"Monash University Endocrine Surgery Unit Department of General Surgery Alfred Hospital  Melbourne Victoria Australia"},{"name":"Department of Surgery Central Clinical School Monash University  Melbourne Victoria Australia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6679-274X","authenticated-orcid":false,"given":"James C.","family":"Lee","sequence":"additional","affiliation":[{"name":"Monash University Endocrine Surgery Unit Department of General Surgery Alfred Hospital  Melbourne Victoria Australia"},{"name":"Department of Surgery Central Clinical School Monash University  Melbourne Victoria Australia"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"311","published-online":{"date-parts":[[2024,12,29]]},"reference":[{"key":"e_1_2_9_2_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ejrad.2013.02.002"},{"key":"e_1_2_9_3_1","doi-asserted-by":"publisher","DOI":"10.1089\/thy.2004.14.926"},{"key":"e_1_2_9_4_1","doi-asserted-by":"publisher","DOI":"10.7326\/0003-4819-126-3-199702010-00009"},{"key":"e_1_2_9_5_1","doi-asserted-by":"publisher","DOI":"10.1136\/bmj.l6670"},{"key":"e_1_2_9_6_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.jacr.2017.01.046"},{"key":"e_1_2_9_7_1","doi-asserted-by":"publisher","DOI":"10.1089\/thy.2021.0558"},{"key":"e_1_2_9_8_1","doi-asserted-by":"publisher","DOI":"10.1007\/s00268\u2010022\u201006798\u20101"},{"key":"e_1_2_9_9_1","doi-asserted-by":"publisher","DOI":"10.21037\/atm\u201021\u20104328"},{"key":"e_1_2_9_10_1","doi-asserted-by":"publisher","DOI":"10.1109\/access.2020.2982767"},{"key":"e_1_2_9_11_1","unstructured":"Hanirex D. K.andK.Kaliyamurthie.2013. \u201cMulti\u2010Classification Approach for Detecting Thyroid Attacks.\u201d"},{"key":"e_1_2_9_12_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ejrad.2019.02.029"},{"key":"e_1_2_9_13_1","doi-asserted-by":"publisher","DOI":"10.1007\/s11227\u2010018\u20102469\u20104"},{"key":"e_1_2_9_14_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"e_1_2_9_15_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-46640-4_25"},{"key":"e_1_2_9_16_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.bspc.2021.103077"},{"key":"e_1_2_9_17_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.667"},{"key":"e_1_2_9_18_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.74"},{"key":"e_1_2_9_19_1","first-page":"1050","volume-title":"Proceedings of International Conference on Machine Learning","author":"Gal Y.","year":"2016"},{"key":"e_1_2_9_20_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41746\u2010022\u201000709\u20103"},{"key":"e_1_2_9_21_1","doi-asserted-by":"publisher","DOI":"10.3390\/s21217241"},{"key":"e_1_2_9_22_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW53098.2021.00251"},{"key":"e_1_2_9_23_1","doi-asserted-by":"publisher","DOI":"10.1109\/TIT.2014.2320500"},{"key":"e_1_2_9_24_1","doi-asserted-by":"publisher","DOI":"10.1016\/0002\u20109343(92)90163\u20106"},{"key":"e_1_2_9_25_1","doi-asserted-by":"publisher","DOI":"10.1159\/000339959"},{"key":"e_1_2_9_26_1","doi-asserted-by":"publisher","DOI":"10.1007\/s12022\u2010011\u20109182\u20107"},{"key":"e_1_2_9_27_1","doi-asserted-by":"publisher","DOI":"10.1159\/000334218"},{"key":"e_1_2_9_28_1","doi-asserted-by":"publisher","DOI":"10.1016\/s2589\u20107500(21)00041\u20108"},{"key":"e_1_2_9_29_1","doi-asserted-by":"publisher","DOI":"10.1089\/thy.2019.0752"},{"key":"e_1_2_9_30_1","doi-asserted-by":"publisher","DOI":"10.1530\/etj\u201021\u20100129"},{"key":"e_1_2_9_31_1","doi-asserted-by":"publisher","DOI":"10.1111\/cen.14612"},{"key":"e_1_2_9_32_1","unstructured":"Rolnick D. A.Veit S.Belongie&N.Shavit2017. \u201cDeep Learning Is Robust to Massive Label Noise.\u201d arXiv preprint arXiv:1705.10694."}],"container-title":["World Journal of Surgery"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1002\/wjs.12464","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T01:13:09Z","timestamp":1760058789000},"score":1,"resource":{"primary":{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/10.1002\/wjs.12464"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,12,29]]},"references-count":31,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2025,4]]}},"alternative-id":["10.1002\/wjs.12464"],"URL":"https:\/\/doi.org\/10.1002\/wjs.12464","archive":["Portico"],"relation":{},"ISSN":["0364-2313","1432-2323"],"issn-type":[{"value":"0364-2313","type":"print"},{"value":"1432-2323","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,12,29]]}}}