{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,14]],"date-time":"2026-04-14T11:24:46Z","timestamp":1776165886837,"version":"3.50.1"},"reference-count":18,"publisher":"Springer Science and Business Media LLC","license":[{"start":{"date-parts":[[2026,4,14]],"date-time":"2026-04-14T00:00:00Z","timestamp":1776124800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,4,14]],"date-time":"2026-04-14T00:00:00Z","timestamp":1776124800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100010391","name":"Tokyo Women's Medical University","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100010391","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Int J CARS"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Purpose<\/jats:title>\n                    <jats:p>Large language models (LLMs) may hold the potential to infer pathological diagnoses from imaging findings such as computed tomography or magnetic resonance imaging (MRI). This retrospective study investigates whether an LLM can accurately predict histopathological types of breast lesions based on descriptive findings in contrast-enhanced breast MRI reports written in natural language.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We retrospectively analyzed findings from diagnostic imaging reports of consecutive cases of contrast-enhanced breast MRI performed between January and December 2024. Textual descriptions of imaging findings were entered into an LLM (OpenAI o3), and its predictions of histopathological types were compared with the actual pathological diagnoses.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>A total of 186 lesions from 180 patients were classified into 10 histopathological types, including lesions with combinations of these types. The LLM o3 generated predictions for eight of these types. For the most prevalent type, invasive ductal carcinoma (123 lesions), the model achieved 83.7% sensitivity, 60.8% specificity, a positive predictive value (PPV) of 76.9%, and a negative predictive value (NPV) of 70.6%. For the second-most frequent type, ductal carcinoma in situ (38 lesions), the model achieved 57.9% sensitivity, 89.6% specificity, PPV of 56.4%, and NPV of 90.2%.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>The latest LLM accurately inferred invasive ductal carcinoma with typical breast MRI findings, supporting the hypothesis that LLMs can predict histopathological types based on imaging descriptions.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1007\/s11548-026-03619-x","type":"journal-article","created":{"date-parts":[[2026,4,14]],"date-time":"2026-04-14T10:01:36Z","timestamp":1776160896000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Estimation of histopathological types from breast MRI findings using a large language model"],"prefix":"10.1007","author":[{"ORCID":"https:\/\/orcid.org\/0009-0005-4308-0746","authenticated-orcid":false,"given":"Rie","family":"Kanasaki","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kazufumi","family":"Suzuki","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Takami","family":"Ota","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Sadako","family":"Akashi-Tanaka","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yoji","family":"Nagashima","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Shuji","family":"Sakai","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"297","published-online":{"date-parts":[[2026,4,14]]},"reference":[{"issue":"1","key":"3619_CR1","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.231040","volume":"308","author":"D Ueda","year":"2023","unstructured":"Ueda D, Mitsuyama Y, Takita H, Horiuchi D, Walston SL, Tatekawa H et al (2023) Diagnostic performance of ChatGPT from patient history and imaging findings on the diagnosis please quizzes. Radiology 308(1):e231040. https:\/\/doi.org\/10.1148\/radiol.231040","journal-title":"Radiology"},{"issue":"1","key":"3619_CR2","doi-asserted-by":"publisher","first-page":"51","DOI":"10.1007\/s11604-024-01643-y","volume":"43","author":"K Suzuki","year":"2025","unstructured":"Suzuki K, Yamada H, Yamazaki H, Honda G, Sakai S (2025) Preliminary assessment of TNM classification performance for pancreatic cancer in Japanese radiology reports using GPT-4. Jpn J Radiol 43(1):51\u201355. https:\/\/doi.org\/10.1007\/s11604-024-01643-y","journal-title":"Jpn J Radiol"},{"issue":"6","key":"3619_CR3","doi-asserted-by":"publisher","first-page":"1415","DOI":"10.2214\/ajr.167.6.8956569","volume":"167","author":"GN Rodenko","year":"1996","unstructured":"Rodenko GN, Harms SE, Pruneda JM, Farrell RS, Evans WP, Copit DS et al (1996) MR imaging in the management before surgery of lobular carcinoma of the breast: correlation with pathology. AJR Am J Roentgenol 167(6):1415\u20131419. https:\/\/doi.org\/10.2214\/ajr.167.6.8956569","journal-title":"AJR Am J Roentgenol"},{"issue":"7","key":"3619_CR4","doi-asserted-by":"publisher","first-page":"1209","DOI":"10.1007\/s00330-004-2275-7","volume":"14","author":"K Schelfout","year":"2004","unstructured":"Schelfout K, Van Goethem M, Kersschot E, Verslegers I, Biltjes I, Leyman P et al (2004) Preoperative breast MRI in patients with invasive lobular breast cancer. Eur Radiol 14(7):1209\u20131216. https:\/\/doi.org\/10.1007\/s00330-004-2275-7","journal-title":"Eur Radiol"},{"issue":"6","key":"3619_CR5","doi-asserted-by":"publisher","first-page":"1805","DOI":"10.2214\/ajr.183.6.01831805","volume":"183","author":"K Kitagawa","year":"2004","unstructured":"Kitagawa K, Sakuma H, Ishida N, Hirano T, Ishihara A, Takeda K (2004) Contrast-enhanced high-resolution MRI of invasive breast cancer: correlation with histopathologic subtypes. AMJ Am J Roentgenol 183(6):1805\u20131809","journal-title":"AMJ Am J Roentgenol"},{"issue":"2","key":"3619_CR6","doi-asserted-by":"publisher","first-page":"313","DOI":"10.2214\/AJR.05.0881","volume":"187","author":"M Tozaki","year":"2006","unstructured":"Tozaki M, Igarashi T, Fukuda KM (2006) Breast MRI using the VIBE sequence: Clustered ring enhancement in the differential diagnosis of lesions showing non-masslike enhancement. AJR Am J Roentgenol 187(2):313\u2013321","journal-title":"AJR Am J Roentgenol"},{"issue":"6","key":"3619_CR7","doi-asserted-by":"publisher","first-page":"545","DOI":"10.1111\/j.1524-4741.2007.00513.x","volume":"13","author":"EL Rosen","year":"2007","unstructured":"Rosen EL, Smith Foley SA, Demartini WB, Eby PR, Peacock S, Lehman CDEL (2007) BI\u2010RADS MRI enhancement characteristics of ductal carcinoma in situ. Breast J 13(6):545\u2013550","journal-title":"Breast J"},{"issue":"2","key":"3619_CR8","doi-asserted-by":"publisher","first-page":"293","DOI":"10.1002\/jmri.22632","volume":"34","author":"RM Mann","year":"2011","unstructured":"Mann RM, Veltman J, Huisman H, Boetes C (2011) Comparison of enhancement characteristics between invasive lobular carcinoma and invasive ductal carcinoma. J Magn Reson Imaging 34(2):293\u2013300. https:\/\/doi.org\/10.1002\/jmri.22632","journal-title":"J Magn Reson Imaging"},{"issue":"6","key":"3619_CR9","doi-asserted-by":"publisher","first-page":"1435","DOI":"10.1002\/jmri.21622","volume":"28","author":"M Kobayashi","year":"2008","unstructured":"Kobayashi M, Kawashima H, Matsui O, Zen Y, Suzuki M, Inokuchi M et al (2008) Two different types of ring\u2010like enhancement on dynamic MR imaging in breast cancer: correlation with the histopathologic findings. J Magn Reson Imaging 28(6):1435\u20131443","journal-title":"J Magn Reson Imaging"},{"issue":"1","key":"3619_CR10","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s12885-020-07557-y","volume":"20","author":"M Ni","year":"2020","unstructured":"Ni M, Zhou X, Liu J, Yu H, Gao Y, Zhang X et al (2020) Prediction of the clinicopathological subtypes of breast cancer using a Fisher discriminant analysis model based on radiomic features of diffusion-weighted MRI. BMC Cancer 20(1):1","journal-title":"BMC Cancer"},{"issue":"10","key":"3619_CR11","doi-asserted-by":"publisher","first-page":"7551","DOI":"10.21037\/qims-24-200","volume":"14","author":"V Sorin","year":"2024","unstructured":"Sorin V, Klang E, Sobeh T, Konen E, Shrot S, Livne A et al (2024) Generative pre-trained transformer (GPT)-4 support for differential diagnosis in neuroradiology. Quant Imaging Med Surg 14(10):7551\u20137560. https:\/\/doi.org\/10.21037\/qims-24-200","journal-title":"Quant Imaging Med Surg"},{"issue":"6","key":"3619_CR12","doi-asserted-by":"publisher","DOI":"10.1111\/liv.70115","volume":"45","author":"L Sheng","year":"2025","unstructured":"Sheng L, Chen Y, Wei H, Che F, Wu Y, Qin Q et al (2025) Large language models for diagnosing focal liver lesions from CT\/MRI reports: a comparative study with radiologists. Liver Int 45(6):e70115","journal-title":"Liver Int"},{"key":"3619_CR13","doi-asserted-by":"publisher","DOI":"10.1016\/j.cmpb.2025.108765","volume":"265","author":"S Ra","year":"2025","unstructured":"Ra S, Kim J, Na I, Ko ES, Park H (2025) Enhancing radiomics features via a large language model for classifying benign and malignant breast tumors in mammography. Comput Methods Programs Biomed 265:108765. https:\/\/doi.org\/10.1016\/j.cmpb.2025.108765","journal-title":"Comput Methods Programs Biomed"},{"issue":"4","key":"3619_CR14","doi-asserted-by":"publisher","first-page":"1938","DOI":"10.1007\/s00330-024-11032-8","volume":"35","author":"Y Mitsuyama","year":"2025","unstructured":"Mitsuyama Y, Tatekawa H, Takita H, Sasaki F, Tashiro A, Oue S et al (2025) Comparative analysis of GPT-4-based ChatGPT\u2019s diagnostic performance with radiologists using real-world radiology reports of brain tumors. Eur Radiol 35(4):1938\u20131947","journal-title":"Eur Radiol"},{"key":"3619_CR15","doi-asserted-by":"publisher","DOI":"10.1016\/j.acra.2025.09.026","author":"DE Tekcan Sanli","year":"2025","unstructured":"Tekcan Sanli DE, Sanli AN (2025) Critical appraisal of imaging- and pathology-based machine learning models for HER2-positive breast cancer: a review of feature selection and biological plausibility. Acad Radiol. https:\/\/doi.org\/10.1016\/j.acra.2025.09.026","journal-title":"Acad Radiol"},{"issue":"11","key":"3619_CR16","first-page":"96448","volume":"17","author":"S Fujisawa","year":"2025","unstructured":"Fujisawa S, Kurokawa R, Hagiwara A, Kurokawa M, Asari Y, Hatano S et al (2025) Large language model-based automated tumor node metastasis staging and resectability assessment for pancreatic cancer in radiology reports with detection of incomplete documentation. Cureus 17(11):96448","journal-title":"Cureus"},{"issue":"1","key":"3619_CR17","doi-asserted-by":"publisher","DOI":"10.1093\/noajnl\/vdae230","volume":"7","author":"M Foltyn-Dumitru","year":"2025","unstructured":"Foltyn-Dumitru M, Rastogi A, Cho J, Schell M, Mahmutoglu MA, Kessler T et al (2025) The potential of GPT-4 advanced data analysis for radiomics-based machine learning models. Neurooncol Adv 7(1):vdae230. https:\/\/doi.org\/10.1093\/noajnl\/vdae230","journal-title":"Neurooncol Adv"},{"key":"3619_CR18","doi-asserted-by":"publisher","DOI":"10.24488\/twmuj.2025010","author":"K Suzuki","year":"2025","unstructured":"Suzuki K, Yamada H, Yamazaki H, Honda G, Sakai S (2025) TNM classification of pancreatic cancer from unstructured radiology reports using a reasoning language model. Tokyo Women\u2019s Medical Univ J. https:\/\/doi.org\/10.24488\/twmuj.2025010","journal-title":"Tokyo Women\u2019s Medical Univ J"}],"container-title":["International Journal of Computer Assisted Radiology and Surgery"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11548-026-03619-x.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s11548-026-03619-x","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s11548-026-03619-x.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,14]],"date-time":"2026-04-14T10:01:38Z","timestamp":1776160898000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s11548-026-03619-x"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,14]]},"references-count":18,"alternative-id":["3619"],"URL":"https:\/\/doi.org\/10.1007\/s11548-026-03619-x","relation":{},"ISSN":["1861-6429"],"issn-type":[{"value":"1861-6429","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,14]]},"assertion":[{"value":"6 January 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"26 March 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"14 April 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors have no conflicts of interest to declare.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}},{"value":"This retrospective study involved human subjects. Patient identifiers were removed, and this study was conducted following the principles outlined in the Declaration of Helsinki.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Human or animal rights"}},{"value":"The Institutional Review Board approved this retrospective study with a waiver for documented informed consent from patients. The research title was published on an institutional website, and patients were allowed to opt out as per the institutional regulations.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Informed consent"}}]}}