{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T04:58:18Z","timestamp":1785387498391,"version":"3.55.0"},"reference-count":17,"publisher":"Springer Science and Business Media LLC","license":[{"start":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T00:00:00Z","timestamp":1785369600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.springernature.com\/gp\/researchers\/text-and-data-mining"},{"start":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T00:00:00Z","timestamp":1785369600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.springernature.com\/gp\/researchers\/text-and-data-mining"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Int J Cardiovasc Imaging"],"DOI":"10.1007\/s10554-026-03785-8","type":"journal-article","created":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T04:03:31Z","timestamp":1785384211000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Benchmarking 54 large language model configurations for CAD-RADS scoring: open-weight models approach human-level agreement"],"prefix":"10.1007","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7218-9306","authenticated-orcid":false,"given":"Veit","family":"Sandfort","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3798-9812","authenticated-orcid":false,"given":"Davis M.","family":"Vigneault","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6991-6557","authenticated-orcid":false,"given":"Martin J.","family":"Willemink","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jie","family":"Wu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9442-6173","authenticated-orcid":false,"given":"Richard L.","family":"Hallett","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8312-0598","authenticated-orcid":false,"given":"Koen","family":"Nieman","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0715-0952","authenticated-orcid":false,"given":"Dominik","family":"Fleischmann","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8227-0757","authenticated-orcid":false,"given":"Domenico","family":"Mastrodicasa","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,7,30]]},"reference":[{"key":"3785_CR1","doi-asserted-by":"publisher","first-page":"e151","DOI":"10.5114\/pjr.2018.75641","volume":"83","author":"A Abdel Razek","year":"2018","unstructured":"Abdel Razek A, Elrakhawy M, Yossof M et al (2018) Inter-observer agreement of the coronary artery disease reporting and data system (cad-rads) in patients with stable chest pain. Pol J Radiol 83:e151\u2013e159. https:\/\/doi.org\/10.5114\/pjr.2018.75641","journal-title":"Pol J Radiol"},{"issue":"3","key":"3785_CR2","doi-asserted-by":"publisher","first-page":"322","DOI":"10.1016\/j.jcct.2025.03.007","volume":"19","author":"P Arnold","year":"2025","unstructured":"Arnold P, Russe M, Bamberg F et al (2025) Performance of large language models for cad-rads 2.0 classification derived from cardiac ct reports. J Cardiovasc Comput Tomogr 19(3):322\u2013330. https:\/\/doi.org\/10.1016\/j.jcct.2025.03.007","journal-title":"J Cardiovasc Comput Tomogr"},{"issue":"1","key":"3785_CR3","doi-asserted-by":"publisher","first-page":"117","DOI":"10.1186\/s13244-019-0806-7","volume":"10","author":"M Basha","year":"2019","unstructured":"Basha M, Aly S, Ismail A et al (2019) The validity and applicability of cad-rads in the management of patients with coronary artery disease. Insights Imaging 10(1):117. https:\/\/doi.org\/10.1186\/s13244-019-0806-7","journal-title":"Insights Imaging"},{"key":"3785_CR4","doi-asserted-by":"publisher","first-page":"110271","DOI":"10.1016\/j.clinimag.2024.110271","volume":"114","author":"E Camur","year":"2024","unstructured":"Camur E, Cesur T, Gunes Y (2024) Can large language models be new supportive tools in coronary computed tomography angiography reporting? Clin Imaging 114:110271. https:\/\/doi.org\/10.1016\/j.clinimag.2024.110271","journal-title":"Clin Imaging"},{"key":"3785_CR5","doi-asserted-by":"publisher","first-page":"e70967","DOI":"10.2196\/70967","volume":"13","author":"Y Chen","year":"2025","unstructured":"Chen Y, Dong M, Sun J et al (2025) Leveraging gpt-4o for automated extraction and categorization of cad-rads features from free-text coronary ct angiography reports: Diagnostic study. JMIR Med Inform 13:e70967. https:\/\/doi.org\/10.2196\/70967","journal-title":"JMIR Med Inform"},{"key":"3785_CR6","doi-asserted-by":"publisher","unstructured":"Cury R, Abbara S, Achenbach S, et\u00a0al (2016) Cad-rads\u2122: Coronary artery disease - reporting and data system: An expert consensus document of the society of cardiovascular computed tomography (scct), the american college of radiology (acr) and the north american society for cardiovascular imaging (nasci). endorsed by the american college of cardiology. Journal of the American College of Radiology : JACR 13(12 Pt A):1458\u20131466.e9. https:\/\/doi.org\/10.1016\/j.jacr.2016.04.024","DOI":"10.1016\/j.jacr.2016.04.024"},{"issue":"5","key":"3785_CR7","doi-asserted-by":"publisher","DOI":"10.1148\/ryct.220183","volume":"4","author":"R Cury","year":"2022","unstructured":"Cury R, Leipsic J, Abbara S et al (2022) CAD-RADS\u21222.0 - 2022 coronary artery disease - reporting and data system an expert consensus document of the Society of Cardiovascular Computed Tomography (SCCT), the American College of Cardiology (ACC), the American College of Radiology (ACR) and the North America Society of Cardiovascular Imaging (NASCI). Radiol Cardiothorac Imaging 4(5):e220183. https:\/\/doi.org\/10.1148\/ryct.220183","journal-title":"Radiol Cardiothorac Imaging"},{"issue":"5","key":"3785_CR8","doi-asserted-by":"publisher","DOI":"10.1093\/jamiaopen\/ooaf097","volume":"8","author":"N Dao","year":"2025","unstructured":"Dao N, Quesada L, Hassan S et al (2025) Generative artificial intelligence for automated data extraction from unstructured medical text. JAMIA Open 8(5):ooaf097. https:\/\/doi.org\/10.1093\/jamiaopen\/ooaf097","journal-title":"JAMIA Open"},{"key":"3785_CR9","doi-asserted-by":"publisher","unstructured":"Gemma Team, Kamath A, Ferret J, et\u00a0al (2025) Gemma 3 technical report. arXiv preprint https:\/\/doi.org\/10.48550\/arXiv.2503.19786, arXiv:2503.19786","DOI":"10.48550\/arXiv.2503.19786"},{"key":"3785_CR10","doi-asserted-by":"publisher","DOI":"10.1007\/s00117-026-01580-z","author":"E Kaba","year":"2026","unstructured":"Kaba E, \u00c7ubuk\u00e7u Y, Uzunibrahimo\u011flu B (2026) Retrieval-augmented generation-enhanced large language models for comprehensive CAD-RADS 2.0 categorization from structured coronary CTA reports. Radiologie (Heidelb). https:\/\/doi.org\/10.1007\/s00117-026-01580-z","journal-title":"Radiologie (Heidelb)"},{"issue":"2","key":"3785_CR11","doi-asserted-by":"publisher","first-page":"217","DOI":"10.1097\/RCT.0000000000001802","volume":"50","author":"H Kaya","year":"2026","unstructured":"Kaya H (2026) Enhancing the cad-rads 2.0 category assignment performance of chatgpt and deepseek through \u201cfew-shot\u2019\u2019 prompting. J Comput Assist Tomogr 50(2):217\u2013221. https:\/\/doi.org\/10.1097\/RCT.0000000000001802","journal-title":"J Comput Assist Tomogr"},{"key":"3785_CR12","doi-asserted-by":"publisher","first-page":"100689","DOI":"10.1016\/j.esmorw.2026.100689","volume":"11","author":"T\u00d3 Kristj\u00e1nsson","year":"2026","unstructured":"Kristj\u00e1nsson T\u00d3, Henriksen AF, Hansen ML et al (2026) Prompting large language models and evaluating inter- and intra-rater agreement for cancer progression assessment from radiology reports. ESMO Real World Data and Digital Oncology 11:100689. https:\/\/doi.org\/10.1016\/j.esmorw.2026.100689","journal-title":"ESMO Real World Data and Digital Oncology"},{"key":"3785_CR13","doi-asserted-by":"publisher","DOI":"10.1186\/s13244-026-02285-6","author":"G Lorusso","year":"2026","unstructured":"Lorusso G, Ruscino G, Spitaleri A et al (2026) Comparative evaluation of large language models for generating CAD-RADS 2.0-compliant diagnostic conclusions in cardiac CT reports. Insights Imaging. https:\/\/doi.org\/10.1186\/s13244-026-02285-6","journal-title":"Insights Imaging"},{"issue":"9","key":"3785_CR14","doi-asserted-by":"publisher","first-page":"817","DOI":"10.3348\/kjr.2025.0293","volume":"26","author":"D Min","year":"2025","unstructured":"Min D, Jin K, Bang S et al (2025) Large language models for cad-rads 2.0 extraction from semi-structured coronary ct angiography reports: A multi-institutional study. Korean J Radiol 26(9):817\u2013831. https:\/\/doi.org\/10.3348\/kjr.2025.0293","journal-title":"Korean J Radiol"},{"issue":"4","key":"3785_CR15","doi-asserted-by":"publisher","first-page":"2303","DOI":"10.1007\/s10278-024-01328-y","volume":"38","author":"M Silbergleit","year":"2025","unstructured":"Silbergleit M, T\u00f3th A, Chamberlin J et al (2025) Chatgpt vs gemini: Comparative accuracy and efficiency in cad-rads score assignment from radiology reports. Journal of imaging informatics in medicine 38(4):2303\u20132311. https:\/\/doi.org\/10.1007\/s10278-024-01328-y","journal-title":"Journal of imaging informatics in medicine"},{"key":"3785_CR16","volume-title":"Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again","author":"E Topol","year":"2019","unstructured":"Topol E (2019) Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again, 1st edn. Basic Books Inc, USA","edition":"1st edn"},{"key":"3785_CR17","doi-asserted-by":"publisher","DOI":"10.1007\/s10278-025-01704-2","author":"W Yoo","year":"2025","unstructured":"Yoo W, Son J, Kim J et al (2025) Large language models versus human readers in CAD-RADS 2.0 categorization of coronary CT angiography reports. J Imaging Inform Med. https:\/\/doi.org\/10.1007\/s10278-025-01704-2","journal-title":"J Imaging Inform Med"}],"container-title":["The International Journal of Cardiovascular Imaging"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10554-026-03785-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10554-026-03785-8","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10554-026-03785-8.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T04:03:32Z","timestamp":1785384212000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10554-026-03785-8"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,30]]},"references-count":17,"alternative-id":["3785"],"URL":"https:\/\/doi.org\/10.1007\/s10554-026-03785-8","relation":{},"ISSN":["1875-8312"],"issn-type":[{"value":"1875-8312","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,7,30]]},"assertion":[{"value":"9 May 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"19 July 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"30 July 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"Data and information were controlled by authors without industry employment (V.S., D.V.). Two authors (M.J.W., J.W.) are industry employees or founders of Segmed, Inc. (Palo Alto, California), a company that delivers machine learning training and validation data for medical imaging. Two authors (D.F., D.M.) are advisors and stockholders of Segmed Inc. Deputy Editor, Radiology: Cardiothoracic Imaging (D.F.). Trainee Deputy Editor, Radiology: Cardiothoracic Imaging (D.M.). Junior Deputy Editor, European Radiology (M.J.W.). For the current project, none of the authors received financial support from the industry. The remaining authors (D.V., R.L.H., K.N.) declare no competing interests relevant to the content of this article.","order":1,"name":"Ethics","label":"Competing Interests","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Due to the non-identifiable nature of the data, IRB approval was not required.","order":2,"name":"Ethics","label":"Ethics Approval","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Not applicable.","order":3,"name":"Ethics","label":"Consent to Participate","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Not applicable.","order":4,"name":"Ethics","label":"Consent to Publish","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"This study did not develop a novel algorithm or model; analyses used standard LLM inference (cloud APIs and\n                      llama.cpp\n                      ) and standard statistical methods (unweighted Cohen\u2019s\n                      \n                      with bootstrap confidence intervals). Boilerplate inference and parsing scripts are available from the corresponding author on reasonable request. As these scripts are thin, frequently-changing wrappers around third-party provider APIs, they carry limited standalone reproducibility value; the exact zero-shot prompt, which is the principal determinant of the results, is instead reproduced in full in this article (Appendix\u00a0A).","order":5,"name":"Ethics","label":"Code Availability","group":{"name":"EthicsHeading","label":"Declarations"}}]}}