{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,16]],"date-time":"2026-09-16T00:59:32Z","timestamp":1789520372644,"version":"build-2803163510"},"reference-count":38,"publisher":"Springer Science and Business Media LLC","issue":"10","license":[{"start":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T00:00:00Z","timestamp":1779753600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T00:00:00Z","timestamp":1779753600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Eur Radiol"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Objectives<\/jats:title>\n                    <jats:p>Large language models (LLMs) are increasingly explored as decision-support tools in medical imaging. However, their ability to align with country-specific guidelines, which often diverge, remains uncertain. We set out to evaluate the geographic neutrality of three state-of-the-art LLMs\u2014GPT-o3, Mistral Large, and DeepSeek R1\u2014and a biomedical LLM (MedGemma 1.5 4B), when applied to neuroradiology scenarios with conflicting U.S. and non-U.S. recommendations.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Materials and methods<\/jats:title>\n                    <jats:p>Vignettes derived from contradictory international guidelines were presented to each model under two conditions: an implicit setting, where no guideline was specified and vignettes were provided in English and French; and an explicit setting, where prompts directed models to follow a named guideline. Performance was reviewed against the target guideline, and mitigation strategies were tested.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Thirty clinical vignettes presenting conflicting guidelines were evaluated by GPT-o3, Mistral Large, and DeepSeek R1. In the implicit setting, all models favored U.S. guidelines, with GPT-o3, Mistral, and DeepSeek aligning with them in 27 of 30 scenarios (90.0%; 95% CI, 74.4\u201396.5). In the explicit setting, adherence declined sharply for non-U.S. recommendations for all models. Providing the complete guideline text was the most effective mitigation strategy, restoring accuracies above 90% across all models.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>Across languages and model origins, LLMs exhibited a systematic bias toward U.S. neuroradiology guidelines, even when explicitly instructed otherwise. This U.S.-centrism likely reflects training data imbalances and raises concerns for safe global deployment. Strategies for local contextualization, such as guideline integration at deployment, are necessary to ensure context-appropriate clinical decision support.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Key Points<\/jats:title>\n                    <jats:p>\n                      <jats:bold>\n                        <jats:italic>Question<\/jats:italic>\n                      <\/jats:bold>\n                      <jats:italic>Do large language models display geographical neutrality in neuroradiology decision support?<\/jats:italic>\n                    <\/jats:p>\n                    <jats:p>\n                      <jats:bold>\n                        <jats:italic>Findings<\/jats:italic>\n                      <\/jats:bold>\n                      <jats:italic>Even models developed in France and China systematically preferred United States guidelines, aligning with them in most implicit scenarios while failing to follow explicit guidelines from other sources.<\/jats:italic>\n                    <\/jats:p>\n                    <jats:p>\n                      <jats:bold>\n                        <jats:italic>Clinical relevance<\/jats:italic>\n                      <\/jats:bold>\n                      <jats:italic>This systematic United States-centric bias poses clinical and legal risks for global deployment. Safe implementation requires specific localization strategies, such as providing full guideline texts, to ensure recommendations align with local practice standards.<\/jats:italic>\n                    <\/jats:p>\n                  <\/jats:sec>","DOI":"10.1007\/s00330-026-12634-0","type":"journal-article","created":{"date-parts":[[2026,5,26]],"date-time":"2026-05-26T11:01:24Z","timestamp":1779793284000},"page":"7444-7453","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Cultural bias in large language models\u2019 ability to follow neuroradiology guidelines"],"prefix":"10.1007","volume":"36","author":[{"given":"Na\u00ebl","family":"Bazerbachi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rapha\u00ebl","family":"Bentegeac","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Giada","family":"Pistilli","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Su-Hwan","family":"Kim","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Philippe","family":"Amouyel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jean-Pierre","family":"Pruvo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lotfi","family":"Hacein-Bey","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aghiles","family":"Hamroun","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Gr\u00e9gory","family":"Kuchcinski","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3618-4791","authenticated-orcid":false,"given":"Bastien","family":"Le Guellec","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,5,26]]},"reference":[{"key":"12634_CR1","doi-asserted-by":"publisher","DOI":"10.1186\/s12880-022-00798-2","volume":"22","author":"E Kjelle","year":"2022","unstructured":"Kjelle E, Andersen ER, Krokeide AM et al (2022) Characterizing and quantifying low-value diagnostic imaging internationally: a scoping review. BMC Med Imaging 22:73","journal-title":"BMC Med Imaging"},{"key":"12634_CR2","doi-asserted-by":"publisher","first-page":"100","DOI":"10.1186\/s13244-022-01241-4","volume":"13","author":"AP Brady","year":"2022","unstructured":"Brady AP, Beets-Tan RG, Brklja\u010di\u0107 B et al (2022) The role of radiologist in the changing world of healthcare: a white paper of the European Society of Radiology (ESR). Insights Imaging 13:100","journal-title":"Insights Imaging"},{"key":"12634_CR3","doi-asserted-by":"publisher","first-page":"168","DOI":"10.1186\/s13244-023-01525-3","volume":"14","author":"M Becker","year":"2023","unstructured":"Becker M (2023) How to prepare for a bright future of radiology in Europe. Insights Imaging 14:168","journal-title":"Insights Imaging"},{"key":"12634_CR4","doi-asserted-by":"publisher","first-page":"1458","DOI":"10.1001\/jama.282.15.1458","volume":"282","author":"MD Cabana","year":"1999","unstructured":"Cabana MD, Rand CS, Powe NR et al (1999) Why don\u2019t physicians follow clinical practice guidelines? A framework for improvement. JAMA 282:1458\u20131465","journal-title":"JAMA"},{"key":"12634_CR5","doi-asserted-by":"publisher","DOI":"10.1186\/s12913-018-3372-7","volume":"18","author":"AM Gransj\u00f8en","year":"2018","unstructured":"Gransj\u00f8en AM, Wiig S, Lysdahl KB, Hofmann BM (2018) Barriers and facilitators for guideline adherence in diagnostic imaging: an explorative study of GPs\u2019 and radiologists\u2019 perspectives. BMC Health Serv Res 18:556","journal-title":"BMC Health Serv Res"},{"key":"12634_CR6","doi-asserted-by":"publisher","first-page":"1581","DOI":"10.2214\/AJR.08.1622","volume":"192","author":"AB Bautista","year":"2009","unstructured":"Bautista AB, Burgos A, Nickel BJ et al (2009) Do clinicians use the American College of Radiology Appropriateness criteria in the management of their patients? AJR Am J Roentgenol 192:1581\u20131585","journal-title":"AJR Am J Roentgenol"},{"key":"12634_CR7","doi-asserted-by":"publisher","first-page":"1166","DOI":"10.1007\/s00330-024-11083-x","volume":"35","author":"C Singer","year":"2025","unstructured":"Singer C, Saban M, Luxenburg O et al (2025) Computed tomography referral guidelines adherence in Europe: insights from a seven-country audit. Eur Radiol 35:1166\u20131177","journal-title":"Eur Radiol"},{"key":"12634_CR8","doi-asserted-by":"publisher","first-page":"373","DOI":"10.1007\/s10140-024-02234-0","volume":"31","author":"FO Herlihy","year":"2024","unstructured":"Herlihy FO, Dempsey PJ, Gorman D et al (2024) Comparison of international guidelines for CT prior to lumbar puncture in patients with suspected meningitis. Emerg Radiol 31:373\u2013379","journal-title":"Emerg Radiol"},{"key":"12634_CR9","doi-asserted-by":"publisher","first-page":"532","DOI":"10.1007\/s00330-024-10938-7","volume":"35","author":"YX Tay","year":"2025","unstructured":"Tay YX, Foley S, Killeen R et al (2025) Impact and effect of imaging referral guidelines on patients and radiology services: a systematic review. Eur Radiol 35:532\u2013541","journal-title":"Eur Radiol"},{"key":"12634_CR10","doi-asserted-by":"publisher","first-page":"172","DOI":"10.1038\/s41586-023-06291-2","volume":"620","author":"K Singhal","year":"2023","unstructured":"Singhal K, Azizi S, Tu T et al (2023) Large language models encode clinical knowledge. Nature 620:172\u2013180","journal-title":"Nature"},{"key":"12634_CR11","doi-asserted-by":"publisher","unstructured":"Adams LC, Truhn D, Busch F et al (2023) Leveraging GPT-4 for post hoc transformation of free-text radiology reports into structured reporting: a multilingual feasibility study. Radiology. https:\/\/doi.org\/10.1148\/radiol.230725","DOI":"10.1148\/radiol.230725"},{"key":"12634_CR12","doi-asserted-by":"publisher","unstructured":"Miller L, Kamel P, Patel J et al (2024) A comparative evaluation of large language model utility in neuroimaging clinical decision support. J Digit Imaging Inform Med. https:\/\/doi.org\/10.1007\/s10278-024-01161-3","DOI":"10.1007\/s10278-024-01161-3"},{"key":"12634_CR13","doi-asserted-by":"publisher","first-page":"1873","DOI":"10.1038\/s41591-025-03626-6","volume":"31","author":"M Omar","year":"2025","unstructured":"Omar M, Soffer S, Agbareia R et al (2025) Sociodemographic biases in medical decision making by large language models. Nat Med 31:1873\u20131881","journal-title":"Nat Med"},{"key":"12634_CR14","doi-asserted-by":"publisher","unstructured":"Suenghataiphorn T, Tribuddharat N, Danpanichkul P, Kulthamrongsri N (2025) Bias in large language models across clinical applications: a systematic review. Preprint at https:\/\/doi.org\/10.48550\/arXiv.2504.02917","DOI":"10.48550\/arXiv.2504.02917"},{"key":"12634_CR15","doi-asserted-by":"publisher","DOI":"10.1186\/s12939-025-02419-0","volume":"24","author":"M Omar","year":"2025","unstructured":"Omar M, Sorin V, Agbareia R et al (2025) Evaluating and addressing demographic disparities in medical large language models: a systematic review. Int J Equity Health 24:57","journal-title":"Int J Equity Health"},{"key":"12634_CR16","doi-asserted-by":"publisher","unstructured":"Johnson RL, Pistilli G, Men\u00e9dez-Gonz\u00e1lez N et al (2022) The ghost in the machine has an American accent: value conflict in GPT-3. Preprint at https:\/\/doi.org\/10.48550\/arXiv.2203.07785","DOI":"10.48550\/arXiv.2203.07785"},{"key":"12634_CR17","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1038\/s41591-024-03425-5","volume":"31","author":"J Gallifant","year":"2025","unstructured":"Gallifant J, Afshar M, Ameen S et al (2025) The TRIPOD-LLM reporting guideline for studies using large language models. Nat Med 31:60\u201369","journal-title":"Nat Med"},{"key":"12634_CR18","doi-asserted-by":"publisher","unstructured":"Sellergren, A, Gao C, Mahvar, F et al (2026) MedGemma 1.5 Technical Report. https:\/\/doi.org\/10.48550\/arXiv.2604.05081","DOI":"10.48550\/arXiv.2604.05081"},{"key":"12634_CR19","doi-asserted-by":"publisher","first-page":"3576","DOI":"10.1007\/s00330-020-06672-5","volume":"30","author":"MP Recht","year":"2020","unstructured":"Recht MP, Dewey M, Dreyer K et al (2020) Integrating artificial intelligence into the clinical practice of radiology: challenges and recommendations. Eur Radiol 30:3576\u20133584","journal-title":"Eur Radiol"},{"key":"12634_CR20","doi-asserted-by":"publisher","first-page":"20230023","DOI":"10.1259\/bjr.20230023","volume":"96","author":"JW Gichoya","year":"2023","unstructured":"Gichoya JW, Thomas K, Celi LA et al (2023) AI pitfalls and what not to do: mitigating bias in AI. Br J Radiol 96:20230023","journal-title":"Br J Radiol"},{"key":"12634_CR21","doi-asserted-by":"publisher","unstructured":"Wu J, Liu X, Li M et al (2024) Clinical text datasets for medical artificial intelligence and large language models\u2014a systematic review. NEJM AI. https:\/\/doi.org\/10.1056\/AIra2400012","DOI":"10.1056\/AIra2400012"},{"key":"12634_CR22","doi-asserted-by":"publisher","unstructured":"Pavone M, Innocenzi C, Macellari N et al (2025) Assessing the accuracy of large language models on European Guidelines for Cervical Cancer: an in silico benchmarking study. BJOG. https:\/\/doi.org\/10.1111\/1471-0528.70095","DOI":"10.1111\/1471-0528.70095"},{"key":"12634_CR23","doi-asserted-by":"publisher","first-page":"178","DOI":"10.1038\/s43856-023-00418-2","volume":"3","author":"O Br\u00fcck","year":"2023","unstructured":"Br\u00fcck O (2023) A bibliometric analysis of geographic disparities in the authorship of leading medical journals. Commun Med 3:178","journal-title":"Commun Med"},{"key":"12634_CR24","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1016\/j.lungcan.2021.01.024","volume":"154","author":"E Pallari","year":"2021","unstructured":"Pallari E, Eriksson M, Billhult A et al (2021) Lung cancer research and its citation on clinical practice guidelines. Lung Cancer 154:44\u201350","journal-title":"Lung Cancer"},{"key":"12634_CR25","unstructured":"Dzieza J (2023) Inside the AI factory. In: Intelligencer. Available via https:\/\/nymag.com\/intelligencer\/article\/ai-artificial-intelligence-humans-technology-business-factory.html. Accessed 15 Sept 2025"},{"key":"12634_CR26","doi-asserted-by":"publisher","first-page":"100066","DOI":"10.1016\/j.ejrai.2025.100066","volume":"5","author":"B Le Guellec","year":"2026","unstructured":"Le Guellec B, Adambounou K, Adams LC et al (2026) PARROT, an open multilingual radiology reports dataset. Eur J Radiol Artif Intell 5:100066","journal-title":"Eur J Radiol Artif Intell"},{"key":"12634_CR27","unstructured":"EUR-Lex (2024) Regulation (EU) 2024\/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300\/2008, (EU) No 167\/2013, (EU) No 168\/2013, (EU) 2018\/858, (EU) 2018\/1139 and (EU) 2019\/2144 and Directives 2014\/90\/EU, (EU) 2016\/797 and (EU) 2020\/1828 (Artificial Intelligence Act) (Text with EEA relevance) http:\/\/data.europa.eu\/eli\/reg\/2024\/1689"},{"key":"12634_CR28","unstructured":"OpenAI (2026) Introducing ChatGPT Health. Available via https:\/\/openai.com\/index\/introducing-chatgpt-health\/. Accessed 10 Feb 2026"},{"key":"12634_CR29","doi-asserted-by":"publisher","first-page":"111933","DOI":"10.1016\/j.ejrad.2025.111933","volume":"183","author":"YX Tay","year":"2025","unstructured":"Tay YX, Foley SJ, Ong ME et al (2025) Using evidence-based imaging referral guidelines to facilitate appropriate imaging: are they all the same? Eur J Radiol 183:111933","journal-title":"Eur J Radiol"},{"key":"12634_CR30","doi-asserted-by":"publisher","first-page":"1329","DOI":"10.1161\/STROKEAHA.123.045154","volume":"55","author":"CM Rapillo","year":"2024","unstructured":"Rapillo CM, Dunet V, Pistocchi S et al (2024) Moving from CT to MRI paradigm in acute ischemic stroke: feasibility, effects on stroke diagnosis and long-term outcomes. Stroke 55:1329\u20131338","journal-title":"Stroke"},{"key":"12634_CR31","doi-asserted-by":"publisher","first-page":"1717","DOI":"10.1148\/rg.2019190142","volume":"39","author":"CA Potter","year":"2019","unstructured":"Potter CA, Vagal AS, Goyal M et al (2019) CT for treatment selection in acute ischemic stroke: a code stroke primer. Radiographics 39:1717\u20131738","journal-title":"Radiographics"},{"key":"12634_CR32","doi-asserted-by":"publisher","first-page":"131","DOI":"10.5853\/jos.2014.16.3.131","volume":"16","author":"BJ Kim","year":"2014","unstructured":"Kim BJ, Kang HG, Kim H-J et al (2014) Magnetic resonance imaging in acute ischemic stroke treatment. J Stroke 16:131\u2013145","journal-title":"J Stroke"},{"key":"12634_CR33","doi-asserted-by":"publisher","first-page":"313","DOI":"10.1001\/jama.2023.9630","volume":"330","author":"M Duffourc","year":"2023","unstructured":"Duffourc M, Gerke S (2023) Generative AI in health care and liability risks for physicians and safety concerns for patients. JAMA 330:313\u2013314","journal-title":"JAMA"},{"key":"12634_CR34","doi-asserted-by":"publisher","unstructured":"Atari M, Xue MJ, Park PS et al (2023) Which humans? Preprint at https:\/\/doi.org\/10.31234\/osf.io\/5b26t","DOI":"10.31234\/osf.io\/5b26t"},{"key":"12634_CR35","doi-asserted-by":"publisher","unstructured":"Torrielli F (2024) Stars, stripes, and silicon: unravelling the ChatGPT\u2019s all-American, monochrome, cis-centric bias. Preprint at https:\/\/doi.org\/10.48550\/arXiv.2410.13868","DOI":"10.48550\/arXiv.2410.13868"},{"key":"12634_CR36","doi-asserted-by":"crossref","unstructured":"Wu W, Xu X, Gao C et al (2025) Assessing and mitigating medical knowledge drift and conflicts in large language models. In: Christodoulopoulos C, Chakraborty T, Rose C, Peng V (eds) Findings of the Association for Computational Linguistics: EMNLP 2025. Association for Computational Linguistics, Suzhou, pp 707\u2013730","DOI":"10.18653\/v1\/2025.findings-emnlp.38"},{"key":"12634_CR37","doi-asserted-by":"publisher","first-page":"1015","DOI":"10.1093\/jamia\/ocaf045","volume":"32","author":"FJ Dorfner","year":"2025","unstructured":"Dorfner FJ, Dada A, Busch F et al (2025) Evaluating the effectiveness of biomedical fine-tuning for large language models on clinical tasks. J Am Med Inf Assoc 32:1015\u20131024","journal-title":"J Am Med Inf Assoc"},{"key":"12634_CR38","doi-asserted-by":"publisher","first-page":"1072","DOI":"10.1089\/neu.2015.4121","volume":"38","author":"MC Cnossen","year":"2021","unstructured":"Cnossen MC, Scholten AC, Lingsma HF et al (2021) Adherence to guidelines in adult patients with traumatic brain injury: a living systematic review. J Neurotrauma 38:1072\u20131085","journal-title":"J Neurotrauma"}],"container-title":["European Radiology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00330-026-12634-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s00330-026-12634-0","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00330-026-12634-0.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,15]],"date-time":"2026-09-15T13:14:22Z","timestamp":1789478062000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s00330-026-12634-0"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,26]]},"references-count":38,"journal-issue":{"issue":"10","published-online":{"date-parts":[[2026,10]]}},"alternative-id":["12634"],"URL":"https:\/\/doi.org\/10.1007\/s00330-026-12634-0","relation":{},"ISSN":["1432-1084"],"issn-type":[{"value":"1432-1084","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,26]]},"assertion":[{"value":"24 November 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"7 April 2026","order":2,"name":"revised","label":"Revised","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"30 April 2026","order":3,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"26 May 2026","order":4,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The scientific guarantor of this publication is Bastien Le Guellec.","order":1,"name":"Ethics","label":"Guarantor","group":{"name":"EthicsHeading","label":"Compliance with ethical standards"}},{"value":"The authors of this manuscript declare no relationships with any companies whose products or services may be related to the subject matter of the article.","order":2,"name":"Ethics","label":"Conflict of interest","group":{"name":"EthicsHeading","label":"Compliance with ethical standards"}},{"value":"One of the authors has significant statistical expertise.","order":3,"name":"Ethics","label":"Statistics and biometry","group":{"name":"EthicsHeading","label":"Compliance with ethical standards"}},{"value":"Written informed consent was not required for this study because the clinical vignettes are fictional.","order":4,"name":"Ethics","label":"Informed consent","group":{"name":"EthicsHeading","label":"Compliance with ethical standards"}},{"value":"Institutional Review Board approval was not required because no data were collected from real patients.","order":5,"name":"Ethics","label":"Ethical approval","group":{"name":"EthicsHeading","label":"Compliance with ethical standards"}},{"value":"None.","order":6,"name":"Ethics","label":"Study subjects or cohorts overlap","group":{"name":"EthicsHeading","label":"Compliance with ethical standards"}},{"order":7,"name":"Ethics","label":"Methodology","group":{"name":"EthicsHeading","label":"Compliance with ethical standards"}}]}}