{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,14]],"date-time":"2026-07-14T21:02:21Z","timestamp":1784062941505,"version":"3.55.0"},"reference-count":11,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2025,8,18]],"date-time":"2025-08-18T00:00:00Z","timestamp":1755475200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,8,18]],"date-time":"2025-08-18T00:00:00Z","timestamp":1755475200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"Universit\u00e4tsklinikum D\u00fcsseldorf. Anstalt \u00f6ffentlichen Rechts"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Clin Neuroradiol"],"published-print":{"date-parts":[[2026,3]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Selection of appropriate imaging sequences protocols for cranial magnetic resonance imaging (MRI) is crucial to address the medical question and adequately support patient care. Inappropriate protocol selection can compromise diagnostic accuracy, extend scan duration, and increase the risk of misdiagnosis. Typically, radiologists determine scanning protocols based on their expertise, a\u00a0process that can be time-consuming and subject to variability. Language models offer the potential to streamline this process. This study investigates the capability of bidirectional encoder representations from transformers (BERT)-based models to suggest appropriate MRI protocols based on referral information.<\/jats:p>\n                  <jats:p>A\u00a0total of 410 anonymized electronic referrals for cranial MRI from a\u00a0local order-entry system were categorized into nine protocol classes by an experienced neuroradiologist. A\u00a0locally hosted instance of four different, pre-trained BERT-based classifiers (BERT, ModernBERT, GottBERT, and medBERT.de) were trained to classify protocols based on referral entries, including preliminary diagnoses, prior treatment history, and clinical questions. Each model was additionally fine-tuned for local language on a\u00a0large dataset of electronic referrals.<\/jats:p>\n                  <jats:p>The model based on medBERT.de with local language fine-tuning was the best-performing model and correctly predicted 81% of all protocols, achieving a\u00a0macro-F1 score of 0.71, macro-precision and macro-recall values of 0.73 and 0.71, respectively. Moreover, we were able to show that local language fine-tuning led to performance improvements across all models.<\/jats:p>\n                  <jats:p>These results demonstrate the potential of language models to predict MRI protocols, even with limited training data. This approach could accelerate and standardize radiological protocol selection, offering significant benefits for clinical workflows.<\/jats:p>","DOI":"10.1007\/s00062-025-01554-z","type":"journal-article","created":{"date-parts":[[2025,8,18]],"date-time":"2025-08-18T14:16:45Z","timestamp":1755526605000},"page":"59-66","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Automated Protocol Suggestions for Cranial MRI Examinations Using Locally Fine-tuned BERT Models"],"prefix":"10.1007","volume":"36","author":[{"ORCID":"https:\/\/orcid.org\/0009-0004-5535-4073","authenticated-orcid":false,"given":"Christian","family":"Boschenriedter","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Christian","family":"Rubbert","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marius","family":"Vach","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Julian","family":"Caspers","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2025,8,18]]},"reference":[{"key":"1554_CR1","doi-asserted-by":"publisher","DOI":"10.1007\/s00330-025-11500-9","author":"S Agarwal","year":"2025","unstructured":"Agarwal\u00a0S, Wood\u00a0D, Murray\u00a0BAK, Wei\u00a0Y, Al Busaidi\u00a0A, Kafiabadi\u00a0S, Guilhem\u00a0E, Lynch\u00a0J, Townend\u00a0M, Mazumder\u00a0A, Barker\u00a0GJ, Cole\u00a0JH, Sasieni\u00a0P, Ourselin\u00a0S, Modat\u00a0M, Booth\u00a0TC. Impact of hospital-specific domain adaptation on BERT-based models to classify neuroradiology reports. Eur Radiol. 2025; https:\/\/doi.org\/10.1007\/s00330-025-11500-9.","journal-title":"Eur Radiol"},{"key":"1554_CR2","doi-asserted-by":"publisher","first-page":"121598","DOI":"10.1016\/j.eswa.2023.121598","volume":"237","author":"KK Bressem","year":"2024","unstructured":"Bressem\u00a0KK, Papaioannou\u00a0J\u2011M, Grundmann\u00a0P, Borchert\u00a0F, Adams\u00a0LC, Liu\u00a0L, Busch\u00a0F, Xu\u00a0L, Loyen\u00a0JP, Niehues\u00a0SM, Augustin\u00a0M, Grosser\u00a0L, Makowski\u00a0MR, Aerts\u00a0HJWL, L\u00f6ser\u00a0A. medBERT.de: a\u00a0comprehensive German BERT model for the medical domain. Expert Syst Appl. 2024;237:121598. https:\/\/doi.org\/10.1016\/j.eswa.2023.121598.","journal-title":"Expert Syst Appl"},{"issue":"5","key":"1554_CR3","doi-asserted-by":"publisher","first-page":"1293","DOI":"10.1007\/s10278-022-00610-1","volume":"35","author":"J Denck","year":"2022","unstructured":"Denck\u00a0J, Haas\u00a0O, Guehring\u00a0J, Maier\u00a0A, Rothgang\u00a0E. Automated protocoling for MRI exams\u2014challenges and solutions. J\u00a0Digit Imaging. 2022;35(5):1293\u2013302. https:\/\/doi.org\/10.1007\/s10278-022-00610-1.","journal-title":"J Digit Imaging"},{"key":"1554_CR4","volume-title":"BERT: pre-training of deep bidirectional transformers for language understanding","author":"J Devlin","year":"2018","unstructured":"Devlin\u00a0J, Chang\u00a0M\u2011W, Lee\u00a0K, Toutanova\u00a0K. BERT: pre-training of deep bidirectional transformers for language understanding. 2018."},{"key":"1554_CR5","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.230877","author":"RJ Gertz","year":"2023","unstructured":"Gertz\u00a0RJ, Bunck\u00a0AC, Lennartz\u00a0S, Dratsch\u00a0T, Iuga\u00a0A\u2011I, Maintz\u00a0D, Kottlors\u00a0J. GPT\u20114 for automated determination of radiologic study and protocol based on radiology request forms: a\u00a0feasibility study. Radiology. 2023; https:\/\/doi.org\/10.1148\/radiol.230877.","journal-title":"Radiology"},{"issue":"5","key":"1554_CR6","doi-asserted-by":"publisher","first-page":"545","DOI":"10.1016\/j.jacr.2015.08.027","volume":"13","author":"DT Ginat","year":"2016","unstructured":"Ginat\u00a0DT, Uppuluri\u00a0P, Christoforidis\u00a0G, Katzman\u00a0G, Lee\u00a0S\u2011K. Identification of neuroradiology MRI protocol errors via a\u00a0quality-driven categorization approach. J\u00a0Am Coll Radiol. 2016;13(5):545\u20138. https:\/\/doi.org\/10.1016\/j.jacr.2015.08.027.","journal-title":"J Am Coll Radiol"},{"key":"1554_CR7","doi-asserted-by":"publisher","DOI":"10.1148\/ryai.230620","author":"HJ Huhtanen","year":"2025","unstructured":"Huhtanen\u00a0HJ, Nyman\u00a0MJ, Karlsson\u00a0A, Hirvonen\u00a0J. Machine learning and deep learning models for automated protocoling of emergency brain MRI using text from clinical referrals. Radiol Artif Intell. 2025; https:\/\/doi.org\/10.1148\/ryai.230620.","journal-title":"Radiol Artif Intell"},{"issue":"5","key":"1554_CR8","doi-asserted-by":"publisher","first-page":"604","DOI":"10.1007\/s10278-018-0066-y","volume":"31","author":"YH Lee","year":"2018","unstructured":"Lee\u00a0YH. Efficiency improvement in a\u00a0busy radiology practice: determination of musculoskeletal magnetic resonance imaging protocol using deep-learning Convolutional neural networks. J\u00a0Digit Imaging. 2018;31(5):604\u201310. https:\/\/doi.org\/10.1007\/s10278-018-0066-y.","journal-title":"J Digit Imaging"},{"key":"1554_CR9","volume-title":"GottBERT: a\u00a0pure German language model","author":"R Scheible","year":"2020","unstructured":"Scheible\u00a0R, Thomczyk\u00a0F, Tippmann\u00a0P, Jaravine\u00a0V, Boeker\u00a0M. GottBERT: a\u00a0pure German language model. 2020."},{"key":"1554_CR10","volume-title":"Smarter, better, faster, longer: a\u00a0modern bidirectional encoder for fast, memory efficient, and long context finetuning and inference","author":"B Warner","year":"2024","unstructured":"Warner\u00a0B, Chaffin\u00a0A, Clavi\u00e9\u00a0B, Weller\u00a0O, Hallstr\u00f6m\u00a0O, Taghadouini\u00a0S, Gallagher\u00a0A, Biswas\u00a0R, Ladhak\u00a0F, Aarsen\u00a0T, Cooper\u00a0N, Adams\u00a0G, Howard\u00a0J, Poli\u00a0I. Smarter, better, faster, longer: a\u00a0modern bidirectional encoder for fast, memory efficient, and long context finetuning and inference. 2024."},{"key":"1554_CR11","unstructured":"Zheng J, Hong H, Liu F, Wang X, Su J, Liang Y, Wu S. Fine-tuning large language models for domain-specific machine translation. 2024;."}],"container-title":["Clinical Neuroradiology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00062-025-01554-z.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s00062-025-01554-z","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00062-025-01554-z.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,23]],"date-time":"2026-03-23T10:08:21Z","timestamp":1774260501000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s00062-025-01554-z"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,18]]},"references-count":11,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2026,3]]}},"alternative-id":["1554"],"URL":"https:\/\/doi.org\/10.1007\/s00062-025-01554-z","relation":{},"ISSN":["1869-1439","1869-1447"],"issn-type":[{"value":"1869-1439","type":"print"},{"value":"1869-1447","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,8,18]]},"assertion":[{"value":"1 June 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"28 July 2025","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"18 August 2025","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"C.\u00a0Boschenriedter, C.\u00a0Rubbert, M.\u00a0Vach and J.\u00a0Caspers declare that they have no competing interests.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}},{"value":"All human and animal studies have been approved by the appropriate ethics committee and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical standards"}}]}}