{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T01:44:18Z","timestamp":1792028658221,"version":"4.4.0"},"reference-count":23,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,10,14]],"date-time":"2026-10-14T00:00:00Z","timestamp":1791936000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,10,14]],"date-time":"2026-10-14T00:00:00Z","timestamp":1791936000000},"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 Exp"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Objectives<\/jats:title>\n                    <jats:p>Cross-border care, teleradiology, and international referral pathways increasingly require radiology report translation, but isolated mistranslations may alter clinical meaning. We evaluated German-to-English machine translation of radiology reports and correction-based localization of text requiring review.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Materials and methods<\/jats:title>\n                    <jats:p>In this retrospective study, 310 de-identified German radiology reports from June and July 2025 were translated into English with DeepL. Eight radiologists assigned 5-point quality scores and corrected perceived errors. Thirty reports were reviewed by all eight readers, and 280 reports were assigned for single-reader evaluation, yielding 520 evaluations. Automated metrics and large language model judges were tested for report-level identification of translations requiring review; a correction-based segment-level edit-prediction model was evaluated on held-out shared reports.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      Most translations were rated good or excellent (339\/520, 65.2%), but clinically important mistranslations involved diagnosis, anatomy, and radiologic terminology. Agreement was low for ordinal quality scores (Krippendorff \u03b1\u2009=\u20090.078,\n                      <jats:italic>p<\/jats:italic>\n                      \u2009=\u20090.006) but high for expert corrections (pairwise \u201cbilingual evaluation understudy\u201d\u2013BLEU\u2009=\u200983.0,\n                      <jats:italic>p<\/jats:italic>\n                      \u2009&lt;\u20090.001). In the conservative primary screen of 30 shared reports, no automated method discriminated translations requiring review (area under receiver operating characteristic curve [AUC] = 0.51\u20130.65;\n                      <jats:italic>p<\/jats:italic>\n                      \u2009=\u20090.100\u20130.920). The final edit-prediction model achieved a concordance index of 0.697 and an AUC of 0.731 for segments edited by at least six of eight readers.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>Machine-translated radiology reports were often acceptable on average, but translations containing errors requiring expert review persisted. Expert corrections were more reproducible than ordinal quality scores and supported localization of text requiring human review.<\/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                      Cross-border radiology workflows increasingly require report translation, but fluent machine translations may still contain meaning-changing errors requiring expert review for safe care.\n                    <\/jats:p>\n                    <jats:p>\n                      <jats:bold>\n                        <jats:italic>Findings<\/jats:italic>\n                      <\/jats:bold>\n                      Most machine translations were rated good or excellent, yet clinically important mistranslations occurred, and automated report-level metrics did not reliably identify reports needing review.\n                    <\/jats:p>\n                    <jats:p>\n                      <jats:bold>\n                        <jats:italic>Relevance statement<\/jats:italic>\n                      <\/jats:bold>\n                      Machine translation may support cross-border radiology workflows, but clinically important mistranslations in otherwise fluent reports require human-in-the-loop oversight and tools that localize text requiring expert review.\n                    <\/jats:p>\n                  <\/jats:sec>","DOI":"10.1186\/s41747-026-00827-7","type":"journal-article","created":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T00:49:35Z","timestamp":1792025375000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Fluent but not fail-safe: machine-translated radiology reports still need expert review"],"prefix":"10.1186","volume":"10","author":[{"given":"Constantin","family":"Seibold","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Timo","family":"Leichenich","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Patrick","family":"Stein","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Edem","family":"Atsiatorme","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Patrick","family":"Kliesch","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"G\u00f6k\u00e7e A.","family":"Sevim","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peter","family":"Trauth","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexander","family":"Kremer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Arved","family":"Bischoff","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hans-Ulrich","family":"Kauczor","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0189-7070","authenticated-orcid":false,"given":"Matthias A.","family":"Fink","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,10,14]]},"reference":[{"key":"827_CR1","doi-asserted-by":"publisher","first-page":"d296","DOI":"10.1136\/bmj.d296","volume":"342","author":"H Legido-Quigley","year":"2011","unstructured":"Legido-Quigley H, Passarani I, Knai C et al (2011) Cross-border healthcare in the European Union: clarifying patients\u2019 rights. BMJ 342:d296. https:\/\/doi.org\/10.1136\/bmj.d296","journal-title":"BMJ"},{"key":"827_CR2","doi-asserted-by":"publisher","first-page":"463","DOI":"10.1007\/s13244-016-0485-6","volume":"7","author":"European Society of Radiology (ESR)","year":"2016","unstructured":"European Society of Radiology (ESR) (2016) ESR teleradiology survey: results. Insights Imaging 7:463\u2013479. https:\/\/doi.org\/10.1007\/s13244-016-0485-6","journal-title":"Insights Imaging"},{"key":"827_CR3","doi-asserted-by":"publisher","first-page":"g7392","DOI":"10.1136\/bmj.g7392","volume":"349","author":"S Patil","year":"2014","unstructured":"Patil S, Davies P (2014) Use of Google Translate in medical communication: evaluation of accuracy. BMJ 349:g7392. https:\/\/doi.org\/10.1136\/bmj.g7392","journal-title":"BMJ"},{"key":"827_CR4","doi-asserted-by":"publisher","unstructured":"Jiao W, Wang W, Huang JT, Wang X, Shi S, Tu Z (2023) Is ChatGPT a good translator? Yes with GPT-4 as the engine. Preprint at https:\/\/doi.org\/10.48550\/arXiv.2301.08745","DOI":"10.48550\/arXiv.2301.08745"},{"key":"827_CR5","doi-asserted-by":"publisher","unstructured":"Zhu W, Liu H, Dong Q et al (2024) Multilingual machine translation with large language models: empirical results and analysis. In: Findings of the association for computational linguistics: NAACL 2024. Association for Computational Linguistics, pp 2765\u20132781. https:\/\/doi.org\/10.18653\/v1\/2024.findings-naacl.176","DOI":"10.18653\/v1\/2024.findings-naacl.176"},{"key":"827_CR6","doi-asserted-by":"publisher","first-page":"56","DOI":"10.1016\/j.jbi.2018.07.018","volume":"85","author":"KN Dew","year":"2018","unstructured":"Dew KN, Turner AM, Choi YK, Bosold A, Kirchhoff K (2018) Development of machine translation technology for assisting health communication: a systematic review. J Biomed Inform 85:56\u201367. https:\/\/doi.org\/10.1016\/j.jbi.2018.07.018","journal-title":"J Biomed Inform"},{"key":"827_CR7","doi-asserted-by":"publisher","first-page":"14","DOI":"10.1148\/radiol.10100252","volume":"257","author":"AA Bankier","year":"2010","unstructured":"Bankier AA, Levine D, Halpern EF, Kressel HY (2010) Consensus interpretation in imaging research: Is there a better way? Radiology 257:14\u201317. https:\/\/doi.org\/10.1148\/radiol.10100252","journal-title":"Radiology"},{"key":"827_CR8","doi-asserted-by":"publisher","first-page":"2817","DOI":"10.1007\/s00330-023-10213-1","volume":"34","author":"K Jeblick","year":"2024","unstructured":"Jeblick K, Schachtner B, Dexl J et al (2024) ChatGPT makes medicine easy to swallow: an exploratory case study on simplified radiology reports. Eur Radiol 34:2817\u20132825. https:\/\/doi.org\/10.1007\/s00330-023-10213-1","journal-title":"Eur Radiol"},{"key":"827_CR9","doi-asserted-by":"publisher","first-page":"124","DOI":"10.1186\/s41747-024-00526-1","volume":"8","author":"R Maroncelli","year":"2024","unstructured":"Maroncelli R, Rizzo V, Pasculli M et al (2024) Probing clarity: AI-generated simplified breast imaging reports for enhanced patient comprehension powered by ChatGPT-4o. Eur Radiol Exp 8:124. https:\/\/doi.org\/10.1186\/s41747-024-00526-1","journal-title":"Eur Radiol Exp"},{"key":"827_CR10","doi-asserted-by":"publisher","first-page":"61","DOI":"10.1186\/s41747-025-00600-2","volume":"9","author":"J Wihl","year":"2025","unstructured":"Wihl J, Rosenkranz E, Schramm S et al (2025) Data extraction from free-text stroke CT reports using GPT-4o and Llama-3.3-70B: the impact of annotation guidelines. Eur Radiol Exp 9:61. https:\/\/doi.org\/10.1186\/s41747-025-00600-2","journal-title":"Eur Radiol Exp"},{"key":"827_CR11","doi-asserted-by":"publisher","first-page":"24","DOI":"10.1186\/s41747-025-00569-y","volume":"9","author":"J Poto\u010dnik","year":"2025","unstructured":"Poto\u010dnik J, Thomas E, Kearney D, Killeen RP, Heffernan EJ, Foley SJ (2025) Can ChatGPT and Gemini justify brain CT referrals? A comparative study with human experts and a custom prediction model. Eur Radiol Exp 9:24. https:\/\/doi.org\/10.1186\/s41747-025-00569-y","journal-title":"Eur Radiol Exp"},{"key":"827_CR12","doi-asserted-by":"publisher","first-page":"7","DOI":"10.1186\/s41747-024-00543-0","volume":"9","author":"I Chouvarda","year":"2025","unstructured":"Chouvarda I, Colantonio S, Verde ASC et al (2025) Differences in technical and clinical perspectives on AI validation in cancer imaging: mind the gap!. Eur Radiol Exp 9:7. https:\/\/doi.org\/10.1186\/s41747-024-00543-0","journal-title":"Eur Radiol Exp"},{"key":"827_CR13","doi-asserted-by":"publisher","unstructured":"Papineni K, Roukos S, Ward T, Zhu WJ (2002) BLEU: a method for automatic evaluation of machine translation. In: Proceedings of the 40th annual meeting of the Association for Computational Linguistics. Association for Computational Linguistics, pp 311\u2013318. https:\/\/doi.org\/10.3115\/1073083.1073135","DOI":"10.3115\/1073083.1073135"},{"key":"827_CR14","doi-asserted-by":"publisher","unstructured":"Popovi\u0107 M (2017) chrF++: words helping character n-grams. In: Proceedings of the second conference on machine translation. Association for Computational Linguistics, pp 612\u2013618. https:\/\/doi.org\/10.18653\/v1\/W17-4770","DOI":"10.18653\/v1\/W17-4770"},{"key":"827_CR15","unstructured":"Zhang T, Kishore V, Wu F, Weinberger KQ, Artzi Y (2020) BERTScore: evaluating text generation with BERT. In: International conference on learning representations. ICLR, 2020"},{"key":"827_CR16","doi-asserted-by":"publisher","unstructured":"Rei R, Stewart C, Farinha AC, Lavie A (2020) COMET: a neural framework for MT evaluation. In: Proceedings of the 2020 conference on empirical methods in natural language processing (EMNLP). Association for Computational Linguistics, pp 2685\u20132702. https:\/\/doi.org\/10.18653\/v1\/2020.emnlp-main.213","DOI":"10.18653\/v1\/2020.emnlp-main.213"},{"key":"827_CR17","doi-asserted-by":"publisher","unstructured":"Freitag M, Rei R, Mathur N et al (2022) Results of WMT22 metrics shared task: stop using BLEU\u2014neural metrics are better and more robust. In: Proceedings of the seventh conference on machine translation (WMT). Association for Computational Linguistics, pp 46\u201368. https:\/\/doi.org\/10.18653\/v1\/2022.wmt-1.2","DOI":"10.18653\/v1\/2022.wmt-1.2"},{"key":"827_CR18","doi-asserted-by":"publisher","unstructured":"Rei R, Treviso M, Guerreiro NM et al (2022) CometKiwi: IST-Unbabel 2022 submission for the quality estimation shared task. In: Proceedings of the seventh conference on machine translation (WMT). Association for Computational Linguistics, pp 634\u2013645. https:\/\/doi.org\/10.18653\/v1\/2022.wmt-1.60","DOI":"10.18653\/v1\/2022.wmt-1.60"},{"key":"827_CR19","doi-asserted-by":"publisher","unstructured":"Zerva C, Blain F, Rei R et al (2022) Findings of the WMT 2022 shared task on quality estimation. In: Proceedings of the seventh conference on machine translation (WMT). Association for Computational Linguistics, pp 69\u201399. https:\/\/doi.org\/10.18653\/v1\/2022.wmt-1.3","DOI":"10.18653\/v1\/2022.wmt-1.3"},{"key":"827_CR20","unstructured":"Graham Y, Baldwin T, Moffat A, Zobel J (2013) Continuous measurement scales in human evaluation of machine translation. In: Proceedings of the 7th linguistic annotation workshop and interoperability with discourse. Association for Computational Linguistics, pp 33\u201341"},{"key":"827_CR21","doi-asserted-by":"publisher","DOI":"10.1016\/j.media.2023.102798","volume":"86","author":"S Yang","year":"2023","unstructured":"Yang S, Wu X, Ge S, Zheng Z, Zhou SK, Xiao L (2023) Radiology report generation with a learned knowledge base and multi-modal alignment. Med Image Anal 86:102798. https:\/\/doi.org\/10.1016\/j.media.2023.102798","journal-title":"Med Image Anal"},{"key":"827_CR22","doi-asserted-by":"publisher","DOI":"10.1148\/radiol.230877","volume":"307","author":"RJ Gertz","year":"2023","unstructured":"Gertz RJ, Bunck AC, Lennartz S et al (2023) GPT-4 for automated determination of radiological study and protocol based on radiology request forms: a feasibility study. Radiology 307:e230877. https:\/\/doi.org\/10.1148\/radiol.230877","journal-title":"Radiology"},{"key":"827_CR23","doi-asserted-by":"publisher","first-page":"16","DOI":"10.1186\/s41747-026-00682-6","volume":"10","author":"M Zanardo","year":"2026","unstructured":"Zanardo M, Albano D, Molinari V et al (2026) Can AI write reports like a radiologist? A blinded evaluation of large language model-generated lumbar spine MRI reports. Eur Radiol Exp 10:16. https:\/\/doi.org\/10.1186\/s41747-026-00682-6","journal-title":"Eur Radiol Exp"}],"container-title":["European Radiology Experimental"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s41747-026-00827-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s41747-026-00827-7","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s41747-026-00827-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T00:49:37Z","timestamp":1792025377000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s41747-026-00827-7"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,14]]},"references-count":23,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2026,12]]}},"alternative-id":["827"],"URL":"https:\/\/doi.org\/10.1186\/s41747-026-00827-7","relation":{},"ISSN":["2509-9280"],"issn-type":[{"value":"2509-9280","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,14]]},"assertion":[{"value":"14 May 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"26 August 2026","order":2,"name":"revised","label":"Revised","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"5 September 2026","order":3,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"14 October 2026","order":4,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"This retrospective study used de-identified radiology reports and was approved by the institutional review board of University Hospital Heidelberg (S-083\/2018). Informed consent was waived because the study was retrospective and all reports were de-identified before analysis.","order":1,"name":"Ethics","label":"Ethics approval and consent to participate","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Not applicable. This manuscript does not contain identifiable person data, images, or videos. All radiology report excerpts were de-identified before analysis and presentation.","order":2,"name":"Ethics","label":"Consent for publication","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare that they have no competing interests.","order":3,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Declarations"}}],"article-number":"158"}}