{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,25]],"date-time":"2026-08-25T22:54:21Z","timestamp":1787698461014,"version":"build-2784847793"},"reference-count":16,"publisher":"Springer Science and Business Media LLC","issue":"11","license":[{"start":{"date-parts":[[2025,7,15]],"date-time":"2025-07-15T00:00:00Z","timestamp":1752537600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,7,15]],"date-time":"2025-07-15T00:00:00Z","timestamp":1752537600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100000646","name":"Japan Society for the Promotion of Science London","doi-asserted-by":"publisher","award":["24K18793"],"award-info":[{"award-number":["24K18793"]}],"id":[{"id":"10.13039\/501100000646","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100004823","name":"Nagoya University","doi-asserted-by":"crossref","id":[{"id":"10.13039\/501100004823","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Ann Nucl Med"],"published-print":{"date-parts":[[2025,11]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:sec>\n                    <jats:title>Objective<\/jats:title>\n                    <jats:p>Vision language models (VLMs) allow visual input to Large Language Models. VLMs have been developing rapidly, and their accuracy is improving rapidly. Their performance in nuclear medicine compared to state-of-the-art models, including reasoning models, is not yet clear. We evaluated state-of-the-art VLMs using problems from the past Japan Nuclear Medicine Board Examination (JNMBE) and assessed their strengths and limitations.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We collected 180 multiple-choice questions from JNMBE (2022\u20132024). About one-third included diagnostic images. We used eight latest VLMs. ChatGPT o1 pro, ChatGPT o1, ChatGPT o3-mini, ChatGPT-4.5, Claude 3.7, Gemini 2.0 Flash thinking, Llama 3.2, and Gemma 3 were tested. Each model answered every question three times in a deterministic setting, and the final answer was set by majority vote. Two board-certified nuclear medicine physicians independently provided reference answers, with a third expert resolving disagreements. We calculated overall accuracy with 95% confidence intervals and performed subgroup analyses by question type, content, and exam year.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Overall accuracies ranged from 36.1% (Gemma 3) to 83.3% (ChatGPT o1\u00a0pro). ChatGPT o1\u00a0pro achieved the highest score (150\/180, 83.3% [95% CI: 77.1\u201388.5%]), followed by ChatGPT o3-mini (82.8%) and ChatGPTo1 (78.9%). All models performed better on text-only questions than on image-based ones; ChatGPT o1\u00a0pro correctly answered 89.5% of text questions versus 66.0% of image questions. VLMs demonstrated limitations in handling with questions on Japanese regulations. ChatGPT 4.5 excelled in neurology-related image-based questions (76.9%). Accuracy was slightly lower from 2022 to 2024 for most models.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusions<\/jats:title>\n                    <jats:p>VLMs demonstrated high accuracy on the JNMBE, especially on text-based questions, but exhibited limitations with image recognition questions. These findings show that VLMs can be a good assistant for text-based questions in medical domains but have limitations when it comes to comprehensive questions that include images. Currently, VLMs cannot replace comprehensive training and expert interpretation. Because VLMs evolve rapidly and exam difficulty varies annually, these findings should be interpreted in that context.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.1007\/s12149-025-02084-x","type":"journal-article","created":{"date-parts":[[2025,7,15]],"date-time":"2025-07-15T05:08:23Z","timestamp":1752556103000},"page":"1258-1266","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":2,"title":["Vision-language model performance on the Japanese Nuclear Medicine Board Examination: high accuracy in text but challenges with image interpretation"],"prefix":"10.1007","volume":"39","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7971-043X","authenticated-orcid":false,"given":"Rintaro","family":"Ito","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Keita","family":"Kato","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marina","family":"Higashi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yumi","family":"Abe","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ryogo","family":"Minamimoto","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Katsuhiko","family":"Kato","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Toshiaki","family":"Taoka","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shinji","family":"Naganawa","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2025,7,15]]},"reference":[{"key":"2084_CR1","unstructured":"OpenAI, Jaech A, Kalai A, Lerer A, Richardson A, El-Kishky A, et al. OpenAI (2024) o1 System Card [Internet]. arXiv [cs.AI]. Available from: http:\/\/arxiv.org\/abs\/2412.16720"},{"key":"2084_CR2","unstructured":"Open AI (2025) OpenAI o3-mini System Card [Internet]. [cited Apr 8]. Available from: https:\/\/cdn.openai.com\/o3-mini-system-card-feb10.pdf"},{"key":"2084_CR3","unstructured":"Open AI. OpenAI GPT-4.5 System Card [Internet]. [cited 2025 Apr 8]. Available from: https:\/\/cdn.openai.com\/gpt-4-5-system-card-2272025.pdf"},{"key":"2084_CR4","unstructured":"Claude 3.7 Sonnet and Claude Code. Anthropic News February. 24"},{"key":"2084_CR5","unstructured":"Deepmind G. Gemini 2.0 Flash Thinking.\u201d Google DeepMind Technologies Page"},{"key":"2084_CR6","unstructured":"Gemma 3 model card [Internet]. Google AI for Developers. [cited 2025 Apr 8]. Available from: https:\/\/ai.google.dev\/gemma\/docs\/core\/model_card_3"},{"key":"2084_CR7","unstructured":"models\/llama3_2\/MODEL_CARD_VISION.md at main \u00b7 meta-llama\/llama-models [Internet]. Github; [cited 2025 Apr 8]. Available from: https:\/\/github.com\/meta-llama\/llama-models\/blob\/main\/models\/llama3_2\/MODEL_CARD_VISION.md"},{"key":"2084_CR8","doi-asserted-by":"publisher","first-page":"1930","DOI":"10.1038\/s41591-023-02448-8","volume":"29","author":"AJ Thirunavukarasu","year":"2023","unstructured":"Thirunavukarasu AJ, Ting DSJ, Elangovan K, Gutierrez L, Tan TF, Ting DSW. Large language models in medicine. Nat Med. 2023;29:1930\u201340. https:\/\/doi.org\/10.1038\/s41591-023-02448-8.","journal-title":"Nat Med"},{"key":"2084_CR9","unstructured":"Dave T, Athaluri SA, Singh S. ChatGPT in medicine: a systematic review. J Med Syst"},{"key":"2084_CR10","doi-asserted-by":"publisher","first-page":"e230582","DOI":"10.1148\/radiol.230582","volume":"307","author":"R Bhayana","year":"2023","unstructured":"Bhayana R, Krishna S, Bleakney RR. Performance of ChatGPT on a radiology board-style examination: insights into current strengths and limitations. Radiology. 2023;307:e230582. https:\/\/doi.org\/10.1148\/radiol.230582.","journal-title":"Radiology"},{"key":"2084_CR11","doi-asserted-by":"publisher","first-page":"e220281","DOI":"10.1148\/ryai.220281","volume":"5","author":"Z Huemann","year":"2023","unstructured":"Huemann Z, Lee C, Hu J, Cho SY, Bradshaw TJ. Domain-adapted large language models for classifying Nuclear Medicine reports. Radiol Artif Intell. 2023;5:e220281. https:\/\/doi.org\/10.1148\/ryai.220281.","journal-title":"Radiol Artif Intell"},{"key":"2084_CR12","doi-asserted-by":"publisher","first-page":"259","DOI":"10.1038\/s41586-023-05881-4","volume":"616","author":"M Moor","year":"2023","unstructured":"Moor M, Banerjee O, Abad ZSH, Krumholz HM, Leskovec J, Topol EJ, et al. Foundation models for generalist medical artificial intelligence. Nature. 2023;616:259\u201365. https:\/\/doi.org\/10.1038\/s41586-023-05881-4.","journal-title":"Nature"},{"key":"2084_CR13","doi-asserted-by":"publisher","first-page":"404","DOI":"10.1007\/s12149-025-02024-9","volume":"39","author":"K Hirata","year":"2025","unstructured":"Hirata K, Matsui Y, Yamada A, Fujioka T, Yanagawa M, Nakaura T, et al. Correction: generative AI and large language models in nuclear medicine: current status and future prospects. Ann Nucl Med. 2025;39:404\u20135. https:\/\/doi.org\/10.1007\/s12149-025-02024-9.","journal-title":"Ann Nucl Med"},{"key":"2084_CR14","doi-asserted-by":"publisher","DOI":"10.1186\/s40658-024-00642-0","author":"R Ito","year":"2024","unstructured":"Ito R, Kato K, Nanataki K, Abe Y, Ogawa H, Minamimoto R, et al. Assessing large language models for Lugano classification of malignant lymphoma in Japanese FDG-PET reports. EJNMMI Rep. 2024. https:\/\/doi.org\/10.1186\/s40658-024-00642-0.","journal-title":"EJNMMI Rep"},{"key":"2084_CR15","doi-asserted-by":"publisher","DOI":"10.1007\/s11604-024-01633-0","author":"T Oura","year":"2024","unstructured":"Oura T, Tatekawa H, Horiuchi D, Matsushita S, Takita H, Atsukawa N, et al. Diagnostic accuracy of vision-language models on Japanese diagnostic radiology, nuclear medicine, and interventional radiology specialty board examinations. Jpn J Radiol. 2024. https:\/\/doi.org\/10.1007\/s11604-024-01633-0.","journal-title":"Jpn J Radiol"},{"key":"2084_CR16","doi-asserted-by":"publisher","first-page":"217","DOI":"10.1007\/s12149-024-01992-8","volume":"39","author":"T Watanabe","year":"2025","unstructured":"Watanabe T, Baba A, Fukuda T, Watanabe K, Woo J, Ojiri H. Role of visual information in multimodal large language model performance: an evaluation using the Japanese nuclear medicine board examination. Ann Nucl Med. 2025;39:217\u201324. https:\/\/doi.org\/10.1007\/s12149-024-01992-8.","journal-title":"Ann Nucl Med"}],"container-title":["Annals of Nuclear Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s12149-025-02084-x.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s12149-025-02084-x\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s12149-025-02084-x.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,27]],"date-time":"2025-10-27T13:30:52Z","timestamp":1761571852000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s12149-025-02084-x"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,7,15]]},"references-count":16,"journal-issue":{"issue":"11","published-print":{"date-parts":[[2025,11]]}},"alternative-id":["2084"],"URL":"https:\/\/doi.org\/10.1007\/s12149-025-02084-x","relation":{},"ISSN":["0914-7187","1864-6433"],"issn-type":[{"value":"0914-7187","type":"print"},{"value":"1864-6433","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,7,15]]},"assertion":[{"value":"16 April 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"4 July 2025","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"15 July 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":"The Department of Innovative BioMedical Visualization, Nagoya University Graduate School of Medicine, was financially supported by CANON Medical Systems. The Department of Integrated Image Information Analysis, Nagoya University Graduate School of Medicine, was financially supported by FUJIFILM Medical Co., Ltd. The authors declare that they have no other conflicts of interest.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}}]}}