{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,11,7]],"date-time":"2026-11-07T01:25:53Z","timestamp":1794014753787,"version":"4.8.0"},"reference-count":70,"publisher":"Springer Science and Business Media LLC","issue":"10","license":[{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T00:00:00Z","timestamp":1791244800000},"content-version":"vor","delay-in-days":5,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Nat Med"],"published-print":{"date-parts":[[2026,10]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Artificial intelligence has high potential for impact in healthcare applications, but its training and deployment are challenging due to diverse data, a complex spectrum of possible tasks and important privacy needs. High-performing foundation models that enable data-efficient fine-tuning for diverse downstream tasks can meaningfully accelerate development in this domain. We introduce MedGemma, a collection of medical vision-language foundation models based on Gemma 3. MedGemma demonstrates advanced medical understanding and reasoning across images and text and multiple medical imaging domains, exceeding the performance of similarly sized generative models while maintaining the general capabilities of the Gemma base models. For out-of-distribution tasks, MedGemma achieves improvements of 2.6\u201310% in medical image question answering, 15.5\u201318.1% in chest X-ray finding classification and 10.8% in agentic evaluations compared with the base models. Our results show that fine-tuning MedGemma can be more effective than fine-tuning the base Gemma 3 model for medical tasks, particularly in the setting of limited training data. We additionally introduce MedSigLIP, a medically tuned vision encoder derived from SigLIP. MedSigLIP powers the visual understanding capabilities of MedGemma and, as an encoder, achieves performance comparable to or better than that of many specialized medical image encoders. Taken together, the MedGemma collection provides a strong foundation of medical image and text capabilities, with the potential to accelerate medical research and the development of downstream applications.<\/jats:p>","DOI":"10.1038\/s41591-026-04626-w","type":"journal-article","created":{"date-parts":[[2026,10,6]],"date-time":"2026-10-06T09:02:00Z","timestamp":1791277320000},"page":"3695-3704","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":2,"title":["An open vision-language model for diverse medical applications"],"prefix":"10.1038","volume":"32","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5260-8013","authenticated-orcid":false,"given":"Andrew","family":"Sellergren","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sahar","family":"Kazemzadeh","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fereshteh","family":"Mahvar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-6613-3581","authenticated-orcid":false,"given":"Atilla","family":"Kiraly","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Madeleine","family":"Traverse","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Timo","family":"Kohlberger","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chufan","family":"Gao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ronnachai","family":"Jaroensri","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shawn","family":"Xu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fayaz","family":"Jamil","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"C\u00edan","family":"Hughes","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3136-9711","authenticated-orcid":false,"given":"Charles","family":"Lau","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-0218-1867","authenticated-orcid":false,"given":"Justin","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liron","family":"Yatziv","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tiffany","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bram","family":"Sterling","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Stefanie Anna","family":"Baby","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Susanna Maria","family":"Baby","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jeremy","family":"Lai","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Samuel","family":"Schmidgall","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lu","family":"Yang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kejia","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Per","family":"Bjornsson","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shashir","family":"Reddy","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ryan","family":"Brush","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kenneth","family":"Philbrick","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bashir","family":"Sadjad","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mercy","family":"Asiedu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ines","family":"Mezerreg","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Howard","family":"Hu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Howard","family":"Yang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Richa","family":"Tiwari","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sunny","family":"Jansen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Preeti","family":"Singh","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1902-3822","authenticated-orcid":false,"given":"Yubin","family":"Kim","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7313-0082","authenticated-orcid":false,"given":"Daniel","family":"McDuff","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Catherine","family":"Kozlowski","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rachelle","family":"Sico","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aishwarya","family":"Kamath","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Johan","family":"Ferret","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shreya","family":"Pathak","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nino","family":"Vieillard","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ramona","family":"Merhej","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sarah","family":"Perrin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tatiana","family":"Matejovicova","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexandre","family":"Ram\u00e9","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Morgane","family":"Riviere","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Louis","family":"Rouillard","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Thomas","family":"Mesnard","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Geoffrey","family":"Cideron","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jean-bastien","family":"Grill","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6656-9732","authenticated-orcid":false,"given":"Sabela","family":"Ramos","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Edouard","family":"Yvinec","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Michelle","family":"Casbon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Elena","family":"Buchatskaya","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jean-Baptiste","family":"Alayrac","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dmitry","family":"Lepikhin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Vlad","family":"Feinberg","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sebastian","family":"Borgeaud","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alek","family":"Andreev","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Cassidy","family":"Hardin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Robert","family":"Dadashi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"L\u00e9onard","family":"Hussenot","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Armand","family":"Joulin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Olivier","family":"Bachem","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Clement","family":"Farabet","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joelle","family":"Barral","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tris","family":"Warkentin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jonathon","family":"Shlens","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-0734-7114","authenticated-orcid":false,"given":"David","family":"Fleet","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7447-6031","authenticated-orcid":false,"given":"Shekoofeh","family":"Azizi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Victor","family":"Cotruta","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Omar","family":"Sanseviero","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Gus","family":"Martins","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Phoebe","family":"Kirk","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Anand","family":"Rao","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3960-6002","authenticated-orcid":false,"given":"Yossi","family":"Matias","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Katherine","family":"Chou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Avinatan","family":"Hassidim","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3023-8824","authenticated-orcid":false,"given":"Dale","family":"Webster","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-3783-3172","authenticated-orcid":false,"given":"Shravya","family":"Shetty","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kavi","family":"Goel","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4079-8275","authenticated-orcid":false,"given":"Yun","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Can","family":"Kirmizibayrak","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1297-0023","authenticated-orcid":false,"given":"David F.","family":"Steiner","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Rory","family":"Pilgrim","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1114-3184","authenticated-orcid":false,"given":"Daniel","family":"Golden","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7593-7066","authenticated-orcid":false,"given":"Lin","family":"Yang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,10,6]]},"reference":[{"key":"4626_CR1","doi-asserted-by":"crossref","unstructured":"AlayracJ.-B. 2022. Flamingo: a visual language model for few-shot learning. In Proc. 36th International Conference on Neural Information Processing Systems (eds Koyejo, S. et al.) 23716\u201323736 (Curran Associates Inc., 2022).","DOI":"10.52202\/068431-1723"},{"key":"4626_CR2","doi-asserted-by":"publisher","unstructured":"Achiam, J. et al. GPT-4 technical report. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2303.08774 (2023).","DOI":"10.48550\/arXiv.2303.08774"},{"key":"4626_CR3","doi-asserted-by":"publisher","unstructured":"Touvron, H. et al. LLaMA: open and efficient foundation language models. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2302.13971 (2023).","DOI":"10.48550\/arXiv.2302.13971"},{"key":"4626_CR4","doi-asserted-by":"publisher","unstructured":"Toma, A. et al. Clinical Camel: an open-source expert-level medical language model with dialogue-based knowledge encoding. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2305.12031 (2023).","DOI":"10.48550\/arXiv.2305.12031"},{"key":"4626_CR5","doi-asserted-by":"publisher","unstructured":"Han, T. et al. MedAlpaca\u2013an open-source collection of medical conversational AI models and training data. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2304.08247 (2023).","DOI":"10.48550\/arXiv.2304.08247"},{"key":"4626_CR6","doi-asserted-by":"crossref","unstructured":"Labrak, Y. et al. BioMistral: a collection of open-source pretrained large language models for medical domains. In Findings of the Association for Computational Linguistics (eds Ku, L. -W., Martins, A. & Srikumar, V.) 5848\u20135864 (2024).","DOI":"10.18653\/v1\/2024.findings-acl.348"},{"key":"4626_CR7","doi-asserted-by":"publisher","unstructured":"Gemma-Team et al. Gemma 3 technical report. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2503.19786 (2025).","DOI":"10.48550\/arXiv.2503.19786"},{"key":"4626_CR8","doi-asserted-by":"publisher","unstructured":"Saab, K. et al. Capabilities of Gemini models in medicine. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2404.18416 (2024).","DOI":"10.48550\/arXiv.2404.18416"},{"key":"4626_CR9","doi-asserted-by":"publisher","unstructured":"Yang, L. et al. Advancing multimodal medical capabilities of Gemini. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2405.03162 (2024).","DOI":"10.48550\/arXiv.2405.03162"},{"key":"4626_CR10","doi-asserted-by":"publisher","DOI":"10.3390\/app11146421","volume":"11","author":"D Jin","year":"2021","unstructured":"Jin, D. et al. What disease does this patient have? A large-scale open domain question answering dataset from medical exams. Appl. Sci. 11, 6421 (2021).","journal-title":"Appl. Sci."},{"key":"4626_CR11","unstructured":"Pal, A., Umapathi, L. K. & Sankarasubbu, M. MedMCQA: a large-scale multi-subject multi-choice dataset for medical domain question answering. In Proceedings of the Conference on Health, Inference, and Learning (eds Flores, G. et al.) 248\u2013260 (PMLR, 2022)."},{"key":"4626_CR12","doi-asserted-by":"crossref","unstructured":"Jin, Q., Dhingra, B., Liu, Z., Cohen, W. W. & Lu, X. PubMedQA: a dataset for biomedical research question answering. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (eds Inui, K. et al.) 2567\u20132577 (2019).","DOI":"10.18653\/v1\/D19-1259"},{"key":"4626_CR13","unstructured":"Hendrycks, D. et al. Measuring massive multitask language understanding. In International Conference on Learning Representations (2021)."},{"key":"4626_CR14","doi-asserted-by":"crossref","unstructured":"Olatunji, T. et al. AfriMed-QA: a pan-african, multi-specialty, medical question-answering benchmark dataset. In Proc. 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (eds Che, W. et al.) 1948\u20131973 (2024).","DOI":"10.18653\/v1\/2025.acl-long.96"},{"key":"4626_CR15","doi-asserted-by":"publisher","unstructured":"Schmidgall, S. et al. AgentClinic: a multimodal agent benchmark to evaluate AI in simulated clinical environments. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2405.07960 (2024).","DOI":"10.48550\/arXiv.2405.07960"},{"key":"4626_CR16","doi-asserted-by":"crossref","unstructured":"DeepSeek-AI. DeepSeek-R1: incentivizing reasoning capability in LLMs via reinforcement learning. Nature 645, 633\u2013638 (2025).","DOI":"10.1038\/s41586-025-09422-z"},{"key":"4626_CR17","unstructured":"Ankit Pal, M. S. OpenBioLLMs: advancing open-source large language models for healthcare and life sciences https:\/\/huggingface.co\/aaditya\/OpenBioLLM-Llama3-70B (2024)."},{"key":"4626_CR18","doi-asserted-by":"publisher","unstructured":"Warner, B. et al. Medmarks: a comprehensive open-source LLM benchmark suite for medical tasks. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2605.01417 (2026).","DOI":"10.48550\/arXiv.2605.01417"},{"key":"4626_CR19","doi-asserted-by":"crossref","unstructured":"Pandit, S. et al. MedHallu: a comprehensive benchmark for detecting medical hallucinations in large language models. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (eds Christodoulopoulos, C. et al.) 2858\u20132873 (2025).","DOI":"10.18653\/v1\/2025.emnlp-main.143"},{"key":"4626_CR20","doi-asserted-by":"crossref","unstructured":"Liu, B. et al. SLAKE: a semantically-labeled knowledge-enhanced dataset for medical visual question answering. In Proceedings of the 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI) 1650\u20131654 (IEEE, 2021).","DOI":"10.1109\/ISBI48211.2021.9434010"},{"key":"4626_CR21","doi-asserted-by":"publisher","DOI":"10.1038\/sdata.2018.251","volume":"5","author":"JJ Lau","year":"2018","unstructured":"Lau, J. J., Gayen, S., Ben Abacha, A. & Demner-Fushman, D. A dataset of clinically generated visual questions and answers about radiology images. Sci. Data 5, 180251 (2018).","journal-title":"Sci. Data"},{"key":"4626_CR22","doi-asserted-by":"publisher","unstructured":"Xu, S. et al. ELIXR: towards a general purpose X-ray artificial intelligence system through alignment of large language models and radiology vision encoders. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2308.01317 (2023).","DOI":"10.48550\/arXiv.2308.01317"},{"key":"4626_CR23","doi-asserted-by":"publisher","first-page":"28541","DOI":"10.52202\/075280-1240","volume":"36","author":"C Li","year":"2023","unstructured":"Li, C. et al. LLaVA-Med: training a large language-and-vision assistant for biomedicine in one day. Adv. Neural Inf. Process. Syst. 36, 28541\u201328564 (2023).","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"4626_CR24","unstructured":"Zhang, K. et al. BiomedGPT: a unified and generalist biomedical generative pre-trained transformer for vision, language, and multimodal tasks. Preprint at arXiv https:\/\/arxiv.org\/abs\/2305.17100v2 (2024)."},{"key":"4626_CR25","unstructured":"Xie, Y. et al. MedTrinity-25M: a large-scale multimodal dataset with multigranular annotations for medicine. In The Thirteenth International Conference on Learning Representations (2025)."},{"key":"4626_CR26","doi-asserted-by":"crossref","unstructured":"Chen, J. et al. Towards injecting medical visual knowledge into multimodal LLMs at scale. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (eds Al-Onaizan, Y., Bansal, M. & Chen, Y. -N.) 7346\u20137370 (2024).","DOI":"10.18653\/v1\/2024.emnlp-main.418"},{"key":"4626_CR27","doi-asserted-by":"publisher","unstructured":"Wang, J. et al. SemiHVision: enhancing medical multimodal models with a semi-human annotated dataset and fine-tuned instruction generation. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2410.14948 (2024).","DOI":"10.48550\/arXiv.2410.14948"},{"key":"4626_CR28","doi-asserted-by":"publisher","unstructured":"Jain, S. et al. RadGraph: extracting clinical entities and relations from radiology reports. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2106.14463 (2021).","DOI":"10.48550\/arXiv.2106.14463"},{"key":"4626_CR29","doi-asserted-by":"publisher","unstructured":"Ostmeier, S. et al. GREEN: Generative radiology report evaluation and error notation. Findings of the Association for Computational Linguistics: EMNLP 2024 374\u2013390 https:\/\/doi.org\/10.18653\/v1\/2024.findings-emnlp.21 (Association for Computational Linguistics, 2024).","DOI":"10.18653\/v1\/2024.findings-emnlp.21"},{"key":"4626_CR30","unstructured":"Zhang, X., Acosta, J. N., Miller, J., Huang, O. & Rajpurkar, P. ReXGradient-160K: a large-scale publicly available dataset of chest radiographs with free-text reports. In Machine Learning for Health (2025)."},{"key":"4626_CR31","doi-asserted-by":"publisher","first-page":"304","DOI":"10.1093\/jamia\/ocv080","volume":"23","author":"D Demner-Fushman","year":"2016","unstructured":"Demner-Fushman, D. et al. Preparing a collection of radiology examinations for distribution and retrieval. J. Am. Med. Inform. Assoc. 23, 304\u2013310 (2016).","journal-title":"J. Am. Med. Inform. Assoc."},{"key":"4626_CR32","doi-asserted-by":"crossref","unstructured":"Pal, A. & Sankarasubbu, M. Gemini goes to Med school: exploring the capabilities of multimodal large language models on medical challenge problems & hallucinations. In Proc. 6th Clinical Natural Language Processing Workshop (eds Naumann, T. et al.) 21\u201346 (2024).","DOI":"10.18653\/v1\/2024.clinicalnlp-1.3"},{"key":"4626_CR33","doi-asserted-by":"publisher","unstructured":"Wu, J. et al. Chest ImaGenome dataset. PhysioNet. https:\/\/doi.org\/10.13026\/wv01-y230 (2021).","DOI":"10.13026\/wv01-y230"},{"key":"4626_CR34","doi-asserted-by":"crossref","unstructured":"Zun, L. Q., Halim, M. Z. B. A. & Fye, G. M. Fine-tuning MedGemma for clinical captioning to enhance multimodal RAG over Malaysia CPGs. In IEEE 13th Conference on Systems, Process & Control (ICSPC) 129\u2013134 (2025).","DOI":"10.1109\/ICSPC68261.2025.11326535"},{"key":"4626_CR35","unstructured":"Alaa, A. et al. Medical large language model benchmarks should prioritize construct validity. In Forty-second International Conference on Machine Learning Position Paper Track (2025)."},{"key":"4626_CR36","doi-asserted-by":"publisher","DOI":"10.1038\/s41591-025-03637-3","volume":"31","author":"F Mahmood","year":"2025","unstructured":"Mahmood, F. A benchmarking crisis in biomedical machine learning. Nat. Med. 31, 1060 (2025).","journal-title":"Nat. Med."},{"key":"4626_CR37","unstructured":"Bedi, S. et al. MedHELM: holistic evaluation of large language models for medical tasks. Nat. Med. 32, 943\u2013951 (2026)."},{"key":"4626_CR38","doi-asserted-by":"publisher","unstructured":"Kim, Y. et al. Medical hallucinations in foundation models and their impact on healthcare. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2503.05777 (2025).","DOI":"10.48550\/arXiv.2503.05777"},{"key":"4626_CR39","doi-asserted-by":"crossref","unstructured":"Muennighoff, N. et al. s1: Simple test-time scaling. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (eds Christodoulopoulos, C. et al.) 20275\u201320321 (Association for Computational Linguistics, 2025).","DOI":"10.18653\/v1\/2025.emnlp-main.1025"},{"key":"4626_CR40","unstructured":"John Snow Labs. JSL-MedLLaMA-3-8B-v2.0 https:\/\/huggingface.co\/johnsnowlabs\/JSL-MedLlama-3-8B-v2.0 (accessed 11 June 2025)."},{"key":"4626_CR41","unstructured":"IQVIA. Introducing IQVIA Medical Reasoning (Med-R1 8B): Best-in-class medical reasoning LLM. https:\/\/www.iqvia.com\/blogs\/2025\/04\/introducing-iqvia-medical-reasoning-med-r1-8b (accessed 11 June 2025)."},{"key":"4626_CR42","doi-asserted-by":"publisher","DOI":"10.1056\/AIoa2300138","volume":"1","author":"T Tu","year":"2024","unstructured":"Tu, T. et al. Towards generalist biomedical AI. NEJM AI 1, AIoa2300138 (2024).","journal-title":"NEJM AI"},{"key":"4626_CR43","doi-asserted-by":"publisher","unstructured":"Deperrois, N. et al. RadVLM: a multitask conversational vision-language model for radiology. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2502.03333 (2025).","DOI":"10.48550\/arXiv.2502.03333"},{"key":"4626_CR44","unstructured":"Endo, M., Krishnan, R., Krishna, V., Ng, A. Y. & Rajpurkar, P. Retrieval-based chest x-ray report generation using a pre-trained contrastive language-image model. In Proceedings of the Machine Learning for Health (eds Roy, S. et al.) Vol. 158, 209\u2013219 (PMLR, 2021)."},{"key":"4626_CR45","doi-asserted-by":"crossref","unstructured":"Miura, Y., Zhang, Y., Tsai, E. B., Langlotz, C. P. & Jurafsky, D. Improving factual completeness and consistency of image-to-text radiology report generation. In Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (eds Toutanova, K. et al.) 5288\u20135304 (2021).","DOI":"10.18653\/v1\/2021.naacl-main.416"},{"key":"4626_CR46","doi-asserted-by":"publisher","unstructured":"Hyland, S. L. et al. MAIRA-1: a specialised large multimodal model for radiology report generation. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2311.13668 (2023).","DOI":"10.48550\/arXiv.2311.13668"},{"key":"4626_CR47","doi-asserted-by":"publisher","unstructured":"Bannur, S. et al. MAIRA-2: grounded radiology report generation. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2406.04449 (2024).","DOI":"10.48550\/arXiv.2406.04449"},{"key":"4626_CR48","doi-asserted-by":"crossref","unstructured":"Chen, Z., Song, Y., Chang, T.-H. & Wan, X. Generating radiology reports via memory-driven transformer. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (eds Webber, B. et al.) 1439\u20131449 (2020).","DOI":"10.18653\/v1\/2020.emnlp-main.112"},{"key":"4626_CR49","doi-asserted-by":"crossref","unstructured":"Moens, M.-F., Huang, X., Specia, L., Yih, & S. W.-T. (eds). Weakly supervised contrastive learning for chest X-ray report generation. In Findings of the Association for Computational Linguistics: EMNLP 2021 4009\u20134015 (Association for Computational Linguistics, 2021).","DOI":"10.18653\/v1\/2021.findings-emnlp.336"},{"key":"4626_CR50","doi-asserted-by":"publisher","first-page":"102633","DOI":"10.1016\/j.artmed.2023.102633","volume":"144","author":"A Nicolson","year":"2023","unstructured":"Nicolson, A., Dowling, J. & Koopman, B. Improving chest x-ray report generation by leveraging warm starting. Artif. Intell. Med. 144, 102633 (2023).","journal-title":"Artif. Intell. Med."},{"key":"4626_CR51","doi-asserted-by":"publisher","first-page":"599","DOI":"10.1038\/s41591-024-03302-1","volume":"31","author":"R Tanno","year":"2025","unstructured":"Tanno, R. et al. Collaboration between clinicians and vision\u2013language models in radiology report generation. Nat. Med. 31, 599\u2013608 (2025).","journal-title":"Nat. Med."},{"key":"4626_CR52","doi-asserted-by":"publisher","unstructured":"Steiner, A. et al. PaliGemma 2: a family of versatile VLMs for transfer. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2412.03555 (2024).","DOI":"10.48550\/arXiv.2412.03555"},{"key":"4626_CR53","first-page":"AIoa2500595","volume":"3","author":"H-Y Zhou","year":"2026","unstructured":"Zhou, H.-Y., Adithan, S., Acosta, J. N., Topol, E. J. & Rajpurkar, P. MedVersa: a generalist learner for multifaceted medical image interpretation. NEJM AI 3, AIoa2500595 (2026).","journal-title":"NEJM AI"},{"key":"4626_CR54","doi-asserted-by":"crossref","unstructured":"Wang, X. et al. ChestX-ray8: hospital-scale chest X-ray database and benchmarks on weakly-supervised classification and localization of common thorax diseases. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition 2097\u20132106 (IEEE, 2017).","DOI":"10.1109\/CVPR.2017.369"},{"key":"4626_CR55","doi-asserted-by":"publisher","DOI":"10.1001\/jamanetworkopen.2024.46615","volume":"7","author":"A Ward","year":"2024","unstructured":"Ward, A. et al. Creating an empirical dermatology dataset through crowdsourcing with web search advertisements. JAMA Netw. Open 7, e2446615 (2024).","journal-title":"JAMA Netw. Open"},{"key":"4626_CR56","doi-asserted-by":"publisher","first-page":"61","DOI":"10.1038\/nature11412","volume":"490","author":"Cancer Genome Atlas Network","year":"2012","unstructured":"Cancer Genome Atlas Network Comprehensive molecular portraits of human breast tumours. Nature 490, 61\u201370 (2012).","journal-title":"Nature"},{"key":"4626_CR57","doi-asserted-by":"publisher","DOI":"10.1016\/j.artmed.2024.102938","volume":"155","author":"I Alonso","year":"2024","unstructured":"Alonso, I., Oronoz, M. & Agerri, R. MedExpQA: multilingual benchmarking of large language models for medical question answering. Artif. Intell. Med. 155, 102938 (2024).","journal-title":"Artif. Intell. Med."},{"key":"4626_CR58","doi-asserted-by":"publisher","first-page":"172","DOI":"10.1038\/s41586-023-06291-2","volume":"620","author":"K Singhal","year":"2023","unstructured":"Singhal, K. et al. Large language models encode clinical knowledge. Nature 620, 172\u2013180 (2023).","journal-title":"Nature"},{"key":"4626_CR59","unstructured":"Abacha, A. B., Agichtein, E., Pinter, Y. & Demner-Fushman, D. Overview of the medical question answering task at TREC 2017 LiveQA. https:\/\/trec.nist.gov\/pubs\/trec26\/papers\/Overview-QA.pdf (2017)."},{"key":"4626_CR60","unstructured":"Zawacki, A. et al. SIIM-ACR pneumothorax segmentation. https:\/\/kaggle.com\/competitions\/siim-acr-pneumothorax-segmentation (2024)."},{"key":"4626_CR61","doi-asserted-by":"publisher","unstructured":"Kather, J. N., Halama, N. & Marx, A. 100,000 histological images of human colorectal cancer and healthy tissue (version v0.1) [dataset]. Zenodo https:\/\/doi.org\/10.5281\/zenodo.1214456 (2018).","DOI":"10.5281\/zenodo.1214456"},{"key":"4626_CR62","doi-asserted-by":"crossref","unstructured":"Codella, N. C. F. et al. Skin lesion analysis toward melanoma detection: a challenge at the 2017 International Symposium on Biomedical Imaging (ISBI), hosted by the International Skin Imaging Collaboration (ISIC). In IEEE 15th International Symposium on Biomedical Imaging, 168\u2013172 (2018).","DOI":"10.1109\/ISBI.2018.8363547"},{"key":"4626_CR63","doi-asserted-by":"publisher","unstructured":"Pal, A. et al. ReXVQA: a large-scale visual question answering benchmark for generalist chest X-ray understanding. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2506.04353 (2025).","DOI":"10.48550\/arXiv.2506.04353"},{"key":"4626_CR64","doi-asserted-by":"publisher","unstructured":"Wu, J. et al. BRIDGE: benchmarking large language models for understanding real-world clinical practice text. Nat. Biomed. Eng. https:\/\/doi.org\/10.1038\/s41551-026-01719-2 (2026).","DOI":"10.1038\/s41551-026-01719-2"},{"key":"4626_CR65","unstructured":"Baharoon, M. et al. RadGame: an AI-powered platform for radiology education. In Proc. 5th Machine Learning for Health Symposium (eds Argaw, P. et al.) Vol. 297, 898\u2013920 (2026)."},{"key":"4626_CR66","doi-asserted-by":"crossref","unstructured":"Zhou, S. et al. Automating expert-level medical reasoning evaluation of large language models. npj Digit. Med. 9, 34 (2026).","DOI":"10.1038\/s41746-025-02208-7"},{"key":"4626_CR67","doi-asserted-by":"publisher","unstructured":"Sung, N.-J., Lee, D., Choi, B. H. & Park, C. J. AMRG: extend vision language models for automatic mammography report generation. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2508.09225 (2025).","DOI":"10.48550\/arXiv.2508.09225"},{"key":"4626_CR68","doi-asserted-by":"crossref","unstructured":"Gautam, S., Riegler, M. A. & Halvorsen, P. Kvasir-VQA-x1: a multimodal dataset for medical reasoning and robust MedVQA in gastrointestinal endoscopy. In Data Engineering in Medical Imaging (eds Bhattarai, B. et al.) 53\u201363 (2025).","DOI":"10.1007\/978-3-032-08009-7_6"},{"key":"4626_CR69","doi-asserted-by":"publisher","unstructured":"Aali, A. et al. MedVAL: toward expert-level medical text validation with language models. npj Digit. Med. https:\/\/doi.org\/10.1038\/s41746-026-03084-5 (2026).","DOI":"10.1038\/s41746-026-03084-5"},{"key":"4626_CR70","doi-asserted-by":"publisher","unstructured":"Cui, X., Huang, T., Wang, Y., Hu, L. & Yin, L. MedFM-Robust: benchmarking robustness of medical foundation models. Preprint at arXiv https:\/\/doi.org\/10.48550\/arXiv.2605.19027 (2026).","DOI":"10.48550\/arXiv.2605.19027"}],"container-title":["Nature Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41591-026-04626-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41591-026-04626-w","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41591-026-04626-w.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,11,7]],"date-time":"2026-11-07T01:03:29Z","timestamp":1794013409000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41591-026-04626-w"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10]]},"references-count":70,"journal-issue":{"issue":"10","published-print":{"date-parts":[[2026,10]]}},"alternative-id":["4626"],"URL":"https:\/\/doi.org\/10.1038\/s41591-026-04626-w","relation":{},"ISSN":["1078-8956","1546-170X"],"issn-type":[{"value":"1078-8956","type":"print"},{"value":"1546-170X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10]]},"assertion":[{"value":"9 March 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"4 August 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"6 October 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"All authors, except for C.L. and C. Kozlowski, are current or former employees of Google LLC, a subsidiary of Alphabet, and may own stock as part of their standard compensation packages.","order":1,"name":"Ethics","label":"Competing interests","group":{"name":"EthicsHeading","label":"Ethics"}}]}}