{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T11:28:11Z","timestamp":1786534091447,"version":"build-2736575974"},"reference-count":43,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T00:00:00Z","timestamp":1786492800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Pharmacol."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>Traditional Chinese Medicine (TCM) rheumatology presents unique challenges for AI-assisted clinical decision support, as the diagnostic process relies heavily on tacit knowledge and individualized reasoning. While Large Language Models (LLMs) have shown promise in medical applications, they remain limited by hallucination risks and inability to replicate expert TCM reasoning. Retrieval-Augmented Generation (RAG) offers a potential solution, yet its application to complex TCM dialectical reasoning remains underexplored.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We developed TCM-CoT-RAG, a hybrid framework combining RAG with Chain-of-Thought (CoT) prompting, grounded in 1,700 expert-curated clinical cases (1,600 for RAG retrieval; 100 for evaluation, including 50 for blinded expert review by three senior TCM rheumatologists). Deployed on Alibaba Cloud, the five system leverages state-of-the-art LLMs (DeepSeek-V3, Qwen3-235B) under a human-in-the-loop paradigm. We designed a dual-tier evaluation: (1) Objective extraction tasks (Task 1\u20132) quantified using F1-scores; (2) Generative tasks (Task 3\u20135) assessed using BERTScore. Two senior TCM rheumatologists (\u226515 years clinical experience) blindly assessed model outputs, and a senior chief expert quantified consistency between model predictions and ground truth (GT). Comprehensive ablation studies (S1-S4, S-Skip) isolated the contributions of each CoT module.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>TCM-CoT-RAG substantially improved diagnostic accuracy across five LLMs. DeepSeek-V3 with full-chain CoT-RAG achieved Entity F1 of 44.89% (+16.45% over baseline) and Formula F1 of 32.13% (+8.74% over baseline), with BERTScore of 0.81 indicating strong semantic alignment with expert reasoning. Ablation confirmed that the complete CoT pipeline was essential\u2014removing any reasoning module caused performance collapse below the zero-shot baseline. Two independent experts validated clinical utility (Cohen\u2019s \u03ba &amp;gt; 0.7). DeepSeek-V3 achieved the highest ground-truth consistency at 81.6%, and consistency metrics were quantified by the third expert holding the most senior professional title.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>This proof-of-concept framework demonstrates the potential of RAG-enhanced CoT reasoning to improve diagnostic consistency in TCM, objectifying the Symptom-Diagnosis-Prescription pipeline. It is important to note that this system is designed as an AI-assisted clinical decision-support tool. All recommendations require validation by qualified TCM practitioners before clinical application.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fphar.2026.1884526","type":"journal-article","created":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T10:29:50Z","timestamp":1786530590000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["TCM-CoT-RAG: a chain-of-thought enhanced retrieval-augmented generation system for clinical decision support in Traditional Chinese Medicine rheumatology"],"prefix":"10.3389","volume":"17","author":[{"given":"Bingbing","family":"Fan","sequence":"first","affiliation":[{"name":"Department of Rheumatology and Immunology, Xiyuan Hospital, China Academy of Chinese Medical Sciences","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yunlong","family":"Fang","sequence":"additional","affiliation":[{"name":"National Center for Integrative Medicine, Department of TCM Rheumatism, China-Japan Friendship Hospital","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zihan","family":"Wang","sequence":"additional","affiliation":[{"name":"National Center for Integrative Medicine, Department of TCM Rheumatism, China-Japan Friendship Hospital","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fang","family":"Ma","sequence":"additional","affiliation":[{"name":"Department of Rheumatology and Immunology, Xiyuan Hospital, China Academy of Chinese Medical Sciences","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,8,12]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"264","DOI":"10.1186\/s12911-019-0963-0","article-title":"Heterogeneous information network based clustering for precision traditional Chinese medicine","volume":"19","author":"Chen","year":"2019","journal-title":"BMC Med. Inf. Decis. Mak."},{"key":"B2","doi-asserted-by":"publisher","first-page":"141","DOI":"10.1038\/s43856-023-00370-1","article-title":"The future landscape of large language models in medicine","volume":"3","author":"Clusmann","year":"2023","journal-title":"Commun. Medicine"},{"key":"B3","doi-asserted-by":"publisher","first-page":"107530","DOI":"10.1016\/j.phrs.2024.107530","article-title":"TCMChat: a generative large language model for traditional Chinese medicine","volume":"210","author":"Dai","year":"2024","journal-title":"Pharmacol. Res."},{"key":"B4","doi-asserted-by":"publisher","first-page":"e52399","DOI":"10.2196\/52399","article-title":"Potential of large language models in health care: Delphi study","volume":"26","author":"Denecke","year":"2024","journal-title":"J. Med. Internet Res."},{"key":"B5","volume-title":"From Local to Global: A Graph Rag Approach to query-focused Summarization","author":"Edge","year":"2024"},{"key":"B6","volume-title":"Chatglm: A Family of Large Language Models from glm-130b to glm-4 all Tools","author":"Glm","year":"2024"},{"key":"B7","doi-asserted-by":"crossref","first-page":"1","DOI":"10.4018\/jkss.2010040101","article-title":"The knowledge CORE","volume":"1","author":"Griffiths","year":"2010","journal-title":"Int. J. Knowl. Syst. Sci."},{"key":"B8","doi-asserted-by":"publisher","first-page":"71","DOI":"10.1186\/s13020-026-01346-8","article-title":"Tuning and clinical application of large language models in traditional Chinese medicine: scoping review","volume":"21","author":"Han","year":"2026","journal-title":"Chin. Med."},{"key":"B9","doi-asserted-by":"publisher","first-page":"1403","DOI":"10.1038\/s41598-025-31038-6","article-title":"A multi module a.i. system for intelligent health insurance support using retrieval augmented generation","volume":"16","author":"Hanmante","year":"2025","journal-title":"Sci. Rep."},{"key":"B10","doi-asserted-by":"publisher","first-page":"682","DOI":"10.1038\/nchembio.118","article-title":"Network pharmacology: the next paradigm in drug discovery","volume":"4","author":"Hopkins","year":"2008","journal-title":"Nat. Chem. Biol."},{"key":"B11","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3703155","article-title":"A survey on hallucination in large language models: principles, taxonomy, challenges, and open questions","volume":"43","author":"Huang","year":"2025","journal-title":"ACM Trans. Inf. Syst."},{"key":"B12","doi-asserted-by":"publisher","first-page":"114","DOI":"10.4258\/hir.2025.31.2.114","article-title":"Large language models in medicine: clinical applications, technical challenges, and ethical considerations","volume":"31","author":"Jung","year":"2025","journal-title":"Hir"},{"key":"B13","doi-asserted-by":"publisher","first-page":"78","DOI":"10.1001\/jama.2023.8288","article-title":"Accuracy of a generative artificial intelligence model in a complex diagnostic challenge","volume":"330","author":"Kanjee","year":"2023","journal-title":"Jama"},{"key":"B14","doi-asserted-by":"publisher","first-page":"15","DOI":"10.4103\/cmac.cmac_2_20","article-title":"Tacit knowledge mining: the key traditional Chinese medical inheritance","volume":"3","author":"Lan","year":"2020","journal-title":"Chin. Med. Cult."},{"key":"B15","first-page":"9459","article-title":"Retrieval-augmented generation for knowledge-intensive nlp tasks","volume":"33","author":"Lewis","year":"2020","journal-title":"Adv. Neural Information Processing Systems"},{"key":"B16","doi-asserted-by":"publisher","first-page":"212","DOI":"10.1007\/s11684-011-0142-x","article-title":"Protecting traditional knowledge of Chinese medicine: concepts and proposals","volume":"5","author":"Liu","year":"2011","journal-title":"Front. Med."},{"key":"B17","volume-title":"Deepseek-v3 Technical Report. arXiv Preprint arXiv:241219437","author":"Liu","year":"2024"},{"key":"B18","doi-asserted-by":"publisher","first-page":"2185","DOI":"10.1109\/JBHI.2024.3510884","article-title":"RTGN: robust traditional Chinese medicine graph networks for patient similarity learning","volume":"29","author":"Long","year":"2024","journal-title":"IEEE J. Biomed. Health Inf."},{"key":"B19","doi-asserted-by":"publisher","first-page":"37","DOI":"10.1016\/j.eng.2023.01.014","article-title":"Artificial intelligence in pharmaceutical sciences","volume":"27","author":"Lu","year":"2023","journal-title":"Engineering"},{"key":"B20","doi-asserted-by":"publisher","first-page":"120","DOI":"10.1038\/s41746-023-00873-0","article-title":"The imperative for regulatory oversight of large language models (or generative AI) in healthcare","volume":"6","author":"Mesk\u00f3","year":"2023","journal-title":"NPJ Digital Medicine"},{"key":"B21","doi-asserted-by":"publisher","first-page":"259","DOI":"10.1038\/s41586-023-05881-4","article-title":"Foundation models for generalist medical artificial intelligence","volume":"616","author":"Moor","year":"2023","journal-title":"Nature"},{"key":"B22","doi-asserted-by":"publisher","first-page":"106362","DOI":"10.1016\/j.ijmedinf.2026.106362","article-title":"Human in the loop artificial intelligence in healthcare: applications, outcomes, and implementation challenges","volume":"213","author":"Olawade","year":"2026","journal-title":"Int. J. Med. Inf."},{"key":"B23","doi-asserted-by":"publisher","first-page":"3580","DOI":"10.1109\/tkde.2024.3352100","article-title":"Unifying large language models and knowledge graphs: a roadmap","volume":"36","author":"Pan","year":"2024","journal-title":"IEEE Trans. Knowl. Data Eng."},{"key":"B24","doi-asserted-by":"publisher","first-page":"3583","DOI":"10.32604\/cmc.2024.055671","article-title":"A review of knowledge graph in traditional Chinese medicine: analysis, construction, application and prospects","volume":"81","author":"Qu","year":"2024","journal-title":"Comput. Mater. & Continua"},{"key":"B25","volume-title":"Openai gpt-5 System Card","author":"Singh","year":"2025"},{"key":"B26","doi-asserted-by":"publisher","first-page":"172","DOI":"10.1038\/s41586-023-06291-2","article-title":"Large language models encode clinical knowledge","volume":"620","author":"Singhal","year":"2023","journal-title":"Nature"},{"key":"B27","doi-asserted-by":"publisher","first-page":"174","DOI":"10.1097\/gscm.0000000000000035","article-title":"Research progress on knowledge discovery in traditional Chinese medicine medical records based on artificial intelligence","volume":"2","author":"Tang","year":"2024","journal-title":"Guidel. Stand. Chin. Med."},{"key":"B28","doi-asserted-by":"publisher","first-page":"627","DOI":"10.1038\/s41746-025-02022-1","article-title":"A generative AI teaching assistant for personalized learning in medical education","volume":"8","author":"Thesen","year":"2025","journal-title":"NPJ Digit. Med."},{"key":"B29","doi-asserted-by":"publisher","first-page":"1930","DOI":"10.1038\/s41591-023-02448-8","article-title":"Large language models in medicine","volume":"29","author":"Thirunavukarasu","year":"2023","journal-title":"Nat. Medicine"},{"key":"B30","article-title":"Research status of literature digitalization for traditional Chinese medicine","author":"Ton","year":"2015","journal-title":"China Digit. Med."},{"key":"B31","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1038\/s41591-018-0300-7","article-title":"High-performance medicine: the convergence of human and artificial intelligence","volume":"25","author":"Topol","year":"2019","journal-title":"Nat. Medicine"},{"key":"B32","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1186\/s13020-025-01295-8","article-title":"GastroTCM: a large language model assistant for gastroenterology in traditional Chinese medicine","volume":"21","author":"Wang","year":"2026","journal-title":"Chin. Med."},{"key":"B33","doi-asserted-by":"crossref","first-page":"24824","DOI":"10.52202\/068431-1800","article-title":"Chain-of-thought prompting elicits reasoning in large language models","volume":"35","author":"Wei","year":"2022","journal-title":"Adv. Neural Information Processing Systems"},{"key":"B34","doi-asserted-by":"publisher","DOI":"10.1109\/JBHI.2025.3612415","article-title":"BianCang: a traditional Chinese medicine large language model","author":"Wei","year":"2025","journal-title":"IEEE J. Biomed. Health Inf."},{"key":"B35","doi-asserted-by":"publisher","first-page":"135","DOI":"10.1038\/s41746-023-00879-8","article-title":"The shaky foundations of large language models and foundation models for electronic health records","volume":"6","author":"Wornow","year":"2023","journal-title":"Npj Digital Medicine"},{"key":"B36","doi-asserted-by":"publisher","first-page":"1207","DOI":"10.1089\/acm.2009.0700","article-title":"Traditional Chinese medicine information digitalization discussion","volume":"16","author":"Xie","year":"2010","journal-title":"J. Altern. Complementary Med."},{"key":"B37","volume-title":"TCM Masters Miner for Knowledge Transfer. 2007 IEEE International Conference on Systems, Man and Cybernetics","author":"Xijin","year":"2007"},{"key":"B38","volume-title":"Superclue: A Comprehensive Chinese Large Language Model Benchmark","author":"Xu","year":"2023"},{"key":"B39","volume-title":"Qwen3 Technical Report. arXiv Preprint arXiv:250509388","author":"Yang","year":"2025"},{"key":"B40","doi-asserted-by":"publisher","first-page":"AIoa2300068","DOI":"10.1056\/aioa2300068","article-title":"Almanac\u2014Retrieval-augmented language models for clinical medicine","volume":"1","author":"Zakka","year":"2024","journal-title":"Nejm Ai"},{"key":"B41","volume-title":"Bertscore: Evaluating Text Generation with Bert","author":"Zhang","year":"2019"},{"key":"B42","doi-asserted-by":"publisher","first-page":"1067","DOI":"10.1142\/s0192415x23500490","article-title":"Advances in the application of traditional Chinese medicine using artificial intelligence: a review","volume":"51","author":"Zhang","year":"2023","journal-title":"Am. Journal Chin. Medicine"},{"key":"B43","doi-asserted-by":"crossref","first-page":"46595","DOI":"10.52202\/075280-2020","article-title":"Judging llm-as-a-judge with mt-bench and chatbot arena","volume":"36","author":"Zheng","year":"2023","journal-title":"Adv. Neural Information Processing Systems"}],"container-title":["Frontiers in Pharmacology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fphar.2026.1884526\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,12]],"date-time":"2026-08-12T10:29:54Z","timestamp":1786530594000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fphar.2026.1884526\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,12]]},"references-count":43,"alternative-id":["10.3389\/fphar.2026.1884526"],"URL":"https:\/\/doi.org\/10.3389\/fphar.2026.1884526","relation":{},"ISSN":["1663-9812"],"issn-type":[{"value":"1663-9812","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,12]]},"article-number":"1884526"}}