Journal of Physiology & Pathology in Korean Medicine (동의생리병리학회지)

The Physiological Society of Korean Medicine and The Society of Pathology in Korean Medicine (대한동의생리학회·한의병리학회)
권호 표지
DOI QR코드

DOI QR Code

Enhancing Korean Medicine Education with Large Language Models: Focusing on the Development of Educational Artificial Intelligence

거대언어모델을 활용한 한의학 교육 강화 : 교육용 인공지능 개발을 중심으로

  • Sa-Yoon Park (Department of Physiology, College of Korean Medicine, Gachon University) ;
  • Chang-Eop Kim (Department of Physiology, College of Korean Medicine, Gachon University)
  • 박사윤 (가천대학교 한의과대학) ;
  • 김창업 (가천대학교 한의과대학)
  • Received : 2023.09.08
  • Accepted : 2023.10.24
  • Published : 2023.10.25
Download PDF
⟨ Previous Next ⟩

Abstract

Large language models (LLMs) have introduced groundbreaking innovations in various fields, including healthcare, where they augment medical diagnosis, decision-making, and facilitate patient-doctor communication through their exceptional contextual understanding and inferential abilities. In the realm of Korean medicine (KM), the utilization of LLMs is highly anticipated. However, it demands additional training with domain-specific KM data for seamless integration of KM knowledge. There are two predominant strategies for training domain-specific LLMs in the KM domain. The first approach entails direct manipulation of the LLM's internals by either pretraining a base model on an extensive corpus of KM data or fine-tuning a pretrained model's parameters using KM-related question-answering datasets. The second approach avoids internal model manipulation and leverages techniques like prompt engineering, retrieval augmented generation, and cognitive augmentation. Domain-specific LLMs specialized for KM hold the potential for diverse applications, ranging from personalized medical education plans and content generation to knowledge integration, curriculum development, automated student assessment, virtual patient simulations, and advanced research and scholarly activities. These advancements are poised to significantly impact the field of KM and medical education at large.

Keywords

Acknowledgement

이 논문은 2022년과 2023년도 정부(교육부)의 재원으로 한국연구재단의 지원을 받아 수행된 기초연구사업임(2022R1F1A1068841, RS-2023-00248152).

References

  1. Eloundou T, Manning S, Mishkin P, Rock D. Gpts are gpts: An early look at the labor market impact potential of large language models. arXiv preprint arXiv:230310130. 2023.
  2. Bubeck S, Chandrasekaran V, Eldan R, Gehrke J, Horvitz E, Kamar E, et al. Sparks of artificial general intelligence: Early experiments with gpt-4. arXiv preprint arXiv:230312712. 2023.
  3. Miao Q, Zheng W, Lv Y, Huang M, Ding W, Wang F-Y. DAO to HANOI via DeSci: AI paradigm shifts from AlphaGo to ChatGPT. IEEE/CAA Journal of Automatica Sinica. 2023;10(4):877-97. https://doi.org/10.1109/JAS.2023.123561
  4. Moor M, Banerjee O, Abad ZSH, Krumholz HM, Leskovec J, Topol EJ, et al. Foundation models for generalist medical artificial intelligence. Nature. 2023;616(7956):259-65. https://doi.org/10.1038/s41586-023-05881-4
  5. Singhal K, Azizi S, Tu T, Mahdavi SS, Wei J, Chung HW, et al. Large language models encode clinical knowledge. Nature. 2023:1-9.
  6. Nori H, King N, McKinney SM, Carignan D, Horvitz E. Capabilities of gpt-4 on medical challenge problems. arXiv preprint arXiv:230313375. 2023.
  7. Kung TH, Cheatham M, Medenilla A, Sillos C, De Leon L, Elepano C, et al. Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLoS digital health. 2023;2(2):e0000198.
  8. Singhal K, Tu T, Gottweis J, Sayres R, Wulczyn E, Hou L, et al. Towards expert-level medical question answering with large language models. arXiv preprint arXiv:230509617, 2023.
  9. Arora A, Arora A. The promise of large language models in health care. The Lancet. 2023;401(10377):641.
  10. Adams LC, Truhn D, Busch F, Kader A, Niehues SM, Makowski MR, et al. Leveraging GPT-4 for post hoc transformation of free-text radiology reports into structured reporting: a multilingual feasibility study. Radiology. 2023;307(4):e230725.
  11. Kraljevic Z, Bean D, Shek A, Bendayan R, Yeung JA, Deng A, et al. Foresight--Deep Generative Modelling of Patient Timelines using Electronic Health Records. arXiv preprint arXiv:221208072. 2022.
  12. Jiang LY, Liu XC, Nejatian NP, Nasir-Moin M, Wang D, Abidin A, et al. Health system-scale language models are all-purpose prediction engines. Nature. 2023:1-6.
  13. Nastasi AJ, Courtright KR, Halpern SD, Weissman GE. Does ChatGPT provide appropriate and equitable medical advice?: A vignette-based, clinical evaluation across care contexts. medRxiv. 2023:2023.02. 25.23286451. 25.23286451
  14. Ayers JW, Poliak A, Dredze M, Leas EC, Zhu Z, Kelley JB, et al. Comparing physician and artificial intelligence chatbot responses to patient questions posted to a public social media forum. JAMA internal medicine. 2023.
  15. Team D. Introducing Duolingo Max, a learning experience powered by GPT-4. Retrieved March. 2023;15:2023.
  16. Jang D, Kim C-E. Exploring the Potential of Large Language models in Traditional Korean Medicine: A Foundation Model Approach to Culturally-Adapted Healthcare. arXiv preprint arXiv:230317807. 2023.
  17. Brown T, Mann B, Ryder N, Subbiah M, Kaplan JD, Dhariwal P, et al. Language models are few-shot learners. Advances in neural information processing systems. 2020;33:1877-901.
  18. Ouyang L, Wu J, Jiang X, Almeida D, Wainwright C, Mishkin P, et al. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems. 2022;35:27730-44.
  19. Touvron H, Lavril T, Izacard G, Martinet X, Lachaux M-A, Lacroix T, et al. Llama: Open and efficient foundation language models. arXiv preprint arXiv:230213971. 2023.
  20. Touvron H, Martin L, Stone K, Albert P, Almahairi A, Babaei Y, et al. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:230709288. 2023.
  21. Wei J, Tay Y, Bommasani R, Raffel C, Zoph B, Borgeaud S, et al. Emergent abilities of large language models. arXiv preprint arXiv:220607682. 2022.
  22. Karpathy A. State of GPT 2023 [Available from: https://build.microsoft.com/en-US/sessions/db3f4859-cd30-4445-a0cd-553c3304f8e2.
  23. Yao S, Yu D, Zhao J, Shafran I, Griffiths TL, Cao Y, et al. Tree of thoughts: Deliberate problem solving with large language models. arXiv preprint arXiv:230510601. 2023.
  24. Frommeyer TC, Fursmidt RM, Gilbert MM, Bett ES. The desire of medical students to integrate artificial intelligence into medical education: an opinion article. Frontiers in Digital Health. 2022;4:831123.
  25. Thirunavukarasu AJ, Ting DSJ, Elangovan K, Gutierrez L, Tan TF, Ting DSW. Large language models in medicine. Nature Medicine. 2023:1-11.
  26. Benoit JR. ChatGPT for Clinical Vignette Generation, Revision, and Evaluation. medRxiv. 2023:2023.02.04.23285478.
  27. Kitamura FC. ChatGPT is shaping the future of medical writing but still requires human judgment. Radiological Society of North America; 2023. p. e230171.
  28. Ali SR, Dobbs TD, Hutchings HA, Whitaker IS. Using ChatGPT to write patient clinic letters. The Lancet Digital Health. 2023;5(4):e179-e81. https://doi.org/10.1016/S2589-7500(23)00048-1
  29. Biswas S. ChatGPT and the future of medical writing. Radiological Society of North America; 2023. p. e223312.
  30. Chen T-J. ChatGPT and other artificial intelligence applications speed up scientific writing. Journal of the Chinese Medical Association. 2023;86(4):351-3. https://doi.org/10.1097/JCMA.0000000000000900