거대언어모델을 활용한 한의학 교육 강화: 교육용 인공지능 개발을 중심으로
Enhancing Korean Medicine Education with Large Language Models: Focusing on the Development of Educational Artificial Intelligence
1
조회
0
다운로드
4
인용
0
참고문헌
키워드
초록
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.
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.
참고문헌 (30)
- Eloundou T, "Gpts are gpts: An early look at the labor market impact potential of large language models", ~ (2023)
- Bubeck S, "Sparks of artificial general intelligence: Early experiments with gpt-4", ~ (2023)
- Miao Q, "DAO to HANOI via DeSci : AI paradigm shifts from AlphaGo to ChatGPT", IEEE, 2023, pp.10(4), 877~897
- Moor M, "Foundation models for generalist medical artificial intelligence", Nature, 616(7956), pp.259~265 (2023)
- Singhal K, "Large language models encode clinical knowledge", Nature, pp.1~9 (2023)
- Nori H, "Capabilities of gpt-4 on medical challenge problems", ~ (2023)
- Kung TH, "Performance of ChatGPT on USMLE : Potential for AI-assisted medical education using large language models", PLoS digital health, 2(2), pp.e0000198~ (2023)
- Singhal K, "Towards expert-level medical question answering with large language models", ~ (2023)
- Arora A, "The promise of large language models in health care", The Lancet, 401(10377), pp.641~ (2023)
- Adams LC, "Leveraging GPT-4 for post hoc transformation of free-text radiology reports into structured reporting : a multilingual feasibility study", Radiology, 307(4), pp.e230725~ (2023)
- Kraljevic Z, "Foresight--Deep Generative Modelling of Patient Timelines using Electronic Health Records", ~ (2022)
- Jiang LY, "Health system-scale language models are all-purpose prediction engines", Nature, pp.1~6 (2023)
- Nastasi AJ, "Does ChatGPT provide appropriate and equitable medical advice?: A vignette-based, clinical evaluation across care contexts", ~ (2023)
- Ayers JW, "Comparing physician and artificial intelligence chatbot responses to patient questions posted to a public social media forum", JAMA internal medicine (2023)
- Team D, "Introducing Duolingo Max, a learning experience powered by GPT-4", Retrieved March, 15, pp.2023~ (2023)
- Jang D, "Exploring the Potential of Large Language models in Traditional Korean Medicine: A Foundation Model Approach to Culturally-Adapted Healthcare", ~ (2023)
- Brown T, "Language models are few-shot learners", Advances in neural information processing systems, 33, pp.1877~1901 (2020)
- Ouyang L, "Training language models to follow instructions with human feedback", Advances in Neural Information Processing Systems, 35, pp.27730~27744 (2022)
- Touvron H, "Llama: Open and efficient foundation language models", ~ (2023)
- Touvron H, "Llama 2: Open foundation and fine-tuned chat models", ~ (2023)
- Wei J, "Emergent abilities of large language models", ~ (2022)
- "State of GPT 2023"[URL]
- Yao S, "Tree of thoughts: Deliberate problem solving with large language models", ~ (2023)
- Frommeyer TC, "The desire of medical students to integrate artificial intelligence into medical education : an opinion article", Frontiers in Digital Health, 4, pp.831123~ (2022)
- Thirunavukarasu AJ, "Large language models in medicine", Nature Medicine, pp.1~11 (2023)
- Benoit JR, "ChatGPT for Clinical Vignette Generation, Revision, and Evaluation", ~ (2023)
- Kitamura FC, "ChatGPT is shaping the future of medical writing but still requires human judgment", Radiological Society of North America, pp.e230171~ (2023)
- Ali SR, "Using ChatGPT to write patient clinic letters", The Lancet Digital Health, 5(4), pp.e179~e181 (2023)
- Biswas S, "ChatGPT and the future of medical writing", Radiological Society of North America, pp.e223312~ (2023)
- Chen T-J, "ChatGPT and other artificial intelligence applications speed up scientific writing", Journal of the Chinese Medical Association, 86(4), pp.351~353 (2023)