생성형 AI를 이용한 작업치료기반 SOAP노트 챗봇에 관한 연구
Development of Task-based SOAP Note Chat-bot Using Generative AI
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
본 연구의 목적은 생성형 AI를 활용하여 작업치료 SOAP 노트를 자동으로 생성하는 챗봇을 개발하고, 이를 통해 작업치료 학생들의 문서화 능력을 향상시키며, 치료 계획 수립에 도움이 되고자 한다. 이를 위해 다양한 출처에서 SOAP 노트를 수집하고, 데이터 클리닝 및 전처리를 통해 모델 훈련에 적합한 데이터셋을 구축하였다. 이후 ChatGPT 모델을 작업치료 SOAP 노트 작성에 최적화하기 위해 fine-tuning을 실시하였으며, 모델의 성능을 평가하기 위해 실제 작업치료 사례를 적용하고, 작업치료 전문가들의 피드백을 반영하여 개선을 진행하였다. 연구 결과, 개발된 챗봇은 주관적정보와 객관적 정보를 입력받아 자동으로 평가(A)와 계획(P)를 생성하는 기능을 성공적으로 구현하였으며, 높은 정확성과 일관성을 보였다. 작업치료 전문가들로부터 긍정적인 피드백을 받았고, 실용적으로 활용될 수 있음을 확인하였다. 향후 다양한 헬스케어 분야로의 확장 가능성을 탐구할 필요가 있으며, 보다 넓은 범위의 데이터를 활용하여 모델의 일반화 성능을 향상시키는 추가 연구가 필요하다. 본연구는 국내 작업치료 및 재활 분야에서 생성형 AI를 활용한 최초의 시도 중 하나로서 여러 가지 중요한 의의를 가지고 있다.
Abstract
The purpose of this study is to develop a chatbot that automatically generates occupational therapy SOAP notes using generative AI, thereby enhancing the documentation skills of occupational therapy students and aiding in treatment planning. To achieve this, SOAP notes were collected from various sources, and a dataset suitable for model training was constructed through data cleaning and preprocessing. Subsequently, the ChatGPT model was fine-tuned to optimize the generation of occupational therapy SOAP notes. The model's performance was evaluated by applying it to actual occupational therapy cases and incorporating feedback from occupational therapy experts. The results of the study indicate that the developed chatbot successfully implemented the function of automatically generating assessment (A) and plan (P) content based on subjective and objective information. The chatbot demonstrated high accuracy and consistency, receiving positive feedback from occupational therapy experts and proving to be practically useful. Further research is needed to explore the potential for expanding the application to various healthcare fields and to improve the model's generalization performance by utilizing a wider range of data. This study holds significant value as one of the first attempts to utilize generative AI in the field of occupational therapy and rehabilitation in Korea.
참고문헌 (10)
- H-A. Jeon, "Development of a casebased nursing education program using generative artificial intelligence", The Journal of Korean Academic Society of Nursing Education, 29(3), pp.234~246 (2023)
- Park, J. H., "Applications and concerns of generative AI : ChatGPT in the field of occupational health", Journal of Korean Society of Occupational and Environmental Hygiene, 33(4), pp.412~418 (2023)
- T. Lim, "Analysis of pre-service teachers' experience with generative AI : Creating animated fairy tale videos", The Journal of Korean Association of Computer Education, 26(5), pp.139~153 (2023)
- The AI index 2023 annual report, "Maslej, N", AI index steering committee, Institute for Human-Centered AI, Stanford University, ~ (2023)
- S-M. Choi, "Use Cases and prospects of ChatGPT :The possibility of caring", Citizen and Humanities, (45), pp.129~161 (2023)
- S. Kang, "Development and application of generative AI-based instructional design platform", The Journal of Korean Association of Computer Education, 26(6), pp.143~153 (2023)
- Y-A. Yang, "Occupational Therapy Assessment" (2022), EduFactory, Publication
- J. Song, "Study on clinical education for nursing in hospitals in Korea", The Journal of Korean academic society of nursing education, 19(2), pp.251~264 (2013)
- E-J. Park, "Meta-analysis on the effect of English education programs using AI Chatbots on language education effectiveness", Journal of the Korea English Education Society, 22(3), pp.113~132 (2023)
- Jovanovic, M, "Generative artificial intelligence: Trends and prospects", Computer, 55(10), pp.107~112 (2022)