DOI QR코드

DOI QR Code

Development of an AI Model to Determine the Relationship between Cerebrovascular Disease and the Work Environment as well as Analysis of Consistency with Expert Judgment

뇌심혈관 질환과 업무 환경의 연관성 판단을 위한 AI 모델의 개발 및 전문가 판단과의 일치도 분석

  • Juyeon Oh (Department of Public Health, Graduate School, Yonsei University) ;
  • Ki-bong Yoo (Department of Health Administration, Yonsei University) ;
  • Ick Hoon Jin (Department of Applied Statistics, Yonsei University) ;
  • Byungyoon Yun (Department of Occupational Health, Department of Preventive Medicine, Yonsei University College of Medicine) ;
  • Juho Sim (Department of Occupational Health, Department of Preventive Medicine, Yonsei University College of Medicine) ;
  • Heejoo Park (Department of Public Health, Graduate School, Yonsei University) ;
  • Jongmin Lee (Department of Occupational Health, Graduate School of Public Health, Yonsei University) ;
  • Jian Lee (Department of Public Health, Graduate School, Yonsei University) ;
  • Jin-Ha Yoon (Department of Occupational Health, Department of Preventive Medicine, Yonsei University College of Medicine)
  • 오주연 (연세대학교 일반대학원 보건학과) ;
  • 유기봉 (연세대학교 보건행정학부) ;
  • 진익훈 (연세대학교 응용통계학과) ;
  • 윤병윤 (연세대학교 의과대학 예방의학교실) ;
  • 심주호 (연세대학교 의과대학 예방의학교실) ;
  • 박희주 (연세대학교 일반대학원 보건학과) ;
  • 이종민 (연세대학교 보건대학원 산업보건전공) ;
  • 이지안 (연세대학교 일반대학원 보건학과) ;
  • 윤진하 (연세대학교 의과대학 예방의학교실)
  • Received : 2024.04.09
  • Accepted : 2024.09.11
  • Published : 2024.09.30

Abstract

Introduction: Acknowledging the global issue of diseases potentially caused by overwork, this study aims to develop an AI model to help workers understand the connection between cerebrocardiovascular diseases and their work environment. Materials and methods: The model was trained using medical and legal expertise along with data from the 2021 occupational disease adjudication certificate by the Industrial Accident Compensation Insurance and Prevention Service. The Polyglot-ko-5.8B model, which is effective for processing Korean, was utilized. Model performance was evaluated through accuracy, precision, sensitivity, and F1-score metrics. Results: The model trained on a comprehensive dataset, including expert knowledge and actual case data, outperformed the others with respective accuracy, precision, sensitivity, and F1-scores of 0.91, 0.89, 0.84, and 0.87. However, it still had limitations in responding to certain scenarios. Discussion: The comprehensive model proved most effective in diagnosing work-related cerebrocardiovascular diseases, highlighting the significance of integrating actual case data in AI model development. Despite its efficacy, the model showed limitations in handling diverse cases and offering health management solutions. Conclusion: The study succeeded in creating an AI model to discern the link between work factors and cerebrocardiovascular diseases, showcasing the highest efficacy with the comprehensively trained model. Future enhancements towards a template-based approach and the development of a user-friendly chatbot webUI for workers are recommended to address the model's current limitations.

Keywords

References

  1. Enforcement decree of the industrial accident compensation insurance act (2023. 7. 1.) article 37 paragraph 5 act no. 18928. 2023a
  2. Enforcement decree of the industrial accident compensation insurance act article 34 paragraph 3 and subparagraph 1 (a) of the appendix act no. 18928. 2023b
  3. Cherry N. Recent advances: Occupational disease. Bmj 1999;318(7195):1397-1399 (https://doi.org/10.1136/bmj.318.7195.1397)
  4. Deaton C, Froelicher ES, Wu LH, Ho C, Shishani K et al. The global burden of cardiovascular disease. Eur J Cardiovasc Nurs 2011;10 Suppl 2:S5-13 (https://doi.org/10.1016/s1474-5151(11)00111-3)
  5. Ding J, Li X, Kang X, Gudivada VN. A case study of the augmentation and evaluation of training data for deep learning. Journal of Data and Information Quality (JDIQ) 2019;11(4):1-22 (https://doi.org/10.1145/3317573)
  6. Guthrie R & Westaway J. Emerging legal concerns with chronic diseases in the australian workplace: Pre-employment medicals, functional capacity evaluations, workers' compensation and disability discrimination. J Law Med 2009;16(5):803-821)
  7. Jang SS, Hwang KH, Ha YG, editors. High quality training set collection using generative adversarial network. 2020 IEEE International Conference on Big Data and Smart Computing (BigComp); 2020: IEEE; (https://doi.org/10.1109/BigComp48618.2020.00-27)
  8. Joo HY, Oh H, Yang J. A survey on open source based large language models. The Journal of Korea Institute of Information, Electronics, Communication Technology 2023;16(4):193-202 (https://doi.org/10.17661/jkiiect.2023.16.4.193)
  9. Kamardeen I. Work stress related cardiovascular diseases among construction professionals. Built Environment Project Asset Management 2022;12(2):223-242 (https://doi.org/10.1108/BEPAM-06-2021-0081)
  10. Kelly S & Romero A. Consequences of sharing invisible chronic health conditions at work: Implications for business communication education and training. Business Communication Research Practice 2019;2(1):13-21 (https://doi.org/10.22682/bcrp.2019.2.1.13)
  11. Kim S & Jung Y. Effect of long working hours on cardiovascular disease in south korean workers: A longitudinal study. Asia Pac J Public Health 2021;33(2-3):213-219 (https://doi.org/10.1177/1010539520979927)
  12. Kim UJ, Choi WJ, Kang SK, Lee W, Ham S et al. Standards for recognition and approval rate of occupational cerebro-cardiovascular diseases in korea. Ann Occup Environ Med 2022;34:e30 (https://doi.org/10.35371/aoem.2022.34.e30)
  13. Ko H, Yang K, Ryu M, Choi T, Yang S et al. A technical report for polyglot-ko: Open-source large-scale korean language models. arXiv preprint arXiv:02254 2023 (https://doi.org/10.48550/arXiv.2306.02254)
  14. Kosis. Mortality trend by cause of death 2022: Korean Statistical Information Service; 2022 [Available from: https://www.index.go.kr/unity/potal/main/EachDtlPageDetail.do?idx_cd=1012
  15. Lin RT, Chien LC, Kawachi I. Nonlinear associations between working hours and overwork-related cerebrovascular and cardiovascular diseases (ccvd). Scientific Reports 2018;8(1):9694 (https://doi.org/10.1038/s41598-018-28141-2)
  16. Lin XV, Mihaylov T, Artetxe M, Wang T, Chen S et al. Few-shot learning with multilingual language models. arXiv preprint arXiv:10668 2021 (https://doi.org/10.48550/arXiv.2112.10668)
  17. Montesinos Lopez OA, Montesinos Lopez A, Crossa J. Overfitting, model tuning, and evaluation of prediction performance. Multivariate statistical machine learning methods for genomic prediction: Springer; 2022. p. 109-139 (https://doi.org/10.1007/978-3-030-89010-0_4)
  18. Mora CD & Bornstein RA. Evaluation of cerebrovascular disease. Neuropsychology: Springer; 1998. p. 171-186 (https://doi.org/10.1007/978-1-4899-1950-2_8)
  19. Park S, Nam J, Lee JK, Oh SS, Kang HT et al. Association between night work and cardiovascular diseases: Analysis of the 3rd korean working conditions survey. Ann Occup Environ Med 2015;27:15 (https://doi.org/10.1186/s40557-015-0064-1)
  20. Picciotto D. [occupational cardiovascular diseases and phlebopathies]. G Ital Med Lav Ergon 2010;32(4 Suppl):160-161)
  21. Roth GA, Abate D, Abate KH, Abay SM, Abbafati C et al. Global, regional, and national age-sex-specific mortality for 282 causes of death in 195 countries and territories, 1980-2017: A systematic analysis for the global burden of disease study 2017. The lancet 2018;392(10159):1736-1788 (https://doi.org/10.1007/978-1-4899-1950-2_8)
  22. Thrift AG, Cadilhac DA, Thayabaranathan T, Howard G, Howard VJ et al. Global stroke statistics. Int J Stroke 2014;9(1):6-18 (https://doi.org/10.1111/ijs.12245)
  23. Trivedi UB & Mishra P, editors. Fine tuning (diagnosis) of machine learning algorithm (model) for optimization. 2022 Second International Conference on Advances in Electrical, Computing, Communication and Sustainable Technologies (ICAECT); 2022: IEEE; (https://doi.org/10.1109/ICAECT54875.2022.9807982)
  24. Vineis P, Demetriou CA, Probst-Hensch N. Long-term effects of air pollution: An exposome meet-in-the-middle approach. Int J Public Health 2020;65(2): 125-127 (https://doi.org/10.1007/s00038-019-01329-7)
  25. Who. Cardiovascular diseases (cvds): World Health Organization; 2017 [Available from: http://www.who.int/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds)
  26. Wu Y, Xiong Y, Wang P, Liu R, Jia X et al. Risk factors of cardiovascular and cerebrovascular diseases in young and middle-aged adults: A meta-analysis. Medicine (Baltimore) 2022;101(48):e32082 (https://doi.org/10.1097/md.0000000000032082)
  27. Zhuang YT, Wu F, Chen C, Pan YH. Challenges and opportunities: From big data to knowledge in ai 2.0. Frontiers of Information Technology Electronic Engineering 2017;18:3-14 (https://doi.org/10.1631/FITEE.1601883)