Phonetics and Speech Sciences (말소리와 음성과학)

Korean Society of Speech Sciences (한국음성학회)
권호 표지
DOI QR코드

DOI QR Code

Machine-learning-based out-of-hospital cardiac arrest (OHCA) detection in emergency calls using speech recognition

119 응급신고에서 수보요원과 신고자의 통화분석을 활용한 머신 러닝 기반의 심정지 탐지 모델

  • Jong In Kim (Interdisciplinary Program in Cognitive Science, Seoul National University) ;
  • Joo Young Lee (Department of Linguistics, Seoul National University) ;
  • Jio Chung (VS Works) ;
  • Dae Jin Shin (SoundMind) ;
  • Dong Hyun Choi (Department of Biomedical Engineering, Seoul National University College of Medicine) ;
  • Ki Hong Kim (Laboratory of Emergency Medical Services, Seoul National University Hospital Biomedical Research Institute) ;
  • Ki Jeong Hong (Laboratory of Emergency Medical Services, Seoul National University Hospital Biomedical Research Institute) ;
  • Sunhee Kim (Department of French Language Education, Seoul National University) ;
  • Minhwa Chung (Interdisciplinary Program in Cognitive Science, Seoul National University)
  • 김종인 (서울대학교 인문대학 인지과학 협동과정) ;
  • 이주영 (서울대학교 인문대학 언어학과) ;
  • 정지오 (브이에스웍스) ;
  • 신대진 (사운드마인드) ;
  • 최동현 (서울대학교 의과대학 의공학교실) ;
  • 김기홍 (서울대학교병원 의생명연구원 응급의료연구실) ;
  • 홍기정 (서울대학교병원 의생명연구원 응급의료연구실) ;
  • 김선희 (서울대학교 사범대학 불어교육과) ;
  • 정민화 (서울대학교 인문대학 인지과학 협동과정)
  • Received : 2023.11.24
  • Accepted : 2023.12.11
  • Published : 2023.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Cardiac arrest is a critical medical emergency where immediate response is essential for patient survival. This is especially true for Out-of-Hospital Cardiac Arrest (OHCA), for which the actions of emergency medical services in the early stages significantly impact outcomes. However, in Korea, a challenge arises due to a shortage of dispatcher who handle a large volume of emergency calls. In such situations, the implementation of a machine learning-based OHCA detection program can assist responders and improve patient survival rates. In this study, we address this challenge by developing a machine learning-based OHCA detection program. This program analyzes transcripts of conversations between responders and callers to identify instances of cardiac arrest. The proposed model includes an automatic transcription module for these conversations, a text-based cardiac arrest detection model, and the necessary server and client components for program deployment. Importantly, The experimental results demonstrate the model's effectiveness, achieving a performance score of 79.49% based on the F1 metric and reducing the time needed for cardiac arrest detection by 15 seconds compared to dispatcher. Despite working with a limited dataset, this research highlights the potential of a cardiac arrest detection program as a valuable tool for responders, ultimately enhancing cardiac arrest survival rates.

심정지는 초기 대응에 따라 생존율과 예후에 영향을 미치는 중요한 응급 상황이다. 특히 병원밖심정지(out-of-hospital cardiac arrest, OHCA)의 경우, 119 구조대의 초기 조치가 심정지 환자의 생존율을 높이는 데 결정적인 역할을 한다. 그러나 국내에서는 수보요원의 수가 제한적이지만 다량의 신고 전화에 응대해야 하는 현실이다. 이런 상황에서 머신러닝 기반의 OHCA 탐지 프로그램은 수보요원의 보조 역할로 심정지 환자의 생존률을 높일 수 있다. 본 연구에서는 이러한 문제를 해결하기 위해 머신러닝 기반의 심정지(OHCA) 탐지 프로그램을 개발하였다. 이 프로그램은 수보요원과 신고자의 통화 녹취록을 분석하여 심정지 여부를 판단한다. 제안한 모델은 수보요원 및 신고자와의 통화를 자동으로 전사하는 모델, 텍스트 기반의 심정지 탐지 모델, 그리고 프로그램 개발을 위한 서버와 클라이언트로 구성되어 있다. 실험 결과, 본 연구에서 제안한 모델은 F1 점수 기준으로 79.49%의 성능을 보였으며, 수보요원과 비교하여 심정지 감지 시간을 15초 단축하였다. 이 연구는 소규모 데이터셋을 사용하였음에도 불구하고, 심정지 기반의 탐지 프로그램이 수보요원의 보조 역할로 심정지 생존률에 기여할 수 있음을 입증하였다.

Keywords

Acknowledgement

본 연구는 소방청 연구개발사업인 "119 구급 신고 정보 표준화 및 자료 활용 방안 연구" 과제의 지원을 받아 진행되었습니다. 119 구급신고 데이터는 세종 소방본부, 제주 소방본부, 서울소방본부의 협조를 받아 녹취 데이터 구축을 하였습니다.

References

  1. Berg, K. M., Cheng, A., Panchal, A. R., Topjian, A. A., Aziz, K., Bhanji, F., Bigham, B. L., ... Lavonas, E. J. (2020). Part 7: Systems of Care: 2020 American Heart Association guidelines for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation, 142(16_suppl_2), S580-S604. https://doi.org/10.1161/CIR.0000000000000899
  2. Blomberg, S. N., Folke, F., Ersboll, A. K., Christensen, H. C., Torp-Pedersen, C., Sayre, M. R., Counts, C. R., ... Lippert, F. K. (2019). Machine learning as a supportive tool to recognize cardiac arrest in emergency calls. Resuscitation, 138, 322-329. https://doi.org/10.1016/j.resuscitation.2019.01.015
  3. Bojanowski, P., Grave, E., Joulin, A., & Mikolov, T. (2017). Enriching word vectors with subword information. Transactions of the Association for Computational Linguistics, 5, 135-146. https://doi.org/10.1162/tacl_a_00051
  4. Byrsell, F., Claesson, A., Ringh, M., Svensson, L., Jonsson, M., Nordberg, P., Forsberg, S., ... Nord, A. (2021). Machine learning can support dispatchers to better and faster recognize out-of-hospital cardiac arrest during emergency calls: A retrospective study. Resuscitation, 162, 218-226. https://doi.org/10.1016/j.resuscitation.2021.02.041
  5. Chung, B. H., Shin, J. H., & Cho, U. H. (2015). Development of 119 integrated emergency management system training simulator system. The Journal of the Korea Institute of Electronic Communication Sciences, 10(4), 461-467. https://doi.org/10.13067/JKIECS.2015.10.4.461
  6. Eisenberg. M. S., Horwood, B. T., Cummins, R. O., Reynolds-Haertle, R., & Hearne, T. R. (1990). Cardiac arrest and resuscitation: A tale of 29 cities. Annals of Emergency Medicine, 19(2), 179-186. https://doi.org/10.1016/S0196-0644(05)81805-0
  7. Gnesin, F., Moller, A. L., Mills, E. H. A., Zylyftari, N., Jensen, B., Boggild, H., Ringgren, K. B., ... Torp-Pedersen, C. (2021). Rapid dispatch for out-of-hospital cardiac arrest is associated with improved survival. Resuscitation, 163, 176-183. https://doi.org/10.1016/j.resuscitation.2021.03.015
  8. Kim, H. S., & Chun, J. H. (2020). Determinants on survival outcomes of sudden out-of-hospital cardiac arrest: A multilevel analysis. The Korean Journal of Emergency Medical Services, 24(2), 7-26. https://doi.org/10.14408/KJEMS.2020.24.2.007
  9. Kim, J., Jeong J., & Kweon, S. (2022). Incidences of out-of-hospital sudden cardiac arrest in the Republic of Korea, 2021. Public Health Weekly Report, 15(51), 3007-3020. https://doi.org/10.56786/PHWR.2022.15.51.3007
  10. Kim, J. H., Choi, A., Kim, M. J., Hyun, H., Kim, S., & Chang, H. J. (2022). Development of a machine-learning algorithm to predict in-hospital cardiac arrest for emergency department patients using a nationwide database. Scientific Reports, 12(1), 21797.
  11. Lee, K., Chung, J., Shin, D., Chung, M., Kang, K., Jang, Y., & Jang, K. (2016). Emergency dispatching based on automatic speech recognition. Phonetics and Speech Sciences, 8(2), 31-39. https://doi.org/10.13064/KSSS.2016.8.2.031
  12. New Jersey Government. (2020). Emergency medical dispatch guidecards. Retrieved from https://www.nj.gov/911/home/highlights/EMD%20Guidecards%202020%20Final.pdf
  13. O'Keeffe, C., Nicholl, J., Turner, J., & Goodacre, S. (2011). Role of ambulance response times in the survival of patients with out-of-hospital cardiac arrest. Emergency Medicine Journal, 28(8), 703-706. https://doi.org/10.1136/emj.2009.086363
  14. Oh, S. Y. (2020). 119 reports in 2021, up 7.1% (nearly 800,000) year-over-year. Retrieved from http://m.safetyin.co.kr/news/articleView.html?idxno=15029
  15. Park, I., Kim, Y., & Kang, S. (2014). A study on the step-by-step survival factors of prehospital cardiac arrest patients. Health and Social Welfare Review, 34(2), 484-513. https://doi.org/10.15709/hswr.2014.34.2.484
  16. Park, J. H. (2022). Use of machine learning based time-to-event analysis for prediction of field return of spontaneous circulation after out-of-hospital cardiac arrest (Doctoral dissertation). Seoul National University, Seoul, Korea.
  17. Phoenix Government. (2016). Fire emergency dispatcher: 911 operator. Retrieved from https://www.phoenix.gov/hrsite/documents/fireguide.pdf
  18. Povey, D., Ghoshal, A., Boulianne, G., Burget, L., Glembek, O., Goel, N., Hannemann, M., ..., Vesely, K. (2011). The Kaldi speech recognition toolkit. Proceeding of the IEEE 2011 Workshop on Automatic Speech Recognition and Understanding (No. CONF). Waikoloa, HI.
  19. Safar, P. (1986). Cerebral resuscitation after cardiac arrest: a review. Circulation, 74(6 Pt 2), IV138-53.
  20. Yoon, H., Park, S., Ahn, J., & Lee, Y. (2020). Major findings of the 2006-2018 survey on acute myocardial infarction. Weekly Health and Disease, 13(1), 77-90.
  21. Vukmir, R. B., & Sodium Bicarbonate Study Group. (2004). Witnessed arrest, but not delayed bystander cardiopulmonary resuscitation improves prehospital cardiac arrest survival. Emergency Medicine Journal, 21(3), 370-373. https://doi.org/10.1136/emj.2003.008383