Journal of the Institute of Electronics Engineers of Korea SC (전자공학회논문지SC)

The Institute of Electronics and Information Engineers (대한전자공학회)
권호 표지

Comparison of Survival Prediction of Rats with Hemorrhagic Shocks Using Artificial Neural Network and Support Vector Machine

출혈성 쇼크를 일으킨 흰쥐에서 인공신경망과 지원벡터기계를 이용한 생존율 비교

  • Jang, Kyung-Hwan (Graduate Program in Biomedical Engineering, Yonsei University) ;
  • Yoo, Tae-Keun (Yonsei Uinversity College of Medicine) ;
  • Nam, Ki-Chang (Clinical Trials Center for Medical Devices, Severance Hospital) ;
  • Choi, Jae-Rim (Graduate Program in Biomedical Engineering, Yonsei University) ;
  • Kwon, Min-Kyung (Brain Korea 21 Project for Medical Science, Yonsei University) ;
  • Kim, Deok-Won (Dept. of Medical Engineering, Yonsei Uinversity College of Medicine)
  • 장경환 (연세대학교 생체공학협동과정) ;
  • 유태근 (연세대학교 의학전문대학원) ;
  • 남기창 (세브란스병원 의료기기 임상시험 센터) ;
  • 최재림 (연세대학교 생체공학협동과정) ;
  • 권민경 (연세대학교 의과학과) ;
  • 김덕원 (연세대학교 의과대학 의학공학교실)
  • Received : 2010.10.22
  • Accepted : 2011.03.10
  • Published : 2011.03.25
Download PDF
⟨ Previous Next ⟩

Abstract

Hemorrhagic shock is a cause of one third of death resulting from injury in the world. Early diagnosis of hemorrhagic shock makes it possible for physician to treat successfully. The objective of this paper was to select an optimal classifier model using physiological signals from rats measured during hemorrhagic experiment. This data set was used to train and predict survival rate using artificial neural network (ANN) and support vector machine (SVM). To avoid over-fitting, we chose the best classifier according to performance measured by a 10-fold cross validation method. As a result, we selected ANN having three hidden nodes with one hidden layer and SVM with Gaussian kernel function as trained prediction model, and the ANN showed 88.9 % of sensitivity, 96.7 % of specificity, 92.0 % of accuracy and the SVM provided 97.8 % of sensitivity, 95.0 % of specificity, 96.7 % of accuracy. Therefore, SVM was better than ANN for survival prediction.

전 세계적으로 상해로 인한 사망자 중 1/3의 직접적인 사망 원인은 출혈성 쇼크이다. 그러나 초기 쇼크에서 이를 정확히 예측할 수 있다면 신속한 치료가 가능하여 그 피해를 줄일 수 있다. 본 논문의 목적은 흰쥐의 대퇴부정맥을 통해 일정량의 출혈을 시키면서 변화하는 생리적 변수인 심박수, 수축기 혈압, 평균 동맥압, 호흡수, 체온 데이터로 최적의 생존 예측 모델을 제시하여 출혈성 쇼크를 조기 진단하는 것이다. 예측 모델로는 최근 많이 연구되는 인공신경망과 지원벡터기계 방법을 사용하였다. 과대적합을 피하고 최적의 모델을 선정하기 위해 10-fold cross validation을 수행하였을 때, 인공신경망의 경우 은닉노드 수가 3개인 모델이 가장 우수한 성능을 보였고, 지원벡터기계에서는 가우시안 커널함수를 이용한 모델이 가장 우수한 성능을 보였다. 평가 데이터 세트를 이용하여 각각의 생존 예측 모델을 평가한 결과 인공신경망의 경우 민감도 88.9 %, 특이도 96.7 %와 정확도 92.0 %를 보였고, 지원벡터기계의 경우 민감도 97.8 %, 특이도 95.0 %와 정확도 96.7 %를 보였다. 따라서 출혈에 따른 흰쥐의 생존 예측에서 지원벡터기계가 인공신경망보다 더 우수한 성능을 보이는 것을 확인하였다.

Keywords

References

  1. World Health Organization, World health statistics 2010, World Health Organization Press, pp. 62-70, 2010.
  2. CJ Murray and AD Lopez, "Alternative projections of mortality and disability by cause 1990-2020: Global Burden of Disease Study," The Lancet, Vol. 349, no. 9064, pp.1498-1504, May. 1997. https://doi.org/10.1016/S0140-6736(96)07492-2
  3. CD Deakin and IR Hicks, "AB or ABC: pre-hospital fluid management in major trauma," Journal of Accident and Emergency Medicine, Vol. 11, no. 3, pp. 154-157, Mar. 1994. https://doi.org/10.1136/emj.11.3.154
  4. I. B. Chang, "Analysis of causes of death of the patients with multiple traumas who died in emergent department," Annu. Meet Korea Neurosurg Soc, Seoul, Korea, Nov. 2007
  5. J. Markovchick and T. Pons, Emergency medicine secrets, 4th edition, Philadelphia, pp. 28-32, 2006.
  6. Martel MJ, MacKinnon KJ, Arsenault MY, Bartellas E, Klein MC, Lane CA, Sprague AE, and Wilson AK, "Hemorrhagic shock," Journal of Obstetrics and Gynaecology Canada, Vol. 24, no. 6, pp. 504-511, Jun. 2002. https://doi.org/10.1016/S1701-2163(16)31098-2
  7. M. Wilson, DP. Davis and R. Coimbra, "Diagnosis and monitoring of hemorrhagic shock during the initial resuscitation of multiple trauma patients: a review," Journal of Emergency Medicine, Vol. 24, no. 4, pp. 413-422, May. 2003. https://doi.org/10.1016/S0736-4679(03)00042-8
  8. J. H. Lee, T. K. Lee, S. W. Chung, H. W. No and D. W. Kim, "Changes in heart rate and heart rate variability induced by hemorrhagic shock in rats shock," Shock, Vol. 31, Suppl 1, pp. 21, Jun. 2009. https://doi.org/10.1097/SHK.0b013e31817d43bf
  9. D. W. Kim, J. L. Choi and Y. S. Park, "Survival prediction in rats with fixed-volum hemorrhage using a logistic regression equation," Shock, Vol. 33, Suppl 1, pp. 14, Jun. 2010.
  10. J. L. Choi, D. W. Kim and Y. S. Park, "Prediction of blood loss in rats during fixed volume hemorrhage using a regression model," Shock, Vol. 33, Suppl 1, pp. 14, Jun. 2010.
  11. X. Li and F. Porikli, "Human state classification and Prediction for Critical Monitoring by real-time bio-signal Analysis," International Conference on Pattern Recognition, Istanbul, Turkey, Aug. 2010.
  12. S. Y. Ji, W. Chen, K. Ward, C. Rickards, K. Ryan, V. Convertino and K. Najarian, "Wavelet based analysis of physiological signals for predictionof severity of hemorrhagic shock," International Conference on Complex Medical Engineering, pp. 1-6, Tempe, USA, April. 2009.
  13. B. Eftekhar, K. Mohammad, HE. Ardebili, M. Ghodsi and E. Ketabchi, "Comparison of artificial neural network and logistic regression models for prediction of mortality in head trauma based on initial clinical data," BMC Medical Informatics and Decision Making, Vol. 5, no. 3, pp. 1-8, Feb. 2005. https://doi.org/10.1186/1472-6947-5-1
  14. SM. Dirusso, T. Sullivan, C. Holly, SN. Cuff and J. Savino, "An artificial neural network as a model for prediction of survival in trauma patients: validation for a regional trauma area," Journal of Trauma, Vol. 49, no. 2 pp. 212-223, Aug 2000. https://doi.org/10.1097/00005373-200008000-00006
  15. SM. Dirusso, AA. Chahine, T. Sullivan, D. Risucci, P. Nealon and S. Cuff, "Development of a model for prediction of survival in pediatric trauma patients: comparision of artificial neural networks and logistic regression," Journal of Pediatric Surgery, Vol. 37, no. 7, pp. 1098-1104, July. 2002. https://doi.org/10.1053/jpsu.2002.33885
  16. 주형, 최재림, 정상원, 김덕원, "출혈성 쇼크를 일으킨 흰쥐에서 인공신경망을 이용한 생존율 예측," 응급의학회지, 제21권, 제3호, 1-7쪽, 2010년 6월.
  17. G. Gutierrez, H. D. Reines and M. E. Wulf-Gutierrez, "Clinical review: hemorrhagic shock," Critical Care, Vol. 8, no. 5, pp. 373-381, Oct. 2004. https://doi.org/10.1186/cc2851
  18. J. Heaton, Introduction to neural networks for JAVA, Second Edition, Heaton Research, pp. 158-159, 2008.
  19. P. N. Tan, M. Steinbch and V. Kumar, Introduction to data mining, Addison Wesley, pp. 250-271, 2005.
  20. I.A. Basheer and M. Hajmeer, "Artificial neural networks: fundamentals, computing, design, and application," Journal of Microbiological Methods, Vol. 43, no. 1, pp. 3-31, Dec. 2000. https://doi.org/10.1016/S0167-7012(00)00201-3
  21. S. Y. Peng, K. C. Wu, J. J. Wang, J. H. Chung, S. K. Peng and Y. H. Lai, "Predicting postoperative nausea and vomiting with the application of an artificial neural network," British Journal of Anaesthesia, Vol. 98, no. 1. pp. 60-65, Aug. 2007. https://doi.org/10.1093/bja/ael282