Advertisement

Health Monitoring Systems Using Machine Learning Techniques

  • Fahmi Ben Rejab
  • Kaouther Nouira
  • Abdelwahed Trabelsi
Part of the Studies in Computational Intelligence book series (SCI, volume 542)

Abstract

This paper describes the steps of building two and efficient monitoring systems in intensive care unit (ICU). We propose two new systems that deal with large data sets and solves the main problems of the current monitoring system. In fact, the current monitoring system in ICU has many issues to detect real states of patients namely critical and normal states. It frequently generates a high number of false alarms having bad effects on the working conditions. Besides, these alarms can threat the patient life by misleading medical staff. Our aim, in this paper, is to avoid false alarms and keep a high level of sensitivity by improving the current monitoring system. In addition, our purpose is to generate groups of patients suffering from similar diseases and building a general model for similar patients. The obtained models will make the classification of new patients possible. To this end, we combine two incremental versions of support vector machines mainly the LASVM and ISVM techniques with the k-prototypes clustering method. The first proposed system is the KP-ISVM which is based on the ISVM technique and the k-prototypes. The second one is called KP-LASVM which takes profits of both the LASVM by reducing the false alarms and the k-prototypes by selecting the appropriate model to start classifying new patients. Both of our proposals are characterized by dealing with large amount of data streams and adding new patients. However, the system using LASVM and k-prototypes i.e. the KP-LASVM has produced the best results compared to the others monitoring systems and based on different evaluation criteria. All experimental results using real-medical databases have been analyzed and have proved the performance of the KP-LASVM.

Keywords

Intensive care unit monitoring system support vector machines LASVM classification k-prototypes 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Baxter, G.D., Monk, A.F., Tan, K., Dear, P.R., Newell, S.J.: Using cognitive task analysis to facilitate the integration of decision support systems into the neonatal intensive care unit. Artificial Intelligence in Medicine 35, 243–257 (2005)CrossRefGoogle Scholar
  2. 2.
    Ben Rejab, F., Nouira, K.: Reducing False Alarms in Intensive Care Units Monitoring System Using Support Vector Machines. CCCM 2010 4, 106–109 (2010)Google Scholar
  3. 3.
    Ben Rejab, F., Nouira, K., Trabelsi, A.: Support Vector Machines versus Multi-layer Perceptrons for Reducing False Alarms in Intensive Care Units. International Journal of Computer Applications, Foundation of Computer Science 49, 41–47 (2012)Google Scholar
  4. 4.
    Ben Rejab, F., Nouira, K., Trabelsi, A.: On the use of the incremental support vector machines for monitoring systems in intensive care unit. In: TAEECE 2013, pp. 266–270 (2013)Google Scholar
  5. 5.
    Ben Rejab, F., Nouira, K., Trabelsi, A.: Monitoring Systems in Intensive Care Units using Incremental Support Vector Machines. In: ICDIPC 2013, pp. 273–280 (2013)Google Scholar
  6. 6.
    Ben Rejab, F., Nouira, K., Trabelsi, A.: Incremental Support Vector Machines for Monitoring Systems in Intensive Care Unit. In: SAI 2013, pp. 496–501 (2013)Google Scholar
  7. 7.
    Bordes, A., Bottou, L.: The Huller: a simple and efficient online svm. In: Gama, J., Camacho, R., Brazdil, P.B., Jorge, A.M., Torgo, L. (eds.) ECML 2005. LNCS (LNAI), vol. 3720, pp. 505–512. Springer, Heidelberg (2005)CrossRefGoogle Scholar
  8. 8.
    Bordes, A., Ertekin, S., Weston, J., Bottou, J.: Fast Kernel Classifiers With Online And Active Learning. Journal of Machine Learning Research 6, 1579–1619 (2005)MATHMathSciNetGoogle Scholar
  9. 9.
    Borowski, M., Siebig, S., Wrede, C., Imhoff, M.: Reducing false alarms of intensive care online monitoring systems: An evaluation of two signal extraction algorithms. In: Computational and Mathematical Methods in Medicine, vol. 2011 (2011)Google Scholar
  10. 10.
    Cauwenberghs, G., Poggio, T.: Incremental and Decremental Support Vector Machine Learning, pp. 409–415 (2000)Google Scholar
  11. 11.
    Chambrin, M.C.: Alarms in the intensive care unit: how can the number of false alarms be reduced. Journal of Critical Care 5, 184–188 (2001)CrossRefGoogle Scholar
  12. 12.
    Chang, C., Lin, C.: LIBSVM: A library for support vector machines. Transactions on Intelligent Systems and Technology 2, 27:1–27:27 (2011)Google Scholar
  13. 13.
    Charbonnie, S., Gentil, S.: A trend-based alarm system to improve patient monitoring in intensive care units. Control Engineering Practice 15, 1039–1050 (2007)CrossRefGoogle Scholar
  14. 14.
    Chen, Y.W., Lin, C.J.: Combining SVMs with various feature selection strategies, http://www.csie.ntu.edu.tw/cjlin/papers/features.pdf
  15. 15.
    Cortes, C., Vapnik, V.: Support vector networks. Machine Learning 20, 273–297 (1995)MATHGoogle Scholar
  16. 16.
    Huang, Z.: Clustering large data sets with mixed numeric and categorical values. In: Proceedings of the 1st Pacific-Asia Conference on Knowledge Discovery and Data Mining, pp. 21–34 (1997)Google Scholar
  17. 17.
    Ivanciuc, O.: Applications of Support Vector Machines in Chemistry. Rev. Comput. Chem. 23, 291–400 (2007)Google Scholar
  18. 18.
    Luo, J., Pronobis, A., Caputo, B., Jensfelt, P.: Incremental learning for place recognition in dynamic environments. In: Proc. IROS 2007 (2007)Google Scholar
  19. 19.
    Ming, H.T., Vojislav, K.: Gene extraction for cancer diagnosis by support vector machines an improvement. Artificial Intelligence in Medicine 35, 185–194 (2005)CrossRefGoogle Scholar
  20. 20.
    MacQueen, J.B.: Some methods for classification and analysis of multivariate observations. In: Proceeding of the Fifth Berkeley Symposium on Math., Stat. and Prob., pp. 281–296 (1967)Google Scholar
  21. 21.
    Nouira, K., Trabelsi, A.: Intelligent monitoring system for intensive care units. Journal of Medical Systems 36, 2309–2318 (2011)CrossRefGoogle Scholar
  22. 22.
    Passerini, A., Lippi, M., Frasconi, P.: Predicting Metal-Binding Sites from Protein Sequence. Transactions on Computational Biology and Bioinformatics 9, 203–213 (2012)CrossRefGoogle Scholar
  23. 23.
    Reslan, Z.A.: Clinical alarm management and noise reduction in hospitals. University of Connecticut, Storrs (2007)Google Scholar
  24. 24.
    Siebig, S., Kuhls, S., Imhoff, M., Langgartner, J., Reng, M., Scholmerich, J., Gather, U., Wrede, C.E.: Collection of annotated data in a clinical validation study for alarm algorithms in intensive care-a methodologic framework. Journal of Critical Care 25, 128–135 (2010)CrossRefGoogle Scholar
  25. 25.
    Tong, T., Koller, D.: Support Vector Machine Active Learning with Applications to Text Classification. Journal of Machine Learning Research, 45–66 (2001)Google Scholar
  26. 26.
    Tsien, C.: Reducing False Alarms in the Intensive Care Unit: A Systematic Comparison of Four Algorithms. In: Proceedings of the AMIA Annual Fall Symposium. American Medical Informatics Association (1997)Google Scholar
  27. 27.
    Zhang, S.W., Pan, Q., Zhang, H.C., Zhang, Y.L., Wang, H.Y.: Classification of protein quaternary structure with support vector machine. Bioinformatics 19, 2390–2396 (2003)CrossRefGoogle Scholar
  28. 28.
    Zhang, S.W., Vucetic, S.: Online Training on a Budget of Support Vector Machines Using Twin Prototypes. Statistical Analysis and Data Mining 3, 149–169 (2010)CrossRefMathSciNetGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  • Fahmi Ben Rejab
    • 1
  • Kaouther Nouira
    • 1
  • Abdelwahed Trabelsi
    • 1
  1. 1.BESTMOD, Institut Supérieur de Gestion de TunisUniversité de TunisLe BardoTunisie

Personalised recommendations