On-Line Extraction of Successive Temporal Sequences from ICU High-Frequency Data for Decision Support Information

  • Sylvie Charbonnier
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2780)


This paper presents a method to extract on line successive temporal sequences from high frequency data monitored in ICU. Successive temporal sequences are expressions such as: “the systolic blood pressure is steady at 120mmHg from time t 0 until time t 1; it is increasing from 120 mmHg to 160mmHg from time t 1 to time t 2 ...”. The method uses a segmentation algorithm that was developed previously and a classification of the segments into temporal patterns. It has seven tuning parameters that are rather easy to tune because they have a physical meaning. The results obtained on simulated data are quite satisfactory. Sequences extracted from real biological data recorded during 14 hours from different patients received the approbation of two clinicians. These temporal sequences can help the health care personnel to take decisions in alarm situations, or can be used as inputs to intelligent alarm systems using inferences on the data.


Segmentation Algorithm Health Care Personnel Successive Sequence Temporal Shape Trend Pattern 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    O’Carrol, T.: Survey of alarms in an intensive therapy units. Anaesthesia 41, 742–744 (1986)CrossRefGoogle Scholar
  2. 2.
    Beneken, J., Van der, A.A.J.: Alarms and their limits in monitoring. J. Clin. Monit. 5, 205–210 (1989)CrossRefGoogle Scholar
  3. 3.
    Coiera, E.: Intelligent monitoring and control of dynamic physiological systems. Artificial Intelligence in Medicine 5, 1–8 (1993)CrossRefGoogle Scholar
  4. 4.
    Uckun, S.: Intelligent systems in patient monitoring and therapy management A survey of research projects. International Journal of Clinical Monitoring and Computing 11, 225–241 (1994)CrossRefGoogle Scholar
  5. 5.
    Steimann, F.: The interpretation of time-varying data with DIAMON-1. Artificial Intelligence in Medicine 8, 343–357 (1996)CrossRefGoogle Scholar
  6. 6.
    Avent, R., Charlton, J.: A critical review of trend-detection methodologies for biomedical systems. Critical Reviews in Biomedical Engineering 17, 621–659 (1990)Google Scholar
  7. 7.
    Haimowitz, I., Phillip, P., Kohane, I.: Clinical monitoring using regression-based trend templates. Artificial Intelligence in Medicine 7, 473–496 (1995)CrossRefGoogle Scholar
  8. 8.
    Shahar, Y.: A framework for knowledge-based temporal abstraction. Artificial Intelligence in Medicine 90, 79–133 (1997)zbMATHGoogle Scholar
  9. 9.
    Hunter, J., McIntosh, N.: Knowledge based event detection in complex time series data. In: Horn, W., Shahar, Y., Lindberg, G., Andreassen, S., Wyatt, J.C. (eds.) AIMDM 1999. LNCS (LNAI), vol. 1620, pp. 271–280. Springer, Heidelberg (1999)CrossRefGoogle Scholar
  10. 10.
    Salatian, A., Hunter, J.R.W.: Deriving trends in historical and real-time continuously sampled medical data. Journal of Intelligent Information Systems 13, 47–71 (1999)CrossRefGoogle Scholar
  11. 11.
    Calvelo, D., Chambrin, M.C., Pomorski, D., Ravaux, P.: Towards symbolisation using data-driven extraction of local trends for ICU monitoring. Artificial Intelligence in Medicine 1-2, 203–223 (2000)CrossRefGoogle Scholar
  12. 12.
    Charbonnier, S., Becq, G., Biot, L., Carry, P., Perdrix, J.P.: Segmentation algorithm for ICU continuously monitored clinical data. In: 15th World IFAC congress (2002)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Sylvie Charbonnier
    • 1
  1. 1.Laboratoire d’Automatique de GrenobleSt Martin d’HèresFrance

Personalised recommendations