Evaluation of Prognostic Factors and Prediction of Chronic Wound Healing Rate by Machine Learning Tools

  • Marko Robnik-Šikonja
  • David Cukjati
  • Igor Kononenko
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2101)


In more than a decade of clinical use of electrical stimulation to accelerate the chronic wound healing each patient and wound were registered and a wound healing process was weekly followed. The controlled study involved a conventional conservative treatment, sham treatment, biphasic pulsed current, and direct current electrical stimulation. A quantity of available data suffices for an analysis with machine learning methods.

So far only a limited number of studies have investigated the wound and patient attributes which affect the chronic wound healing. There is none to our knowledge to include the treatment attributes. The aims of our study are to determine effects of the wound, patient and treatment attributes on the wound healing process and to propose a system for prediction of the wound healing rate.

In the first step of our analysis we determined which wound and patient attributes play a predominant role in the wound healing process. Then we investigated a possibility to predict the wound healing rate at the beginning of the treatment based on the initial wound, patient and treatment attributes. Finally we discussed the possibility to enhance the wound healing rate prediction accuracy by predicting it after a few weeks of the wound healing follow-up.

By using the attribute estimation algorithms ReliefF and RReliefF we obtained a ranking of the prognostic factors which was comprehensi- ble to field experts. We also used regression and classification trees to build models for prediction of the wound healing rate. The obtained results are encouraging and may form a basis of an expert system for the chronic wound healing rate prediction. If the wound healing rate is known, then the provided information can help to formulate the appro- priate treatment decisions and orient resources to those individuals with poor prognosis.


Wound Healing Spinal Cord Injury Regression Tree Wound Healing Process Treatment Attribute 
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. J.A. Birke, A. Novick, C. A. Patout, and W. C. Coleman. Healing rates of plantar ulcers in leprosy and diabetes. Leprosy Rev., 63:365–374, 1992.Google Scholar
  2. L. Breiman, J. H. Friedman, R. A. Olshen, and C. J. Stone. Classification and regression trees. Wadsworth Inc., Belmont, California, 1984.zbMATHGoogle Scholar
  3. D. Cukjati, R. Karba, S. Reberšek, and D. Miklavčič. Modeling of chronic wound healing dynamics. Med.Biol.Eng.Comput., 38:339–347, 2000.CrossRefGoogle Scholar
  4. D. Cukjati, S. Reberšek, and D. Miklavčič. A reliable method of determining wound healing rate. Med.Biol.Eng.Comput., 2001a. (in press).Google Scholar
  5. D. Cukjati, M. Robnik-Šikonja, S. Reberšek, I. Kononenko, and D. Miklavčič. Prognostic factors, prediction of chronic wound healing and electrical stimulation. Medical & Biological Engineering & Computing, 2001b. (submitted).Google Scholar
  6. T.G. Dietterich. Machine learning research: Four current directions. AI Magazine, 18(4):97–136, 1997.Google Scholar
  7. J.A. Feedar and L. C. Kloth. Conservative management of chronic wounds. In L. C. Kloth and J.M. McCulloch, editors, Wound Healing: Alternatives in Management, pages 135–172. F. A. Davis Co., Philadelphia, 1990.Google Scholar
  8. A. Jerčinovič, R. Karba, L. Vodovnik, A. Stefanovska, P. Krošelj, R. Turk, I. Džidič, H. Benko, and R. Šavrin. Low frequency pulsed current and pressure ulcer healing. IEEE Trans.Rehab.Eng., 2(4):225–233, 1994.CrossRefGoogle Scholar
  9. M. Johnson. Using cluster analysis to develop a healing typology in vascular ulcers. J.Vasc.Nurs., 15:45–49, 1997.CrossRefGoogle Scholar
  10. R. Karba, D. Šemrov, L. Vodovnik, H. Benko, and R. Šavrin. DC electrical stimulation for chronic wound healing enhancement. Part 1. Clinical study and determination of electrical field distribution in the numerical wound model. Bioelectrochemistry and Bioenergetics, 43:265–270, 1997.CrossRefGoogle Scholar
  11. I. Kononenko, E. Šimec, and M. Robnik-Šikonja. Overcoming the myopia of inductive learning algorithms with RELIEFF. Applied Intelligence, 7:39–55, 1997.CrossRefGoogle Scholar
  12. I.R. Lyman, J. H. Tenery, and R. P. Basson. Corelation between decrease in bacterial load and rate of wound healing. Surg.Gynecol.Obstet., 130(4):616–620, 1970.Google Scholar
  13. J.R. Quinlan. C4.5: Programs for Machine Learning. Morgan Kaufmann, 1993.Google Scholar
  14. M. Robnik Šikonja. CORE — a system that predicts continuous variables. In Proceedings of Electrotehnical and Computer Science Conference, pages B145–148, 1997.Google Scholar
  15. M. Robnik Šikonja and I. Kononenko. An adaptation of Relief for attribute estimation in regression. In Douglas H. Fisher, editor, Machine Learning: Proceedings of the Fourteenth International Conference (ICML’97), pages 296–304. Morgan Kaufmann, 1997.Google Scholar
  16. Marko Robnik-Šikonja and Igor Kononenko. Comprehensible interpretation of Relief’s estimates. 2001. (submitted).Google Scholar
  17. D.J. Shea. Pressure sores classification and management. Clin.Orthop.Rel.Res., 112: 89–100, 1975.Google Scholar
  18. A.I. Skene, J. M. Smith, C. J. Doré, A. Charlett, and J. D. Lewis. Venous leg ulcers: a prognostic index to predict time to healing. BMJ, 305:1119–1121, 1992.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2001

Authors and Affiliations

  • Marko Robnik-Šikonja
    • 1
  • David Cukjati
    • 2
  • Igor Kononenko
    • 1
  1. 1.Faculty of Computer and Information ScienceUniversity of LjubljanaLjubljanaSlovenia
  2. 2.Faculty of Electrical EngineeringUniversity of LjubljanaLjubljanaSlovenia

Personalised recommendations