Predicting Colorectal Cancer Recurrence: A Hybrid Neural Networks-Based Approach

  • Rob Smithies
  • Said Salhi
  • Nat Queen
Part of the Operations Research/Computer Science Interfaces Series book series (ORCS, volume 32)


This study presents a new expert system for predicting colorectal cancer recurrence that combines a relaxed form of stepwise regression, a new clustering strategy designed to cope with mixed attribute types and missing data, and ensembles of neural networks trained via a new robust global optimisation method. Results show that implementation of the expert system leads to an improvement in the prediction of colorectal cancer recurrence approaching levels of accuracy required by medical practitioners for clinical use.


Metaheuristic expert system regression K-means missing data neural network ensemble RPROP local search tabu search 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Aarts, E. and Lenstra, J. (1997). Local Search in Combinatorial Optimization. John Wiley and Sons, Chichester, UK.Google Scholar
  2. Barr, L., Cowan, R., and Nicolson, M. (1999). Oncology. Churchill Livingstone, London, UK.Google Scholar
  3. Bhatavdekar, J., Patel, D., Chikhlikar, P., Shah, N., Vora, H., Ghosh, N., and Trivedi, T. (2001). Molecular markers are predictors of recurrence and survival in patients with Dukes B and Dukes C colorectal adenocarcinoma. Diseases of the Colon & Rectum, 44:523–533.Google Scholar
  4. Cancer Guidance Sub-group of the Clinical Outcomes Group (1997). Improving outcomes in colon cancer: The Research Evidence. Technical report, NHS Executive, Department of Health, London, UK.Google Scholar
  5. Draper, N. and Smith, H. (1981). Applied Regression Analysis (2nd edition). John Wiley and Sons, New York, NY.Google Scholar
  6. Grumett, S., Quirke, P., Kerr, D., McConkey, C., Stahlsmidt, M., and Barnwell, J. (2002). A pathology sub-study of putative prognostic markers in colorectal cancer. In Proceedings of the International Symposium on Predictive Oncology and Intervention Strategies, Paris, France.Google Scholar
  7. Igel, C. and Hüsken, M. (2000). Improving the Rprop Learning Algorithm. In Proceedings of the Second International Symposium on Neural Computation, NC2000, pages 115–121, Berlin, Germany. ICSC Academic Press.Google Scholar
  8. IMPACT (1995). Efficacy of adjuvant fluorouracil and folinic acid in colon cancer. Lancet, 345:939–944.Google Scholar
  9. Le Cun, Y. (1989). Generalization and Network Design Strategies. Technical report, CRG-TR-89-4. Department of Computer Science, University of Toronto, Toronto, Canada.Google Scholar
  10. Linde, Y., Buzo, A., and Gray, R. (1980). The K-means algorithm. IEEE Transactions on Communication, 28:84.CrossRefGoogle Scholar
  11. Medsker, L. (1994). Hybrid Neural Network and Expert Systems. Kluwer Academic Publishers, Norwell, MA.Google Scholar
  12. Papachristodoulou, A., Kouskos, E., Markopoulos, C., Karatzas, G., Kouraklis, G., and Kostakis, A. (1987). Local recurrence after radical surgery for colorectal cancer. Physical Review Letters, 59:2229–2232.CrossRefPubMedGoogle Scholar
  13. Patterson, D. (1996). Artificial Neural Networks: Theory and Applications. Prentice Hall, Singapore.Google Scholar
  14. Prechelt, L. (1994). PROBEN1 — A set of benchmarks and bench-marking rules for neural network training algorithms. Technical report, 21/94, Fakultät für Informatik, Universität Karlsruhe, Karlsruhe, Germany.Google Scholar
  15. Riedmiller, M. and Braun, H. (1993). A direct adaptive method for faster backpropagation learning: The RPROP algorithm. In Proceedings of the IEEE International Conference on Neural Networks (ICNN), volume 16, pages 586–591, Piscataway, NJ. IEEE Press.Google Scholar
  16. Rumelhart, D. and McClelland, J. (1986). Parallel Distributed Processing. MIT Press, Cambridge, MA.Google Scholar
  17. Salhi, S. (1998). Heuristic Search Methods: in Modern Methods for Business Research (G. Marcoulides, Ed.), chapter 6, pages 147–175. Lawrence Erlbaum Associates.Google Scholar
  18. Salhi, S. and Queen, N. (2004). A hybrid algorithm for detecting global and local minima when optimizing functions with many minima. European Journal of Operational Research, 155:51–67.CrossRefMathSciNetGoogle Scholar
  19. Silva, F. and Almeida, L. (1990). Advanced Neural Computers, pages 151–158. North-Holland.Google Scholar
  20. Smithies, R. (2004). Neural Network Heuristics for Real-World Classification with an Application to Predict Cancer Recurrence. PhD Thesis, University of Birmingham, Birmingham, UK.Google Scholar
  21. Smithies, R., Salhi, S., and Queen, N. (2004). Adaptive Hybrid Learning for Neural Networks. Neural Computation, 16:139–157.CrossRefPubMedGoogle Scholar
  22. United Kingdom Co-ordinating Committee on Cancer Research (1998). QUASAR1 (Quick and Simple and Reliable): a UKCCCR study of adjuvant chemotherapy for colorectal cancer. Technical report, Clinical Trials Unit, Institute of Clinical Research, The Medical School, University of Birmingham, Birmingham, UK.Google Scholar

Copyright information

© Springer Science+Business Media, Inc. 2005

Authors and Affiliations

  • Rob Smithies
    • 1
  • Said Salhi
    • 1
  • Nat Queen
    • 1
  1. 1.School of Mathematics and StatisticsUniversity of BirminghamEdgbastonUK

Personalised recommendations