Annals of Biomedical Engineering

, Volume 28, Issue 11, pp 1362–1369

EEG Spike Detection With a Kohonen Feature Map

  • C. Kurth
  • F. Gilliam
  • B. J. Steinhoff
Article

Abstract

Artificial neural networks are widely used for pattern recognition tasks. For spike detection in electroencephalography (EEG), feedforward networks trained by the backpropagation algorithm are preferred by most authors. Opposed to this, we examined the off-line spike detection abilities of a Kohonen feature map (KFM), which is different from feedforward networks in certain aspects. The EEG data for the training set were obtained from patients with intractable partial epilepsies of mesiotemporal (n = 2) or extratemporal (n = 2) origin. For each patient the training set for the KFM included the same patterns of background activity and artifacts as well as the typical individual spike patterns. Three different-sized networks were examined (15 × 15 cells, 25 × 25 cells, and 60 × 60 cells in the Kohonen layer). To investigate the quality of spike detection the results obtained with the KFM were compared with the findings of two board-certified electroencephalographers. Application of a threshold based on the partial invariance of spike recognition against translation of the EEG provided an average sensitivity and selectivity of 80.2% at crossover threshold (71%–86%) depending on the networksize and noise. Multichannel EEG processing in real time will be available soon. In conclusion, pattern-based automated spike detection with a KFM is a promising approach in clinical epileptology and seems to be at least as accurate as other more-established methods of spike detection. © 2000 Biomedical Engineering Society.

PAC00: 8719Nn, 8780-y

Electroencephalography Epilepsy Neural networks Spike detection Kohonen feature map 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

REFERENCES

  1. 1.
    Barry, E., N. M. Sussman, M. J. O'Connor, and R. N. Harner. Presurgical electroencephalographic patterns and outcome from anterior temporal lobectomy. Arch. Neurol. 49:21-27, 1992.Google Scholar
  2. 2.
    Chee, M. W. L., H. H. Morris III, M. A. Antar, P. C. Van Ness, D. S. Dinner, P. Rehm, and V. Salanova. Presurgical evaluation of temporal lobe epilepsy using interictal temporal spikes and positrone emission tomography. Arch. Neurol. 50:45-48, 1993.Google Scholar
  3. 3.
    Chung, M. Y., T. S. Walczak, D. V. Lewis, D. V. Dawson, and R. Radtke. Temporal lobectomy and independent bitemporal interictal activity: What degree of lateralization is sufficient? Epilepsia 32:195-201, 1991.Google Scholar
  4. 4.
    Davey, B. L. K., W. R. Fright, G. J. Caroll, and R. D. Jones. Expert system approach to detection of epileptiform activity in the EEG. Med. Biol. Eng. Comput. 27:365-370, 1989.Google Scholar
  5. 5.
    Dingle, A. A., R. D. Jones, G. J. Caroll, and W. R. Fright. A multistage system to detect epileptiform activity in the EEG. IEEE Trans. Biomed. Eng. 40:1260-1268, 1993.Google Scholar
  6. 6.
    Dreyfus, H. L., and S. E. Dreyfus. Mind Over Machine. New York: The Free Press, 1985.Google Scholar
  7. 7.
    Gabor, A. J., and M. Seyal. Automated interictal EEG spike detection using artificial neural networks. Electroencepha-logr. Clin. Neurophysiol. 83:271-280, 1992.Google Scholar
  8. 8.
    Gilliam, F., S. Bowling, E. Bilir, J. Thomas, E. Faught, R. Morawetz, C. Palmer, J. Hugg, and R. Kuzniecky. Association of combined MRI, interictal EEG and ictal EEG results with outcome and pathology after temporal lobectomy. Epilepsia 38:1315-1320, 1998.Google Scholar
  9. 9.
    Glover, Jr., J. R., P. Y. Ktonas, N. Raghavan, J. M. Urunñela, S. S. Velamuri, and E. L. Reilly. A multichannel signal processor for the detection of epileptogenic sharp transients in the EEG. IEEE Trans. Biomed. Eng. 33:1121-1128, 1986.Google Scholar
  10. 10.
    Glover, Jr., J. R., N. Raghavan, P. Y. Ktonas, and J. D. Frost, Jr. Context-based automated detection of epileptogenic sharp transients in the EEG: Elimination of false positives. IEEE Trans. Biomed. Eng. 36:519-527, 1989.Google Scholar
  11. 11.
    Godoy, J., H. Lüders, D. S. Dinner, H. H. Morris III., E. Wyllie, and D. Murphy. Significance of sharp waves in rou-tine EEGs after epilepsy surgery. Epilepsia 33:285-288, 1992.Google Scholar
  12. 12.
    Gotman, J., and P. Gloor. Automatic recognition and quanti-fication of interictal epileptic activity in the human scalp EEG. Electroencephalogr. Clin. Neurophysiol. 41:513-529, 1976.Google Scholar
  13. 13.
    Gotman, J., J. R. Ives, and R. Gloor. Automatic recognition of interictal epileptic activity in prolonged EEG recordings. Electroencephalogr. Clin. Neurophysiol. 46:510-520, 1979.Google Scholar
  14. 14.
    Gotman, J., and L. Y. Wang. State-dependent spike detection: Concepts and preliminary results. Electroencephalogr. Clin. Neurophysiol. 79:11-19, 1991.Google Scholar
  15. 15.
    Gotman, J., and L. Y. Wang. State-dependent spike detection: Validation. Electroencephalogr. Clin. Neurophysiol. 83:12-18, 1992.Google Scholar
  16. 16.
    Jando, G., R. M. Siegel, Z. Horvath, and G. Buzsaki. Pattern recognition of the electroencephalogram by artificial neural networks. Electroencephalogr. Clin. Neurophysiol. 86:100-109, 1993.Google Scholar
  17. 17.
    Kanner, A. M., H. H. Morris III., H. O. Lüders, D. S. Dinner, P. Van Ness, and E. Wyllie. Usefulness of unilateral interictal sharp waves of temporal lobe origin in prolonged video-EEG monitoring studies. Epilepsia 34:884-889, 1993.Google Scholar
  18. 18.
    Kohonen, T. Analysis of a simple self-organizing process. Biol. Cybern. 40:135-140, 1982.Google Scholar
  19. 19.
    Kohonen, T. Self-Organization and Associative Memory. New York: Springer, 1984.Google Scholar
  20. 20.
    Kohonen, T. The self-organizing map. Proc. IEEE 78:1464-1480, 1990.Google Scholar
  21. 21.
    Kohonen, T. Self-Organizing Maps, 2nd ed. New York: Springer, 1997.Google Scholar
  22. 22.
    Morris III., H. H., A. Kanner, H. O. Lüders, D. Murphy, D. S. Dinner, E. Wyllie, and P. Kotagal. Sharp waves at the sphenoidal electrode accurately identify a mesio-temporal epileptogenic focus? Epilepsia 30:532-539, 1989.Google Scholar
  23. 23.
    Özdamar, O., I. Yaylali, P. Jayakar, and C. N. Lopez. Multilevel neural network system for EEG spike detection. In: Computer-Based Medical Systems. Proceedings of the 4th Annual IEEE Symposium, edited by I. N. Bankmann and J. E. Tsitlik. Washington, DC: IEEE Computer Society, 1991, pp. 272-279.Google Scholar
  24. 24.
    Peltoranta, M., and G. Pfurtscheller. Neural network-based classification of nonaveraged event-related EEG responses. IIG Report No. 328, 1992.Google Scholar
  25. 25.
    Pilcher, W. H., D. W. Roberts, H. F. Flanigin, P. H. Crandall, H. G. Wieser, G. A. Ojemann, and W. J. Peacock. Complications of epilepsy surgery. In: Surgical Treatment of the Epilepsies, 2nd ed., edited by J. Engel, Jr. New York: Raven, 1993, pp. 565-581.Google Scholar
  26. 26.
    Risinger, M. W., J. Engel, Jr., P. C. Van Ness, T. R. Henry, and P. H. Crandall. Ictal localization of temporal lobe seizures with scalp/sphenoidal recordings. Neurology 39:1288-1293, 1989.Google Scholar
  27. 27.
    27 Rumelhart, D. E., G. E. Hinton, and R. J. Williams. Learning internal representation by error propagation. In: Parallel Distributed Processing, Vol. I, edited by D. E. Rumelhard and J. L. McCalland. Cambridge, MA: MIT, 1986.Google Scholar
  28. 28.
    Spatt, J., G. Pelzl, and B. Mamoli. Reliability of automatic and visual analysis of interictal spikes in lateralizing an epileptic focus during video-EEG monitoring. Electroencephalogr. Clin. Neurophysiol. 103:421-425, 1997.Google Scholar
  29. 29.
    29 Steinhoff, B. J., N. K. So, S. Lim, and H. O. Lüders. Ictal scalp EEG in temporal lobe epilepsy with unitemporal versus bitemporal interictal epileptiform discharges. Neurology 45:889-896, 1995.Google Scholar
  30. 30.
    Ultsch, A. Self-organized feature maps for monitoring and knowledge acquisition of a chemical process. Proceedings of the International Conference on Artificial Neural Networks, 1993 (unpublished).Google Scholar
  31. 31.
    Webber, W. R. S., B. Litt, R. P. Lesser, R. S. Fisher, and I. Bankman. Automatic EEG spike detection: what should the computer imitate. Electroencephalogr. Clin. Neurophysiol. 87:364-373, 1993.Google Scholar
  32. 32.
    Webber, W. R. S., B. Litt, K. Wilson, and R. P. Lesser. Practical detection of epileptiform discharges (EDs) in the EEG using an artificial neural network: A comparison of raw and parameterized EEG data. Electroencephalogr. Clin. Neurophysiol. 91:194-204, 1994.Google Scholar
  33. 33.
    Wilson, K., W. R. S. Webber, R. P. Lesser, R. S. Fisher, R. C. Eberhart, and R. W. Dobbins. Detection of epileptiform spikes in the EEG using a patient-independent neural net-work. In: Computer-Based Medical Systems. Proceedings of the 4th IEEE Symposium, edited by I. N. Bankmann and J. E. Tsitlik. Washington, DC: IEEE Computer Society, 1991, pp. 264-271.Google Scholar
  34. 34.
    Wilson, S. B., R. N. Harner, B. R. Duffy, B. R. Tharp, M. R. Nuwer, and M. R. Sperling. Spike detection. I. Correlation and reliability of human experts. Electroencephalogr. Clin. Neurophysiol. 98:186-198, 1996.Google Scholar

Copyright information

© Biomedical Engineering Society 2000

Authors and Affiliations

  • C. Kurth
    • 1
  • F. Gilliam
    • 2
  • B. J. Steinhoff
    • 1
  1. 1.Department of Clinical NeurophysiologyUniversity of GöttingenGermany
  2. 2.Department of NeurologyUniversity of WashingtonSt. Louis

Personalised recommendations