EEG Sleep Staging Using Vectorial Autoregressive Models

  • Arnon Cohen
  • Felix Flomen
  • Nir Drori


Sleep studies require the use of several channels of EEG. The analysis of vector EEG, exhibits significant advantages over scalar analysis. Novel algorithms for segmentation, classification and compression of vector EEG are described. The statistics of the suggested measures for segmentation and classification are discussed. The algorithms were evaluated on four patients, yielding mean correct sleep staging of about 85%.


Sleep Stage Distortion Measure Inverse Filter Neighbor Rule Codebook Size 
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. Bodenstein, G., and Praetorius, H. M., 1977, Feature extraction from the electroencephalogram by adaptive segmentation, Proc. IEEE, 35: 642–652.CrossRefGoogle Scholar
  2. Franaszczuk, P. J., Blinowska, K. J. and Kowalczyk, M., 1985, The application of parametric multichannel spectral estimates in the study of electrical brain activity, Biol. Cybern., 51: 239–247.zbMATHCrossRefGoogle Scholar
  3. Gersch. W. and Yonemoto. J., 1977a, Parametric time series models for multivariate EEG analysis, Computers in Biomed. Res., 10: 113–125.CrossRefGoogle Scholar
  4. Gersch, W. Yonemoto, J., and Naitoh, P., 1977b, Automatic classification of multivariate EEGs using an amount of information and the eigenvalues of parametric time series model features, Computers in Biomed. Res., 17: 352–361.Google Scholar
  5. Gersch, W.. Martinelli, F., Yonemoto, J., Low, M.D. and McEwan, J. A.. 1979, Automatic classification of EEGs: Kullback-Leibler nearest neighbor rules, Science, 205: 193–195.CrossRefGoogle Scholar
  6. Gray, R. M., Buzo, A., and Gray. A., H., 1980, Distortion measures for speech processing, IEEE Trans. Acoust. Speech & Sig. Proc. ASSP-28: 367–376.MathSciNetzbMATHCrossRefGoogle Scholar
  7. Hannan. E. J., 1970, Multiple Time Series. John Wiley & Sons, N.Y.zbMATHCrossRefGoogle Scholar
  8. Haykin, S. and Kesler. S., 1979. Prediction error filtering and maximum entropy spectral estimation, in: Haykin. S., (ed.), Topics in Applied Physics. 34: 9–70, Springer Verlag, Berlin.Google Scholar
  9. Isaksson, A., Wennberg, A., and Zetterberg, L. H., 1981, Computer analysis of EEG signals with parametric models. Proc. IEEE. 69: 451–461.CrossRefGoogle Scholar
  10. Jenkins. G. M. and Watts, D. G.. 1986, Spectral Analysis and its Applications, Holden Day, San Francisco, Ca.Google Scholar
  11. Kay. S. M.. 1988, Modem Spectral Estimation: Theory and Applications, Prentice Hall. Engelwood Cliffs. N.J.Google Scholar
  12. Linde,Y.. Buzo. A., and Gray. R. M.. 1980. An algorithm for vector quantizer design. IEEE Trans. Comm., COM-28: 84–95.Google Scholar
  13. Marple, S. L., 1987, Digital Spectral Analysis with Applications. Prentice Hall, Engelwood Cliffs, N.J.Google Scholar
  14. Nhiro. I.. Hideyuki, S.. Akira, I., and Nobus, S., 1980. Computer classification of the EEG time series by Kulback information measure, Int. J. Syst. Sci., 11: 677–687.CrossRefGoogle Scholar
  15. Priestly, M. B., Spectral Analysis and Time Series, 1981, Academic Press. N.Y.Google Scholar
  16. Sanderson, A. C. Segen, J., and Richey, E., 1980, Hierarchical modeling of EEG signals, IEEE Trans. Patt. Anal. Mach. Intell., PAMI-2: 405–414.Google Scholar
  17. Whitle, P., 1963, On the fitting of multivariate autoregression, and the approximate canonical factorization of spectral density matrix, Biometrika, 50: 129–134.MathSciNetGoogle Scholar

Copyright information

© Springer Science+Business Media New York 1996

Authors and Affiliations

  • Arnon Cohen
    • 1
  • Felix Flomen
    • 1
  • Nir Drori
    • 1
  1. 1.Electrical and Computer Engineering Department Biomedical Engineering ProgramBen-Gurion UniversityBeer-ShevaIsrael

Personalised recommendations