Skip to main content
SpringerLink
Log in

Automatic recognition of vigilance state by using a wavelet-based artificial neural network

  • Original Article
  • Published:
Neural Computing & Applications Aims and scope Submit manuscript

Abstract

In this study, 5-s long sequences of full-spectrum electroencephalogram (EEG) recordings were used for classifying alert versus drowsy states in an arbitrary subject. EEG signals were obtained from 30 healthy subjects and the results were classified using a wavelet-based neural network. The wavelet-based neural network model, employing the multilayer perceptron (MLP), was used for the classification of EEG signals. A multilayer perceptron neural network (MLPNN) trained with the Levenberg–Marquardt algorithm was used to discriminate the alertness level of the subject. In order to determine the MLPNN inputs, spectral analysis of EEG signals was performed using the discrete wavelet transform (DWT) technique. The MLPNN was trained, cross-validated, and tested with training, cross-validation, and testing sets, respectively. The correct classification rate was 93.3% alert, 96.6% drowsy, and 90% sleep. The classification results showed that the MLPNN trained with the Levenberg–Marquardt algorithm was effective for discriminating the vigilance state of the subject.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2a–c
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7

Similar content being viewed by others

References

  1. Vuckovic A, Radivojevic V, Chen ACN, Popovic D (2002) Automatic recognition of alertness and drowsiness from EEG by an artificial neural network. Med Eng Phys 24:349–360

    Article  Google Scholar 

  2. Güler İ, Kiymik MK, Akin M, Alkan A (2001) AR spectral analysis of EEG signals by using maximum likelihood estimation. Comput Biol Med 31:441–450

    Article  Google Scholar 

  3. Herrmann CS, Arnold T, Visbeck A, Hundemer HP, Hopf HC (2001) Adaptive frequency decomposition of EEG with subsequent expert system analysis. Comput Biol Med 31:407–427

    Article  Google Scholar 

  4. Peters BO, Pfurtscheller G, Flyvbjerg H (1998) Mining multi-channel EEG for its information content: an ANN-based method for a brain–computer interface. Neural Netw 11:1429–1433

    Article  Google Scholar 

  5. Khahill M, Duchene J (1999) Detection and classification of multiple events in piecewise stationary signals. J Signal Processing S0165-1684(98):236–239

    Google Scholar 

  6. Jung TP, Makeig S, Stensmo M, Sejnowski TJ (1997) Estimating alertness from the EEG power spectrum. IEEE Trans Biomed Eng 44:60–69

    Article  Google Scholar 

  7. Conradt R, Brandenburg U, Penzel T, Hasan J, Varri A, Peter JH (1999) Vigilance transitions in reaction time test: a method of describing the state of alertness more objectively. Clin Neurophysiol 110:1499–1509

    Article  Google Scholar 

  8. Robert C, Gaudy JF, Limoge A (2002) Electroencephalogram processing using neural networks. Clin Neurophysiol 113:694–701

    Article  Google Scholar 

  9. Principe JC, Gala SK, Chang TG (1989) Sleep staging automation based on the theory of evidence. IEEE Trans Biomed Eng 36:503–509

    Article  Google Scholar 

  10. De Gennaro L, Ferrara M, Bertini M (2001) The boundary between wakefulness and sleep: quantitative electroencephalographic changes during the sleep onset period. Neuroscience 107(1):1–11

    Article  Google Scholar 

  11. Wilson BJ, Bracewell TD (2000) Alertness monitor using neural networks for EEG analysis. Proc Neur Net Signal Process X (ISPS) 2:814–820

    Google Scholar 

  12. Tsoi AC, So DSC, Sergejew A (1994) Classification of electroencephalogram using artificial neural networks. In: Cowan JD, Tesauro G, Alspector J (eds) Advances in neural information processing systems 6. Morgan Kaufmann, San Mateo, California, pp 1151–1158

  13. Peters BO, Pfurtscheller G, Flyvbjerg H (2001) Automatic differentiation of multichannel EEG signals. IEEE Trans Biomed Eng 48:111–116

    Article  Google Scholar 

  14. McKeown MJ, Humphries C, Achermann P, Borbely AA, Sejnowski TJ (1997) A new method for detecting state changes in EEG: exploratory application to sleep data. J Sleep Res 7:48–56

    Article  Google Scholar 

  15. Basheer IA, Hajmeer M (2000) Artificial neural networks: fundamentals, computing, design, and application. J Microbiological Methods 43:3–31

    Article  CAS  PubMed  Google Scholar 

  16. Haykin S (1994) Neural networks: a comprehensive foundation. Macmillan, New York

    Google Scholar 

  17. Guler NF, Ubeyli ED (2004) Wavelet-based neural network analysis of ophthalmic artery Doppler signals. Comput Biol Med 34(7):601–613

    Article  Google Scholar 

  18. Adeli H, Zhou Z, Dadmehr N (2003) Analysis of EEG records in an epileptic patient using wavelet transform. J Neurosci Methods 123:69–87

    Article  Google Scholar 

  19. Shimada T, Shiina T, Saito Y (2000) Detection of characteristic waves of sleep EEG by neural network analysis. IEEE Trans Biomed Eng 47:369–379

    Article  Google Scholar 

  20. Akay M, Marsic I, Medl A, Bu GA (1998) A system for medical consultation and education using multimodal human/machine communication. IEEE Trans Inf Tech Biomed 2:282–291

    Article  Google Scholar 

  21. Daubechies I (1996) Where do wavelets come from? A personal point of view. Proc IEEE 84:510–513

    Article  Google Scholar 

  22. Cohen A, Kovacevic J (1996) Wavelets: the mathematical background. Proc IEEE 84:514–522

    Article  Google Scholar 

  23. Rioul O, Vetterli M (1991) Wavelet and signal processing. IEEE Signal Proc Mag 84(4):14–38

    Article  Google Scholar 

  24. Chaudhuri BB, Bhattacharya U (2000) Efficient training and improved performance of multilayer perceptron in pattern classification. Neurocomputing 34:11–27

    Article  Google Scholar 

  25. Hagan MT, Menhaj MB (1994) Training feedforward networks with the Marquardt algorithm. IEEE Trans Neural Netw 5(6): 989–993

    Google Scholar 

  26. De Carli F, Nobili L, Gelcich P, Ferrillo F (1999) A method for the automatic detection of arousals during sleep. Sleep 22:561–572

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Electrical and Electronics Engineering, Kahramanmaraş Sütçü İmam University, Kahramanmaras, 46100, Turkey

    Abdulhamit Subasi & M. Kemal Kiymik

  2. Department of Electrical and Electronics Engineering, Dicle University, Diyarbakir, Turkey

    Mehmet Akin

  3. Gulhane Military Medicine Academy, Biomedical Engineering Center, Ankara, Turkey

    Osman Erogul

Authors
  1. Abdulhamit Subasi
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. M. Kemal Kiymik
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Mehmet Akin
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Osman Erogul
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Abdulhamit Subasi.

Appendix

Appendix

1.1 The measure of sensitivity, selectivity, and specificity

The performance of a particular run of the program or a particular reading by an expert was evaluated in terms of sensitivity, selectivity, and specificity:

$$\begin{aligned} {\text{Sensitivity}} = \frac{{{\text{TP}}}}{{{\text{TP}} + {\text{FN}}}} \times 100\% & \\ {\text{Selectivity}} = \frac{{{\text{TP}}}}{{{\text{TP}} + {\text{FP}}}} \times 100\% & \\ {\text{Specificity}} = \frac{{{\text{TN}}}}{{{\text{TN}} + {\text{FP}}}} \times 100\% & \\ \end{aligned} $$

The specificity was computed only in the context of the discriminant analysis, in which each fixed length basic epoch was classified as true positive (TP), false positive (FP), true negative (TN), or false negative (FN). In subsequent analyses, variable length vigilance states, marked by one observer, were compared to the reference set and the individual events were considered as TP (if an overlapping occurred), FP, or FN. We believed that, in this case, TN counting, and consequently, specificity evaluation, was nonsensical [26].

Rights and permissions

Reprints and permissions

About this article

Cite this article

Subasi, A., Kiymik, M.K., Akin, M. et al. Automatic recognition of vigilance state by using a wavelet-based artificial neural network. Neural Comput & Applic 14, 45–55 (2005). https://doi.org/10.1007/s00521-004-0441-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00521-004-0441-0

Keywords

Search

Navigation