Journal of Biomedical Engineering Research (대한의용생체공학회:의공학회지)

The Korean Society of Medical and Biological Engineering (대한의용생체공학회)
권호 표지

Power Spectrum Estimation of EEG Signal Using Robust Filter

로버스트 필터를 이용한 EEG 신호의 스펙트럼 추정

  • 김택수 (연세대학교 공과대학 전기공학과) ;
  • 허재만 (연세대학교 공과대학 전기공학과, 순천향대학교 공과대학 전기공학과)
  • Published : 1992.06.01
Download PDF
⟨ Previous Next ⟩

Abstract

Background EEG signals can be represented as the sum of a conventional AR process and an innovation process. It Is know that conventional estimation techniques, such as least square estimates (LSE) or Gaussian maximum likelihood estimates (MLE-G ) are optimal when the innovation process satisfies the Gaussian or presumed distribution. When the data are contaminated by outliers, however, these assumptions are not met and the power spectrum estimated by conventional estimation techniques may be fatally biased. EEG signal may be affected by artifacts, which are outliers in the statistical term. So the robust filtering estimation technique is used against those artifacts and it performs well for the contaminated EEG signal.

Keywords

References

  1. Annals. Math. Statist. v.35 A General Qualitative Definition of Robustness F.R.Hampel
  2. Annals. Math. Statist. Robust Estimation of a Location Parmeter P.J.Huber
  3. Robust Statistics P.J.Huber
  4. Robust Statistics F.R.Hampel;E.M.Ronchetti;P.J.Rousseuw;W.A.Stahel
  5. J. R. Statist. Soc. Ser. B. v.34 Outliers in Time Series Fox,A.J.
  6. J. R. Statist. Soc. B. v.41 no.3 Robust Estimation of Power Spectra B.Kleiner;R.D.Martin;D.J.Thomson
  7. EEG Technology, Butterworths R.Coopee;J.W.Osslton;J.C.Shaw
  8. EEG Primer R.Spehlmann
  9. Journal of the American Statistical Association v.69 no.346 The Influnce Curve and Its Role in Robust Estimation F.R.Hampel
  10. Medical and Biomedical Engineering Analogue simulator of e.e.g. signals based on spectral components L.H.Zetterberg;Kjeli Ablin