Stationary common spatial patterns for brain-computer interfacing

Wojciech Samek, Carmen Vidaurre, Klaus Muller, Motoaki Kawanabe

Research output: Contribution to journalArticle

134 Citations (Scopus)

Abstract

Classifying motion intentions in brain-computer interfacing (BCI) is a demanding task as the recorded EEG signal is not only noisy and has limited spatial resolution but it is also intrinsically non-stationary. The non-stationarities in the signal may come from many different sources, for instance, electrode artefacts, muscular activity or changes of task involvement, and often deteriorate classification performance. This is mainly because features extracted by standard methods like common spatial patterns (CSP) are not invariant to variations of the signal properties, thus should also change over time. Although many extensions of CSP were proposed to, for example, reduce the sensitivity to noise or incorporate information from other subjects, none of them tackles the non-stationarity problem directly. In this paper, we propose a method which regularizes CSP towards stationary subspaces (sCSP) and show that this increases classification accuracy, especially for subjects who are hardly able to control a BCI. We compare our method with the state-of-the-art approaches on different datasets, show competitive results and analyse the reasons for the improvement.

Original languageEnglish
Article number026013
JournalJournal of Neural Engineering
Volume9
Issue number2
DOIs
Publication statusPublished - 2012 Apr 1
Externally publishedYes

Fingerprint

Brain
Electroencephalography
Artifacts
Electrodes
Noise
Datasets

ASJC Scopus subject areas

  • Biomedical Engineering
  • Cellular and Molecular Neuroscience

Cite this

Stationary common spatial patterns for brain-computer interfacing. / Samek, Wojciech; Vidaurre, Carmen; Muller, Klaus; Kawanabe, Motoaki.

In: Journal of Neural Engineering, Vol. 9, No. 2, 026013, 01.04.2012.

Research output: Contribution to journalArticle

Samek, Wojciech ; Vidaurre, Carmen ; Muller, Klaus ; Kawanabe, Motoaki. / Stationary common spatial patterns for brain-computer interfacing. In: Journal of Neural Engineering. 2012 ; Vol. 9, No. 2.
@article{36b7f143b3454d06856941a2b264e66f,
title = "Stationary common spatial patterns for brain-computer interfacing",
abstract = "Classifying motion intentions in brain-computer interfacing (BCI) is a demanding task as the recorded EEG signal is not only noisy and has limited spatial resolution but it is also intrinsically non-stationary. The non-stationarities in the signal may come from many different sources, for instance, electrode artefacts, muscular activity or changes of task involvement, and often deteriorate classification performance. This is mainly because features extracted by standard methods like common spatial patterns (CSP) are not invariant to variations of the signal properties, thus should also change over time. Although many extensions of CSP were proposed to, for example, reduce the sensitivity to noise or incorporate information from other subjects, none of them tackles the non-stationarity problem directly. In this paper, we propose a method which regularizes CSP towards stationary subspaces (sCSP) and show that this increases classification accuracy, especially for subjects who are hardly able to control a BCI. We compare our method with the state-of-the-art approaches on different datasets, show competitive results and analyse the reasons for the improvement.",
author = "Wojciech Samek and Carmen Vidaurre and Klaus Muller and Motoaki Kawanabe",
year = "2012",
month = "4",
day = "1",
doi = "10.1088/1741-2560/9/2/026013",
language = "English",
volume = "9",
journal = "Journal of Neural Engineering",
issn = "1741-2560",
publisher = "IOP Publishing Ltd.",
number = "2",

}

TY - JOUR

T1 - Stationary common spatial patterns for brain-computer interfacing

AU - Samek, Wojciech

AU - Vidaurre, Carmen

AU - Muller, Klaus

AU - Kawanabe, Motoaki

PY - 2012/4/1

Y1 - 2012/4/1

N2 - Classifying motion intentions in brain-computer interfacing (BCI) is a demanding task as the recorded EEG signal is not only noisy and has limited spatial resolution but it is also intrinsically non-stationary. The non-stationarities in the signal may come from many different sources, for instance, electrode artefacts, muscular activity or changes of task involvement, and often deteriorate classification performance. This is mainly because features extracted by standard methods like common spatial patterns (CSP) are not invariant to variations of the signal properties, thus should also change over time. Although many extensions of CSP were proposed to, for example, reduce the sensitivity to noise or incorporate information from other subjects, none of them tackles the non-stationarity problem directly. In this paper, we propose a method which regularizes CSP towards stationary subspaces (sCSP) and show that this increases classification accuracy, especially for subjects who are hardly able to control a BCI. We compare our method with the state-of-the-art approaches on different datasets, show competitive results and analyse the reasons for the improvement.

AB - Classifying motion intentions in brain-computer interfacing (BCI) is a demanding task as the recorded EEG signal is not only noisy and has limited spatial resolution but it is also intrinsically non-stationary. The non-stationarities in the signal may come from many different sources, for instance, electrode artefacts, muscular activity or changes of task involvement, and often deteriorate classification performance. This is mainly because features extracted by standard methods like common spatial patterns (CSP) are not invariant to variations of the signal properties, thus should also change over time. Although many extensions of CSP were proposed to, for example, reduce the sensitivity to noise or incorporate information from other subjects, none of them tackles the non-stationarity problem directly. In this paper, we propose a method which regularizes CSP towards stationary subspaces (sCSP) and show that this increases classification accuracy, especially for subjects who are hardly able to control a BCI. We compare our method with the state-of-the-art approaches on different datasets, show competitive results and analyse the reasons for the improvement.

UR - http://www.scopus.com/inward/record.url?scp=84857884326&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84857884326&partnerID=8YFLogxK

U2 - 10.1088/1741-2560/9/2/026013

DO - 10.1088/1741-2560/9/2/026013

M3 - Article

C2 - 22350439

AN - SCOPUS:84857884326

VL - 9

JO - Journal of Neural Engineering

JF - Journal of Neural Engineering

SN - 1741-2560

IS - 2

M1 - 026013

ER -