Asynchronous, adaptive BCI using movement imagination training and rest-state inference

Siamac Fazli, Márton Danóczy, Motoaki Kawanabe, Florin Popescu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

The current study introduces an adaptive Bayesian learning scheme which discriminates between left hand movement imagination, right hand movement imagination and idle (i.e. "no-command") state in an EEG Brain Computer Interface. Unlike previous BCI designs using minimal training, the user does not have to continuously imagine a movement in order to control a cursor. Rather, the cursor reacts meaningfully only when a trained movement imagination is produced. The algorithmic approach was to compute Gaussian probability distributions in log-variance of main Common Spatial Patterns for each movement class, infer from these a prior distribution of idle-class, and allow each distribution to adapt during feedback BCI performance. By producing a markedly different but complexity constrained partition of feature space than with LDA classifiers, allowing the classifier to adapt and introducing an intermediary state driven by the classifier output through a dynamic control law, 90% level classification accuracy was achieved with less than 5 seconds activation time from cued onset.

Original languageEnglish
Title of host publicationProceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008
Pages85-90
Number of pages6
Publication statusPublished - Dec 1 2008
Externally publishedYes
EventIASTED International Conference on Artificial Intelligence and Applications, AIA 2008 - Innsbruck, Austria
Duration: Feb 13 2008Feb 15 2008

Publication series

NameProceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008

Conference

ConferenceIASTED International Conference on Artificial Intelligence and Applications, AIA 2008
CountryAustria
CityInnsbruck
Period2/13/082/15/08

Fingerprint

Classifiers
Brain computer interface
Electroencephalography
Probability distributions
Chemical activation
Feedback

Keywords

  • Asynchronous design
  • Bayesian inference
  • Brain-computer interface
  • Idle state
  • Machine learning

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications

Cite this

Fazli, S., Danóczy, M., Kawanabe, M., & Popescu, F. (2008). Asynchronous, adaptive BCI using movement imagination training and rest-state inference. In Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008 (pp. 85-90). (Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008).

Asynchronous, adaptive BCI using movement imagination training and rest-state inference. / Fazli, Siamac; Danóczy, Márton; Kawanabe, Motoaki; Popescu, Florin.

Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008. 2008. p. 85-90 (Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Fazli, S, Danóczy, M, Kawanabe, M & Popescu, F 2008, Asynchronous, adaptive BCI using movement imagination training and rest-state inference. in Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008. Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008, pp. 85-90, IASTED International Conference on Artificial Intelligence and Applications, AIA 2008, Innsbruck, Austria, 2/13/08.
Fazli S, Danóczy M, Kawanabe M, Popescu F. Asynchronous, adaptive BCI using movement imagination training and rest-state inference. In Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008. 2008. p. 85-90. (Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008).
Fazli, Siamac ; Danóczy, Márton ; Kawanabe, Motoaki ; Popescu, Florin. / Asynchronous, adaptive BCI using movement imagination training and rest-state inference. Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008. 2008. pp. 85-90 (Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008).
@inproceedings{5337e94737c244389ee0951b988a5ad7,
title = "Asynchronous, adaptive BCI using movement imagination training and rest-state inference",
abstract = "The current study introduces an adaptive Bayesian learning scheme which discriminates between left hand movement imagination, right hand movement imagination and idle (i.e. {"}no-command{"}) state in an EEG Brain Computer Interface. Unlike previous BCI designs using minimal training, the user does not have to continuously imagine a movement in order to control a cursor. Rather, the cursor reacts meaningfully only when a trained movement imagination is produced. The algorithmic approach was to compute Gaussian probability distributions in log-variance of main Common Spatial Patterns for each movement class, infer from these a prior distribution of idle-class, and allow each distribution to adapt during feedback BCI performance. By producing a markedly different but complexity constrained partition of feature space than with LDA classifiers, allowing the classifier to adapt and introducing an intermediary state driven by the classifier output through a dynamic control law, 90{\%} level classification accuracy was achieved with less than 5 seconds activation time from cued onset.",
keywords = "Asynchronous design, Bayesian inference, Brain-computer interface, Idle state, Machine learning",
author = "Siamac Fazli and M{\'a}rton Dan{\'o}czy and Motoaki Kawanabe and Florin Popescu",
year = "2008",
month = "12",
day = "1",
language = "English",
isbn = "9780889867093",
series = "Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008",
pages = "85--90",
booktitle = "Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008",

}

TY - GEN

T1 - Asynchronous, adaptive BCI using movement imagination training and rest-state inference

AU - Fazli, Siamac

AU - Danóczy, Márton

AU - Kawanabe, Motoaki

AU - Popescu, Florin

PY - 2008/12/1

Y1 - 2008/12/1

N2 - The current study introduces an adaptive Bayesian learning scheme which discriminates between left hand movement imagination, right hand movement imagination and idle (i.e. "no-command") state in an EEG Brain Computer Interface. Unlike previous BCI designs using minimal training, the user does not have to continuously imagine a movement in order to control a cursor. Rather, the cursor reacts meaningfully only when a trained movement imagination is produced. The algorithmic approach was to compute Gaussian probability distributions in log-variance of main Common Spatial Patterns for each movement class, infer from these a prior distribution of idle-class, and allow each distribution to adapt during feedback BCI performance. By producing a markedly different but complexity constrained partition of feature space than with LDA classifiers, allowing the classifier to adapt and introducing an intermediary state driven by the classifier output through a dynamic control law, 90% level classification accuracy was achieved with less than 5 seconds activation time from cued onset.

AB - The current study introduces an adaptive Bayesian learning scheme which discriminates between left hand movement imagination, right hand movement imagination and idle (i.e. "no-command") state in an EEG Brain Computer Interface. Unlike previous BCI designs using minimal training, the user does not have to continuously imagine a movement in order to control a cursor. Rather, the cursor reacts meaningfully only when a trained movement imagination is produced. The algorithmic approach was to compute Gaussian probability distributions in log-variance of main Common Spatial Patterns for each movement class, infer from these a prior distribution of idle-class, and allow each distribution to adapt during feedback BCI performance. By producing a markedly different but complexity constrained partition of feature space than with LDA classifiers, allowing the classifier to adapt and introducing an intermediary state driven by the classifier output through a dynamic control law, 90% level classification accuracy was achieved with less than 5 seconds activation time from cued onset.

KW - Asynchronous design

KW - Bayesian inference

KW - Brain-computer interface

KW - Idle state

KW - Machine learning

UR - http://www.scopus.com/inward/record.url?scp=44449093082&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=44449093082&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9780889867093

T3 - Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008

SP - 85

EP - 90

BT - Proceedings of the IASTED International Conference on Artificial Intelligence and Applications, AIA 2008

ER -