Subject-specific time-frequency selection for multi-class motor imagery-based BCIs using few Laplacian EEG channels

Yuan Yang, Sylvain Chevallier, Joe Wiart, Isabelle Bloch

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

The essential task of a motor imagery brain–computer interface (BCI) is to extract the motor imagery-related features from electroencephalogram (EEG) signals for classifying motor intentions. However, the optimal frequency band and time segment for extracting such features differ from subject to subject. In this work, we aim to improve the multi-class classification and to reduce the required EEG channel in motor imagery-based BCI by subject-specific time-frequency selection. Our method is based on a criterion namely Fisher discriminant analysis-type F-score to simultaneously select the optimal frequency band and time segment for multi-class classification. The proposed method uses only few Laplacian EEG channels (C3, Cz and C4) located around the sensorimotor area for classification. Applied to a standard multi-class BCI dataset (BCI competition III dataset IIIa), our method leads to better classification performance and smaller standard deviation across subjects compared to the state-of-art methods. Moreover, adding artifacts contaminated trials to the training dataset does not necessarily deteriorate our classification results, indicating that our method is tolerant to artifacts.

Original languageEnglish (US)
Pages (from-to)302-311
Number of pages10
JournalBiomedical Signal Processing and Control
Volume38
DOIs
StatePublished - Sep 1 2017

Fingerprint

Imagery (Psychotherapy)
Electroencephalography
Artifacts
Frequency bands
Discriminant Analysis
Discriminant analysis
Datasets

Keywords

  • Brain–computer interfaces
  • FDA-type F-score
  • Motor imagery
  • Multi-class classification
  • Time-frequency selection

ASJC Scopus subject areas

  • Signal Processing
  • Health Informatics

Cite this

@article{1a8de9b54555414caf3e107b1168ba08,
title = "Subject-specific time-frequency selection for multi-class motor imagery-based BCIs using few Laplacian EEG channels",
abstract = "The essential task of a motor imagery brain–computer interface (BCI) is to extract the motor imagery-related features from electroencephalogram (EEG) signals for classifying motor intentions. However, the optimal frequency band and time segment for extracting such features differ from subject to subject. In this work, we aim to improve the multi-class classification and to reduce the required EEG channel in motor imagery-based BCI by subject-specific time-frequency selection. Our method is based on a criterion namely Fisher discriminant analysis-type F-score to simultaneously select the optimal frequency band and time segment for multi-class classification. The proposed method uses only few Laplacian EEG channels (C3, Cz and C4) located around the sensorimotor area for classification. Applied to a standard multi-class BCI dataset (BCI competition III dataset IIIa), our method leads to better classification performance and smaller standard deviation across subjects compared to the state-of-art methods. Moreover, adding artifacts contaminated trials to the training dataset does not necessarily deteriorate our classification results, indicating that our method is tolerant to artifacts.",
keywords = "Brain–computer interfaces, FDA-type F-score, Motor imagery, Multi-class classification, Time-frequency selection",
author = "Yuan Yang and Sylvain Chevallier and Joe Wiart and Isabelle Bloch",
year = "2017",
month = "9",
day = "1",
doi = "10.1016/j.bspc.2017.06.016",
language = "English (US)",
volume = "38",
pages = "302--311",
journal = "Biomedical Signal Processing and Control",
issn = "1746-8094",
publisher = "Elsevier BV",

}

Subject-specific time-frequency selection for multi-class motor imagery-based BCIs using few Laplacian EEG channels. / Yang, Yuan; Chevallier, Sylvain; Wiart, Joe; Bloch, Isabelle.

In: Biomedical Signal Processing and Control, Vol. 38, 01.09.2017, p. 302-311.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Subject-specific time-frequency selection for multi-class motor imagery-based BCIs using few Laplacian EEG channels

AU - Yang, Yuan

AU - Chevallier, Sylvain

AU - Wiart, Joe

AU - Bloch, Isabelle

PY - 2017/9/1

Y1 - 2017/9/1

N2 - The essential task of a motor imagery brain–computer interface (BCI) is to extract the motor imagery-related features from electroencephalogram (EEG) signals for classifying motor intentions. However, the optimal frequency band and time segment for extracting such features differ from subject to subject. In this work, we aim to improve the multi-class classification and to reduce the required EEG channel in motor imagery-based BCI by subject-specific time-frequency selection. Our method is based on a criterion namely Fisher discriminant analysis-type F-score to simultaneously select the optimal frequency band and time segment for multi-class classification. The proposed method uses only few Laplacian EEG channels (C3, Cz and C4) located around the sensorimotor area for classification. Applied to a standard multi-class BCI dataset (BCI competition III dataset IIIa), our method leads to better classification performance and smaller standard deviation across subjects compared to the state-of-art methods. Moreover, adding artifacts contaminated trials to the training dataset does not necessarily deteriorate our classification results, indicating that our method is tolerant to artifacts.

AB - The essential task of a motor imagery brain–computer interface (BCI) is to extract the motor imagery-related features from electroencephalogram (EEG) signals for classifying motor intentions. However, the optimal frequency band and time segment for extracting such features differ from subject to subject. In this work, we aim to improve the multi-class classification and to reduce the required EEG channel in motor imagery-based BCI by subject-specific time-frequency selection. Our method is based on a criterion namely Fisher discriminant analysis-type F-score to simultaneously select the optimal frequency band and time segment for multi-class classification. The proposed method uses only few Laplacian EEG channels (C3, Cz and C4) located around the sensorimotor area for classification. Applied to a standard multi-class BCI dataset (BCI competition III dataset IIIa), our method leads to better classification performance and smaller standard deviation across subjects compared to the state-of-art methods. Moreover, adding artifacts contaminated trials to the training dataset does not necessarily deteriorate our classification results, indicating that our method is tolerant to artifacts.

KW - Brain–computer interfaces

KW - FDA-type F-score

KW - Motor imagery

KW - Multi-class classification

KW - Time-frequency selection

UR - http://www.scopus.com/inward/record.url?scp=85025147770&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85025147770&partnerID=8YFLogxK

U2 - 10.1016/j.bspc.2017.06.016

DO - 10.1016/j.bspc.2017.06.016

M3 - Article

VL - 38

SP - 302

EP - 311

JO - Biomedical Signal Processing and Control

JF - Biomedical Signal Processing and Control

SN - 1746-8094

ER -