Selective Feature Generation Method Based on Time Domain Parameters and Correlation Coefficients for Filter-Bank-CSP BCI Systems

Yongkoo Park, Wonzoo Chung

Research output: Contribution to journalArticle

Abstract

This paper presents a novel motor imagery (MI) classification algorithm using filter-bank common spatial pattern (FBCSP) features based on MI-relevant channel selection. In contrast to existing channel selection methods based on global CSP features, the proposed algorithm utilizes the Fisher ratio of time domain parameters (TDPs) and correlation coefficients: the channel with the highest Fisher ratio of TDPs, named principle channel, is selected and a supporting channel set for the principle channel that consists of highly correlated channels to the principle channel is generated. The proposed algorithm using the FBCSP features generated from the supporting channel set for the principle channel significantly improved the classification performance. The performance of the proposed method was evaluated using BCI Competition III Dataset IVa (18 channels) and BCI Competition IV Dataset I (59 channels).

Original languageEnglish
JournalSensors (Basel, Switzerland)
Volume19
Issue number17
DOIs
Publication statusPublished - 2019 Aug 30

Fingerprint

Filter banks
correlation coefficients
Imagery (Psychotherapy)
filters
coefficients
imagery
Datasets

Keywords

  • brain-computer interfaces (BCIs)
  • common spatial pattern (CSP)
  • correlation coefficient
  • motor-imagery (MI)
  • time domain parameters

ASJC Scopus subject areas

  • Analytical Chemistry
  • Atomic and Molecular Physics, and Optics
  • Biochemistry
  • Instrumentation
  • Electrical and Electronic Engineering

Cite this

@article{459c0a2a91c6404bafbba8a5c6c88380,
title = "Selective Feature Generation Method Based on Time Domain Parameters and Correlation Coefficients for Filter-Bank-CSP BCI Systems",
abstract = "This paper presents a novel motor imagery (MI) classification algorithm using filter-bank common spatial pattern (FBCSP) features based on MI-relevant channel selection. In contrast to existing channel selection methods based on global CSP features, the proposed algorithm utilizes the Fisher ratio of time domain parameters (TDPs) and correlation coefficients: the channel with the highest Fisher ratio of TDPs, named principle channel, is selected and a supporting channel set for the principle channel that consists of highly correlated channels to the principle channel is generated. The proposed algorithm using the FBCSP features generated from the supporting channel set for the principle channel significantly improved the classification performance. The performance of the proposed method was evaluated using BCI Competition III Dataset IVa (18 channels) and BCI Competition IV Dataset I (59 channels).",
keywords = "brain-computer interfaces (BCIs), common spatial pattern (CSP), correlation coefficient, motor-imagery (MI), time domain parameters",
author = "Yongkoo Park and Wonzoo Chung",
year = "2019",
month = "8",
day = "30",
doi = "10.3390/s19173769",
language = "English",
volume = "19",
journal = "Sensors (Switzerland)",
issn = "1424-8220",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "17",

}

TY - JOUR

T1 - Selective Feature Generation Method Based on Time Domain Parameters and Correlation Coefficients for Filter-Bank-CSP BCI Systems

AU - Park, Yongkoo

AU - Chung, Wonzoo

PY - 2019/8/30

Y1 - 2019/8/30

N2 - This paper presents a novel motor imagery (MI) classification algorithm using filter-bank common spatial pattern (FBCSP) features based on MI-relevant channel selection. In contrast to existing channel selection methods based on global CSP features, the proposed algorithm utilizes the Fisher ratio of time domain parameters (TDPs) and correlation coefficients: the channel with the highest Fisher ratio of TDPs, named principle channel, is selected and a supporting channel set for the principle channel that consists of highly correlated channels to the principle channel is generated. The proposed algorithm using the FBCSP features generated from the supporting channel set for the principle channel significantly improved the classification performance. The performance of the proposed method was evaluated using BCI Competition III Dataset IVa (18 channels) and BCI Competition IV Dataset I (59 channels).

AB - This paper presents a novel motor imagery (MI) classification algorithm using filter-bank common spatial pattern (FBCSP) features based on MI-relevant channel selection. In contrast to existing channel selection methods based on global CSP features, the proposed algorithm utilizes the Fisher ratio of time domain parameters (TDPs) and correlation coefficients: the channel with the highest Fisher ratio of TDPs, named principle channel, is selected and a supporting channel set for the principle channel that consists of highly correlated channels to the principle channel is generated. The proposed algorithm using the FBCSP features generated from the supporting channel set for the principle channel significantly improved the classification performance. The performance of the proposed method was evaluated using BCI Competition III Dataset IVa (18 channels) and BCI Competition IV Dataset I (59 channels).

KW - brain-computer interfaces (BCIs)

KW - common spatial pattern (CSP)

KW - correlation coefficient

KW - motor-imagery (MI)

KW - time domain parameters

UR - http://www.scopus.com/inward/record.url?scp=85071743768&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85071743768&partnerID=8YFLogxK

U2 - 10.3390/s19173769

DO - 10.3390/s19173769

M3 - Article

C2 - 31480390

AN - SCOPUS:85071743768

VL - 19

JO - Sensors (Switzerland)

JF - Sensors (Switzerland)

SN - 1424-8220

IS - 17

ER -