An Automatic Subject Specific Channel Selection Method for Enhancing Motor Imagery Classification in EEG-BCI using Correlation

Pramod Gaur, Karl McCreadie, Ram Bilas Pachori, Hui Wang, Girijesh Prasad

Research output: Contribution to journalArticlepeer-review

Abstract

A motor imagery (MI) based brain-computer interface (BCI) decodes the motor intention from the electroencephalogram (EEG) of a subject and translates this into a control signal. These intentions are hence classified as different cognitive tasks, e.g. left and right hand movements. A challenge indeveloping a BCI is handling the high dimensionality of the data recordedfrom multichannel EEG signals which are highly subject-specific. Designing aportable BCI whilst minimizing EEG channel number is a challenge. To thisend, this paper presents a method to reduce the channel count with the goalof reducing computational complexity whilst maintaining a sufficient level ofaccuracy, by utilising an automatic subject-specific channel selection methodcreated using the Pearson correlation coefficient. This method computes thecorrelation between EEG signals and helps to select highly correlated EEGchannels for a particular subject without compromising classification accuracy (CA). Common spatial patterns (CSP) are used to analyse imaginedleft and right hand movements and the method is evaluated on both BCICompetition III Dataset IIIa and right hand and foot imagined tasks on BCICompetition III Dataset IVa. For both datasets, a minimum number of EEGchannels are identified with an average channel reduction of 65.45% whilstdemonstrating an increase of >5% in CA using channel Cz as a reference.
Original languageEnglish
Number of pages19
JournalBiomedical Signal Processing and Control
Publication statusAccepted/In press - 10 Mar 2021

Keywords

  • Brain-computer interface
  • motor-imagery
  • common spatial patterns,
  • linear discriminant analysis,
  • channel selection.

Fingerprint Dive into the research topics of 'An Automatic Subject Specific Channel Selection Method for Enhancing Motor Imagery Classification in EEG-BCI using Correlation'. Together they form a unique fingerprint.

Cite this