A Simplified Hybrid EEG/fNIRS based Brain Computer Interface Approach for Motor Task Classification

Guangming Zhu1, Rihui Li2, Tingting Zhang1, Dandan Lou1, Ruirong Wang3, Yingchun Zhang

  • 1Guangdong Provincial Work-Injury Rehabilitation Hospital
  • 2University of Houston
  • 3Hangzhou Dianzi University

Details

11:30 - 13:30 | Fri 26 May | Emerald III, Rose, Narcissus & Jasmine | FrPS1T1.36

Session: Poster I

Abstract

Hybrid Brain Computer Interfaces (BCI) has shown great promise for neuro-prosthetics and assistive devices in rehabilitation. However, the complexity of the BCI system and time cost for classification of motor tasks limit its applications. To overcome these changelings, concurrent Electroencephalography (EEG) and functional Near-Infrared Spectroscopy (fNIRS) recording approach was proposed in this study and tested on three healthy volunteers with a left-right hand grasping paradigm. A wavelet-based method was employed to extract the wavelet approximation coefficients from EEG signals and the slope information was employed to discriminate the concentration change of Oxygenated hemoglobin (HbO) during left-right hand grasping tasks. To maximize the valuable information carried in the two modalities, we proposed an approach based on principle component analysis (PCA) to integrate the features of fNIRS and EEG signals. Two classifiers, including support vector machine (SVM) and linear discriminant analysis (LDA) were applied to identify and estimate the control signals associated with left-right hand grasping tasks. The present experimental result demonstrates that the complement of EEG and fNIRS can significantly improve the classification accuracy by 3~9% on average. The reduction of dimensionality by PCA could achieve a reduction of time complexity and computational complexity with little loss in accuracy.