Current Source Density Estimation Enhances the Performance of Motor-Imagery related Brain-Computer Interface

Dheeraj Rathee, Haider Raza, Girijesh Prasad, Hubert Cecotti

    Research output: Contribution to journalArticle

    9 Citations (Scopus)

    Abstract

    The objective is to evaluate the impact of EEG referencing schemes and spherical surface Laplacian (SSL) methods on the classification performance of motor-imagery (MI)-related brain-computer interface systems. Two EEG referencing schemes: common referencing and common average referencing and three surface Laplacian methods: current source density (CSD), finite difference method, and SSL using realistic head model were implemented separately for pre-processing of the EEG signals recorded at the scalp. A combination of filter bank common spatial filter for features extraction and support vector machine for classification was used for both pairwise binary classifications and four-class classification of MI tasks. The study provides three major outcomes: 1) the CSD method performs better than CR, providing a significant improvement of 3.02% and 5.59% across six binary classification tasks and four-class classification task, respectively; 2) the combination of a greater number of channels at the pre-processing stage as compared with the feature extraction stage yields better classification accuracies for all the Laplacian methods; and 3) the efficiency of all the surface Laplacian methods reduced significantly in the case of a fewer number of channels considered during the pre-processing.
    LanguageEnglish
    Pages2461-2471
    Number of pages10
    JournalIEEE Transactions on Neural Systems and Rehabilitation Engineering
    Volume25
    Issue number12
    Early online date13 Jul 2017
    DOIs
    Publication statusE-pub ahead of print - 13 Jul 2017

    Fingerprint

    Brain-Computer Interfaces
    Brain computer interface
    Imagery (Psychotherapy)
    Electroencephalography
    Feature extraction
    Processing
    Filter banks
    Computer Systems
    Scalp
    Finite difference method
    Support vector machines
    Head

    Keywords

    • Motor imagery
    • Brain-computer interface
    • Preprocessing
    • Spatial filtering
    • Spherical surface Laplacian.

    Cite this

    @article{5fc51eb1240647ff9b35545170312ecb,
    title = "Current Source Density Estimation Enhances the Performance of Motor-Imagery related Brain-Computer Interface",
    abstract = "The objective is to evaluate the impact of EEG referencing schemes and spherical surface Laplacian (SSL) methods on the classification performance of motor-imagery (MI)-related brain-computer interface systems. Two EEG referencing schemes: common referencing and common average referencing and three surface Laplacian methods: current source density (CSD), finite difference method, and SSL using realistic head model were implemented separately for pre-processing of the EEG signals recorded at the scalp. A combination of filter bank common spatial filter for features extraction and support vector machine for classification was used for both pairwise binary classifications and four-class classification of MI tasks. The study provides three major outcomes: 1) the CSD method performs better than CR, providing a significant improvement of 3.02{\%} and 5.59{\%} across six binary classification tasks and four-class classification task, respectively; 2) the combination of a greater number of channels at the pre-processing stage as compared with the feature extraction stage yields better classification accuracies for all the Laplacian methods; and 3) the efficiency of all the surface Laplacian methods reduced significantly in the case of a fewer number of channels considered during the pre-processing.",
    keywords = "Motor imagery, Brain-computer interface, Preprocessing, Spatial filtering, Spherical surface Laplacian.",
    author = "Dheeraj Rathee and Haider Raza and Girijesh Prasad and Hubert Cecotti",
    year = "2017",
    month = "7",
    day = "13",
    doi = "10.1109/TNSRE.2017.2726779",
    language = "English",
    volume = "25",
    pages = "2461--2471",
    journal = "IEEE Transactions on Neural Systems and Rehabilitation Engineering",
    issn = "1534-4320",
    number = "12",

    }

    Current Source Density Estimation Enhances the Performance of Motor-Imagery related Brain-Computer Interface. / Rathee, Dheeraj; Raza, Haider; Prasad, Girijesh; Cecotti, Hubert.

    In: IEEE Transactions on Neural Systems and Rehabilitation Engineering, Vol. 25, No. 12, 13.07.2017, p. 2461-2471.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - Current Source Density Estimation Enhances the Performance of Motor-Imagery related Brain-Computer Interface

    AU - Rathee, Dheeraj

    AU - Raza, Haider

    AU - Prasad, Girijesh

    AU - Cecotti, Hubert

    PY - 2017/7/13

    Y1 - 2017/7/13

    N2 - The objective is to evaluate the impact of EEG referencing schemes and spherical surface Laplacian (SSL) methods on the classification performance of motor-imagery (MI)-related brain-computer interface systems. Two EEG referencing schemes: common referencing and common average referencing and three surface Laplacian methods: current source density (CSD), finite difference method, and SSL using realistic head model were implemented separately for pre-processing of the EEG signals recorded at the scalp. A combination of filter bank common spatial filter for features extraction and support vector machine for classification was used for both pairwise binary classifications and four-class classification of MI tasks. The study provides three major outcomes: 1) the CSD method performs better than CR, providing a significant improvement of 3.02% and 5.59% across six binary classification tasks and four-class classification task, respectively; 2) the combination of a greater number of channels at the pre-processing stage as compared with the feature extraction stage yields better classification accuracies for all the Laplacian methods; and 3) the efficiency of all the surface Laplacian methods reduced significantly in the case of a fewer number of channels considered during the pre-processing.

    AB - The objective is to evaluate the impact of EEG referencing schemes and spherical surface Laplacian (SSL) methods on the classification performance of motor-imagery (MI)-related brain-computer interface systems. Two EEG referencing schemes: common referencing and common average referencing and three surface Laplacian methods: current source density (CSD), finite difference method, and SSL using realistic head model were implemented separately for pre-processing of the EEG signals recorded at the scalp. A combination of filter bank common spatial filter for features extraction and support vector machine for classification was used for both pairwise binary classifications and four-class classification of MI tasks. The study provides three major outcomes: 1) the CSD method performs better than CR, providing a significant improvement of 3.02% and 5.59% across six binary classification tasks and four-class classification task, respectively; 2) the combination of a greater number of channels at the pre-processing stage as compared with the feature extraction stage yields better classification accuracies for all the Laplacian methods; and 3) the efficiency of all the surface Laplacian methods reduced significantly in the case of a fewer number of channels considered during the pre-processing.

    KW - Motor imagery

    KW - Brain-computer interface

    KW - Preprocessing

    KW - Spatial filtering

    KW - Spherical surface Laplacian.

    U2 - 10.1109/TNSRE.2017.2726779

    DO - 10.1109/TNSRE.2017.2726779

    M3 - Article

    VL - 25

    SP - 2461

    EP - 2471

    JO - IEEE Transactions on Neural Systems and Rehabilitation Engineering

    T2 - IEEE Transactions on Neural Systems and Rehabilitation Engineering

    JF - IEEE Transactions on Neural Systems and Rehabilitation Engineering

    SN - 1534-4320

    IS - 12

    ER -