Motor imagery task classification using a signal-dependent orthogonal transform based feature extraction

Mostefa Mesbah, A. Khorshidtalab, H. Baali, A. Al-Ani

    Research output: Chapter in Book/Conference paperConference paper

    3 Citations (Scopus)
    10 Downloads (Pure)

    Abstract

    © Springer International Publishing Switzerland 2015. In this paper, we present the results of classifying electroencephalographic (EEG) signals into four motor imagery tasks using a new method for feature extraction. This method is based on a signal-dependent orthogonal transform, referred to as LP-SVD, defined as the left singular vectors of the LPC filter impulse response matrix. Using a logistic tree based model classifier, the extracted features are mapped into one of four motor imagery movements, namely left hand, right hand, foot, and tongue. The proposed technique-based classification performance was benchmarked against those based on two widely used linear transform for feature extraction methods, namely discrete cosine transform (DCT) and adaptive autoregressive (AAR). By achieving an accuracy of 67.35 %, the LP-SVD based method outperformed the other two by large margins (+25 % compared to DCT and +6 % compared to AAR-based methods).
    Original languageEnglish
    Title of host publicationNeural Information Processing
    Place of PublicationUSA
    PublisherSpringer
    Pages1-9
    Volume9490
    ISBN (Print)9783319265346
    DOIs
    Publication statusPublished - 2015
    EventNeural Information Processing: 22nd International Conference, ICONIP 2015 - Istanbul, Turkey
    Duration: 9 Nov 201512 Nov 2015

    Conference

    ConferenceNeural Information Processing: 22nd International Conference, ICONIP 2015
    CountryTurkey
    CityIstanbul
    Period9/11/1512/11/15

    Fingerprint

    Discrete cosine transforms
    Singular value decomposition
    Feature extraction
    Impulse response
    Logistics
    Classifiers

    Cite this

    Mesbah, M., Khorshidtalab, A., Baali, H., & Al-Ani, A. (2015). Motor imagery task classification using a signal-dependent orthogonal transform based feature extraction. In Neural Information Processing (Vol. 9490, pp. 1-9). USA: Springer. https://doi.org/10.1007/978-3-319-26535-3_1
    Mesbah, Mostefa ; Khorshidtalab, A. ; Baali, H. ; Al-Ani, A. / Motor imagery task classification using a signal-dependent orthogonal transform based feature extraction. Neural Information Processing. Vol. 9490 USA : Springer, 2015. pp. 1-9
    @inproceedings{be9db620a25241f4971fa022e89215dc,
    title = "Motor imagery task classification using a signal-dependent orthogonal transform based feature extraction",
    abstract = "{\circledC} Springer International Publishing Switzerland 2015. In this paper, we present the results of classifying electroencephalographic (EEG) signals into four motor imagery tasks using a new method for feature extraction. This method is based on a signal-dependent orthogonal transform, referred to as LP-SVD, defined as the left singular vectors of the LPC filter impulse response matrix. Using a logistic tree based model classifier, the extracted features are mapped into one of four motor imagery movements, namely left hand, right hand, foot, and tongue. The proposed technique-based classification performance was benchmarked against those based on two widely used linear transform for feature extraction methods, namely discrete cosine transform (DCT) and adaptive autoregressive (AAR). By achieving an accuracy of 67.35 {\%}, the LP-SVD based method outperformed the other two by large margins (+25 {\%} compared to DCT and +6 {\%} compared to AAR-based methods).",
    author = "Mostefa Mesbah and A. Khorshidtalab and H. Baali and A. Al-Ani",
    year = "2015",
    doi = "10.1007/978-3-319-26535-3_1",
    language = "English",
    isbn = "9783319265346",
    volume = "9490",
    pages = "1--9",
    booktitle = "Neural Information Processing",
    publisher = "Springer",
    address = "Netherlands",

    }

    Mesbah, M, Khorshidtalab, A, Baali, H & Al-Ani, A 2015, Motor imagery task classification using a signal-dependent orthogonal transform based feature extraction. in Neural Information Processing. vol. 9490, Springer, USA, pp. 1-9, Neural Information Processing: 22nd International Conference, ICONIP 2015, Istanbul, Turkey, 9/11/15. https://doi.org/10.1007/978-3-319-26535-3_1

    Motor imagery task classification using a signal-dependent orthogonal transform based feature extraction. / Mesbah, Mostefa; Khorshidtalab, A.; Baali, H.; Al-Ani, A.

    Neural Information Processing. Vol. 9490 USA : Springer, 2015. p. 1-9.

    Research output: Chapter in Book/Conference paperConference paper

    TY - GEN

    T1 - Motor imagery task classification using a signal-dependent orthogonal transform based feature extraction

    AU - Mesbah, Mostefa

    AU - Khorshidtalab, A.

    AU - Baali, H.

    AU - Al-Ani, A.

    PY - 2015

    Y1 - 2015

    N2 - © Springer International Publishing Switzerland 2015. In this paper, we present the results of classifying electroencephalographic (EEG) signals into four motor imagery tasks using a new method for feature extraction. This method is based on a signal-dependent orthogonal transform, referred to as LP-SVD, defined as the left singular vectors of the LPC filter impulse response matrix. Using a logistic tree based model classifier, the extracted features are mapped into one of four motor imagery movements, namely left hand, right hand, foot, and tongue. The proposed technique-based classification performance was benchmarked against those based on two widely used linear transform for feature extraction methods, namely discrete cosine transform (DCT) and adaptive autoregressive (AAR). By achieving an accuracy of 67.35 %, the LP-SVD based method outperformed the other two by large margins (+25 % compared to DCT and +6 % compared to AAR-based methods).

    AB - © Springer International Publishing Switzerland 2015. In this paper, we present the results of classifying electroencephalographic (EEG) signals into four motor imagery tasks using a new method for feature extraction. This method is based on a signal-dependent orthogonal transform, referred to as LP-SVD, defined as the left singular vectors of the LPC filter impulse response matrix. Using a logistic tree based model classifier, the extracted features are mapped into one of four motor imagery movements, namely left hand, right hand, foot, and tongue. The proposed technique-based classification performance was benchmarked against those based on two widely used linear transform for feature extraction methods, namely discrete cosine transform (DCT) and adaptive autoregressive (AAR). By achieving an accuracy of 67.35 %, the LP-SVD based method outperformed the other two by large margins (+25 % compared to DCT and +6 % compared to AAR-based methods).

    U2 - 10.1007/978-3-319-26535-3_1

    DO - 10.1007/978-3-319-26535-3_1

    M3 - Conference paper

    SN - 9783319265346

    VL - 9490

    SP - 1

    EP - 9

    BT - Neural Information Processing

    PB - Springer

    CY - USA

    ER -

    Mesbah M, Khorshidtalab A, Baali H, Al-Ani A. Motor imagery task classification using a signal-dependent orthogonal transform based feature extraction. In Neural Information Processing. Vol. 9490. USA: Springer. 2015. p. 1-9 https://doi.org/10.1007/978-3-319-26535-3_1