Simultaneously optimizing spatial spectral features based on mutual information for EEG classification

Jianjun Meng, Lin Yao, Xinjun Sheng, Dingguo Zhang, Xiangyang Zhu

Research output: Contribution to journalArticle

25 Citations (Scopus)

Abstract

High performance of the brain-computer interface (BCI) needs efficient algorithms to extract discriminative features from raw electroencephalography (EEG) signals. In this paper, we present a novel scheme to extract spatial spectral features for the motor imagery-based BCI. The learning task is formulated by maximizing the mutual information between spatial spectral features (MMISS) and class labels, by which a unique objective function directly related to Bayes classification error is optimized. The spatial spectral features are assumed to follow a parametric Gaussian distribution, which has been validated by the normal distribution Mardia's test, and under this assumption the estimation of mutual information is derived. We propose a gradient based alternative and iterative learning algorithm to optimize the cost function and derive the spatial and spectral filters simultaneously. The experimental results on dataset IVa of BCI competition III and dataset IIa of BCI competition IV show that the proposed MMISS is able to efficiently extract discriminative features from motor imagery-based EEG signals to enhance the classification accuracy compared to other existing algorithms.

Original languageEnglish (US)
Article number6871337
Pages (from-to)227-240
Number of pages14
JournalIEEE Transactions on Biomedical Engineering
Volume62
Issue number1
DOIs
StatePublished - Jan 1 2015

Fingerprint

Brain computer interface
Electroencephalography
Gaussian distribution
Normal distribution
Cost functions
Learning algorithms
Labels

Cite this

Simultaneously optimizing spatial spectral features based on mutual information for EEG classification. / Meng, Jianjun; Yao, Lin; Sheng, Xinjun; Zhang, Dingguo; Zhu, Xiangyang.

In: IEEE Transactions on Biomedical Engineering, Vol. 62, No. 1, 6871337, 01.01.2015, p. 227-240.

Research output: Contribution to journalArticle

Meng, Jianjun ; Yao, Lin ; Sheng, Xinjun ; Zhang, Dingguo ; Zhu, Xiangyang. / Simultaneously optimizing spatial spectral features based on mutual information for EEG classification. In: IEEE Transactions on Biomedical Engineering. 2015 ; Vol. 62, No. 1. pp. 227-240.
@article{ccf6dc1b079d4b9db8493c51071f58fd,
title = "Simultaneously optimizing spatial spectral features based on mutual information for EEG classification",
abstract = "High performance of the brain-computer interface (BCI) needs efficient algorithms to extract discriminative features from raw electroencephalography (EEG) signals. In this paper, we present a novel scheme to extract spatial spectral features for the motor imagery-based BCI. The learning task is formulated by maximizing the mutual information between spatial spectral features (MMISS) and class labels, by which a unique objective function directly related to Bayes classification error is optimized. The spatial spectral features are assumed to follow a parametric Gaussian distribution, which has been validated by the normal distribution Mardia's test, and under this assumption the estimation of mutual information is derived. We propose a gradient based alternative and iterative learning algorithm to optimize the cost function and derive the spatial and spectral filters simultaneously. The experimental results on dataset IVa of BCI competition III and dataset IIa of BCI competition IV show that the proposed MMISS is able to efficiently extract discriminative features from motor imagery-based EEG signals to enhance the classification accuracy compared to other existing algorithms.",
author = "Jianjun Meng and Lin Yao and Xinjun Sheng and Dingguo Zhang and Xiangyang Zhu",
year = "2015",
month = "1",
day = "1",
doi = "10.1109/TBME.2014.2345458",
language = "English (US)",
volume = "62",
pages = "227--240",
journal = "IEEE Transactions on Biomedical Engineering",
issn = "0018-9294",
publisher = "IEEE Computer Society",
number = "1",

}

TY - JOUR

T1 - Simultaneously optimizing spatial spectral features based on mutual information for EEG classification

AU - Meng, Jianjun

AU - Yao, Lin

AU - Sheng, Xinjun

AU - Zhang, Dingguo

AU - Zhu, Xiangyang

PY - 2015/1/1

Y1 - 2015/1/1

N2 - High performance of the brain-computer interface (BCI) needs efficient algorithms to extract discriminative features from raw electroencephalography (EEG) signals. In this paper, we present a novel scheme to extract spatial spectral features for the motor imagery-based BCI. The learning task is formulated by maximizing the mutual information between spatial spectral features (MMISS) and class labels, by which a unique objective function directly related to Bayes classification error is optimized. The spatial spectral features are assumed to follow a parametric Gaussian distribution, which has been validated by the normal distribution Mardia's test, and under this assumption the estimation of mutual information is derived. We propose a gradient based alternative and iterative learning algorithm to optimize the cost function and derive the spatial and spectral filters simultaneously. The experimental results on dataset IVa of BCI competition III and dataset IIa of BCI competition IV show that the proposed MMISS is able to efficiently extract discriminative features from motor imagery-based EEG signals to enhance the classification accuracy compared to other existing algorithms.

AB - High performance of the brain-computer interface (BCI) needs efficient algorithms to extract discriminative features from raw electroencephalography (EEG) signals. In this paper, we present a novel scheme to extract spatial spectral features for the motor imagery-based BCI. The learning task is formulated by maximizing the mutual information between spatial spectral features (MMISS) and class labels, by which a unique objective function directly related to Bayes classification error is optimized. The spatial spectral features are assumed to follow a parametric Gaussian distribution, which has been validated by the normal distribution Mardia's test, and under this assumption the estimation of mutual information is derived. We propose a gradient based alternative and iterative learning algorithm to optimize the cost function and derive the spatial and spectral filters simultaneously. The experimental results on dataset IVa of BCI competition III and dataset IIa of BCI competition IV show that the proposed MMISS is able to efficiently extract discriminative features from motor imagery-based EEG signals to enhance the classification accuracy compared to other existing algorithms.

UR - http://www.scopus.com/inward/record.url?scp=84919935225&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84919935225&partnerID=8YFLogxK

U2 - 10.1109/TBME.2014.2345458

DO - 10.1109/TBME.2014.2345458

M3 - Article

VL - 62

SP - 227

EP - 240

JO - IEEE Transactions on Biomedical Engineering

JF - IEEE Transactions on Biomedical Engineering

SN - 0018-9294

IS - 1

M1 - 6871337

ER -