Experiments on Using combined short window bivariate autoregression for EEG classification

Research output: A Conference proceeding or a Chapter in BookConference contribution

3 Citations (Scopus)

Abstract

In EEG-based classification problem, most of currently used features are univariate and extracted from single channels. However EEG signals recorded from multiple channels for a brain activity are correlated, features extracted from the EEG signals should reflect relationships among those channels. For this reason, we propose and apply a bivariate feature called Combined Short-Window BiVariate AutoRegres-sive model (CSWBVAR) for EEG classification problems. Given a pair of channels, we firstly divide each of them in to overlapping segments or short windows, and then estimate BVAR parameters for each pair of segments. CSWBVAR is formed by combining extracted BVAR parameters together with a pre-defined overlapping window parameter. We analyzed and compared CSWBVAR feature and univariate feature using the dataset III for motor imagery problem of BCI Competition II (2003). Preliminary results show that using CSWBVAR feature can improve classification accuracy up to 7% comparing with using univariate one with the same linear-kernel SVM classifier
Original languageEnglish
Title of host publicationProceeding Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2011
EditorsMetin Akay
Place of PublicationBoston
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages372-375
Number of pages4
Volume1
ISBN (Print)9781424441402
DOIs
Publication statusPublished - 2011
Event5th International IEEE EMBS Conference on Neural Engineering - Cancun, Cancun, Mexico
Duration: 27 Apr 20111 May 2011

Conference

Conference5th International IEEE EMBS Conference on Neural Engineering
Abbreviated titleNER 2011
CountryMexico
CityCancun
Period27/04/111/05/11

Fingerprint

Electroencephalography
Experiments
Brain
Classifiers

Cite this

Tran, D., Huang, X., & Sharma, D. (2011). Experiments on Using combined short window bivariate autoregression for EEG classification. In M. Akay (Ed.), Proceeding Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2011 (Vol. 1, pp. 372-375). Boston: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/NER.2011.5910564
Tran, Dat ; Huang, Xu ; Sharma, Dharmendra. / Experiments on Using combined short window bivariate autoregression for EEG classification. Proceeding Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2011. editor / Metin Akay. Vol. 1 Boston : IEEE, Institute of Electrical and Electronics Engineers, 2011. pp. 372-375
@inproceedings{a1884f21254b491eb9db60078a038f2b,
title = "Experiments on Using combined short window bivariate autoregression for EEG classification",
abstract = "In EEG-based classification problem, most of currently used features are univariate and extracted from single channels. However EEG signals recorded from multiple channels for a brain activity are correlated, features extracted from the EEG signals should reflect relationships among those channels. For this reason, we propose and apply a bivariate feature called Combined Short-Window BiVariate AutoRegres-sive model (CSWBVAR) for EEG classification problems. Given a pair of channels, we firstly divide each of them in to overlapping segments or short windows, and then estimate BVAR parameters for each pair of segments. CSWBVAR is formed by combining extracted BVAR parameters together with a pre-defined overlapping window parameter. We analyzed and compared CSWBVAR feature and univariate feature using the dataset III for motor imagery problem of BCI Competition II (2003). Preliminary results show that using CSWBVAR feature can improve classification accuracy up to 7{\%} comparing with using univariate one with the same linear-kernel SVM classifier",
keywords = "Brain Modelling, EEG, Feature Extraction",
author = "Dat Tran and Xu Huang and Dharmendra Sharma",
year = "2011",
doi = "10.1109/NER.2011.5910564",
language = "English",
isbn = "9781424441402",
volume = "1",
pages = "372--375",
editor = "Metin Akay",
booktitle = "Proceeding Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2011",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
address = "United States",

}

Tran, D, Huang, X & Sharma, D 2011, Experiments on Using combined short window bivariate autoregression for EEG classification. in M Akay (ed.), Proceeding Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2011. vol. 1, IEEE, Institute of Electrical and Electronics Engineers, Boston, pp. 372-375, 5th International IEEE EMBS Conference on Neural Engineering, Cancun, Mexico, 27/04/11. https://doi.org/10.1109/NER.2011.5910564

Experiments on Using combined short window bivariate autoregression for EEG classification. / Tran, Dat; Huang, Xu; Sharma, Dharmendra.

Proceeding Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2011. ed. / Metin Akay. Vol. 1 Boston : IEEE, Institute of Electrical and Electronics Engineers, 2011. p. 372-375.

Research output: A Conference proceeding or a Chapter in BookConference contribution

TY - GEN

T1 - Experiments on Using combined short window bivariate autoregression for EEG classification

AU - Tran, Dat

AU - Huang, Xu

AU - Sharma, Dharmendra

PY - 2011

Y1 - 2011

N2 - In EEG-based classification problem, most of currently used features are univariate and extracted from single channels. However EEG signals recorded from multiple channels for a brain activity are correlated, features extracted from the EEG signals should reflect relationships among those channels. For this reason, we propose and apply a bivariate feature called Combined Short-Window BiVariate AutoRegres-sive model (CSWBVAR) for EEG classification problems. Given a pair of channels, we firstly divide each of them in to overlapping segments or short windows, and then estimate BVAR parameters for each pair of segments. CSWBVAR is formed by combining extracted BVAR parameters together with a pre-defined overlapping window parameter. We analyzed and compared CSWBVAR feature and univariate feature using the dataset III for motor imagery problem of BCI Competition II (2003). Preliminary results show that using CSWBVAR feature can improve classification accuracy up to 7% comparing with using univariate one with the same linear-kernel SVM classifier

AB - In EEG-based classification problem, most of currently used features are univariate and extracted from single channels. However EEG signals recorded from multiple channels for a brain activity are correlated, features extracted from the EEG signals should reflect relationships among those channels. For this reason, we propose and apply a bivariate feature called Combined Short-Window BiVariate AutoRegres-sive model (CSWBVAR) for EEG classification problems. Given a pair of channels, we firstly divide each of them in to overlapping segments or short windows, and then estimate BVAR parameters for each pair of segments. CSWBVAR is formed by combining extracted BVAR parameters together with a pre-defined overlapping window parameter. We analyzed and compared CSWBVAR feature and univariate feature using the dataset III for motor imagery problem of BCI Competition II (2003). Preliminary results show that using CSWBVAR feature can improve classification accuracy up to 7% comparing with using univariate one with the same linear-kernel SVM classifier

KW - Brain Modelling

KW - EEG

KW - Feature Extraction

U2 - 10.1109/NER.2011.5910564

DO - 10.1109/NER.2011.5910564

M3 - Conference contribution

SN - 9781424441402

VL - 1

SP - 372

EP - 375

BT - Proceeding Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2011

A2 - Akay, Metin

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - Boston

ER -

Tran D, Huang X, Sharma D. Experiments on Using combined short window bivariate autoregression for EEG classification. In Akay M, editor, Proceeding Annual International Conference of the IEEE Engineering in Medicine and Biology Society 2011. Vol. 1. Boston: IEEE, Institute of Electrical and Electronics Engineers. 2011. p. 372-375 https://doi.org/10.1109/NER.2011.5910564