Automated EEG feature selection for brain computer interfaces

Michael Schröder; Martin Bogdan; Wolfgang Rosenstiel; Thilo Hinterberger; Niels Birbaumer

doi:10.1109/CNE.2003.1196906

Automated EEG feature selection for brain computer interfaces

Michael Schröder, Martin Bogdan, Wolfgang Rosenstiel, Thilo Hinterberger, Niels Birbaumer

Istituto di Neuroriabilitazione Motoria San Camillo

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

A brain computer interface (BCI) utilizes signals derived from electroencephalography (EEG) to establish a connection between a person's state of mind and a computer based signal processing system that interprets the EEG signals. The choice of suitable features of the available EEG signals is crucial for good BCI communication. The optimal set of features is strongly dependent on the subjects and on the used experimental paradigm. Based upon EEG data of an existing BCI system, we present a wrapper method for the automated selection of features. The proposed method combines a genetic algorithm (GA) for the selection of feature with a support vector machine (SVM) for their evaluation. Applying this GA-SVM method to data of several subjects and two different experimental paradigms, we show that our approach leads to enhanced or even optimal classification accuracy.

Original language	English
Title of host publication	International IEEE/EMBS Conference on Neural Engineering, NER
Publisher	IEEE Computer Society
Pages	626-629
Number of pages	4
Volume	2003-January
ISBN (Print)	0780375793
DOIs	https://doi.org/10.1109/CNE.2003.1196906
Publication status	Published - 2003
Event	1st International IEEE EMBS Conference on Neural Engineering - Capri Island, Italy Duration: Mar 20 2003 → Mar 22 2003

Other

Other	1st International IEEE EMBS Conference on Neural Engineering
Country/Territory	Italy
City	Capri Island
Period	3/20/03 → 3/22/03

Keywords

Brain computer interfaces
Communication system control
Computer interfaces
Electroencephalography
Genetic algorithms
Neurons
Signal detection
Signal processing
Support vector machine classification
Support vector machines

ASJC Scopus subject areas

Artificial Intelligence
Mechanical Engineering

Access to Document

10.1109/CNE.2003.1196906

Cite this

Schröder, M, Bogdan, M, Rosenstiel, W, Hinterberger, T & Birbaumer, N 2003, Automated EEG feature selection for brain computer interfaces. in International IEEE/EMBS Conference on Neural Engineering, NER. vol. 2003-January, 1196906, IEEE Computer Society, pp. 626-629, 1st International IEEE EMBS Conference on Neural Engineering, Capri Island, Italy, 3/20/03. https://doi.org/10.1109/CNE.2003.1196906

@inproceedings{531057f5ec16431594339ab2a8a53a16,

title = "Automated EEG feature selection for brain computer interfaces",

abstract = "A brain computer interface (BCI) utilizes signals derived from electroencephalography (EEG) to establish a connection between a person's state of mind and a computer based signal processing system that interprets the EEG signals. The choice of suitable features of the available EEG signals is crucial for good BCI communication. The optimal set of features is strongly dependent on the subjects and on the used experimental paradigm. Based upon EEG data of an existing BCI system, we present a wrapper method for the automated selection of features. The proposed method combines a genetic algorithm (GA) for the selection of feature with a support vector machine (SVM) for their evaluation. Applying this GA-SVM method to data of several subjects and two different experimental paradigms, we show that our approach leads to enhanced or even optimal classification accuracy.",

keywords = "Brain computer interfaces, Communication system control, Computer interfaces, Electroencephalography, Genetic algorithms, Neurons, Signal detection, Signal processing, Support vector machine classification, Support vector machines",

author = "Michael Schr{\"o}der and Martin Bogdan and Wolfgang Rosenstiel and Thilo Hinterberger and Niels Birbaumer",

year = "2003",

doi = "10.1109/CNE.2003.1196906",

language = "English",

isbn = "0780375793",

volume = "2003-January",

pages = "626--629",

booktitle = "International IEEE/EMBS Conference on Neural Engineering, NER",

publisher = "IEEE Computer Society",

note = "1st International IEEE EMBS Conference on Neural Engineering ; Conference date: 20-03-2003 Through 22-03-2003",

}

TY - GEN

T1 - Automated EEG feature selection for brain computer interfaces

AU - Schröder, Michael

AU - Bogdan, Martin

AU - Rosenstiel, Wolfgang

AU - Hinterberger, Thilo

AU - Birbaumer, Niels

PY - 2003

Y1 - 2003

N2 - A brain computer interface (BCI) utilizes signals derived from electroencephalography (EEG) to establish a connection between a person's state of mind and a computer based signal processing system that interprets the EEG signals. The choice of suitable features of the available EEG signals is crucial for good BCI communication. The optimal set of features is strongly dependent on the subjects and on the used experimental paradigm. Based upon EEG data of an existing BCI system, we present a wrapper method for the automated selection of features. The proposed method combines a genetic algorithm (GA) for the selection of feature with a support vector machine (SVM) for their evaluation. Applying this GA-SVM method to data of several subjects and two different experimental paradigms, we show that our approach leads to enhanced or even optimal classification accuracy.

AB - A brain computer interface (BCI) utilizes signals derived from electroencephalography (EEG) to establish a connection between a person's state of mind and a computer based signal processing system that interprets the EEG signals. The choice of suitable features of the available EEG signals is crucial for good BCI communication. The optimal set of features is strongly dependent on the subjects and on the used experimental paradigm. Based upon EEG data of an existing BCI system, we present a wrapper method for the automated selection of features. The proposed method combines a genetic algorithm (GA) for the selection of feature with a support vector machine (SVM) for their evaluation. Applying this GA-SVM method to data of several subjects and two different experimental paradigms, we show that our approach leads to enhanced or even optimal classification accuracy.

KW - Brain computer interfaces

KW - Communication system control

KW - Computer interfaces

KW - Electroencephalography

KW - Genetic algorithms

KW - Neurons

KW - Signal detection

KW - Signal processing

KW - Support vector machine classification

KW - Support vector machines

UR - http://www.scopus.com/inward/record.url?scp=84942614432&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84942614432&partnerID=8YFLogxK

U2 - 10.1109/CNE.2003.1196906

DO - 10.1109/CNE.2003.1196906

M3 - Conference contribution

AN - SCOPUS:84942614432

SN - 0780375793

VL - 2003-January

SP - 626

EP - 629

BT - International IEEE/EMBS Conference on Neural Engineering, NER

PB - IEEE Computer Society

T2 - 1st International IEEE EMBS Conference on Neural Engineering

Y2 - 20 March 2003 through 22 March 2003

ER -

Automated EEG feature selection for brain computer interfaces

Abstract

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this