Comparison of feature and classifier algorithms for online automatic sleep staging based on a single EEG signal

Mustafa Radha, Gary Garcia Molina, Mannes Poel, Giulio Tononi

Abstract

Automatic sleep staging on an online basis has recently emerged as a research topic motivated by fundamental sleep research. The aim of this paper is to find optimal signal processing methods and machine learning algorithms to achieve online sleep staging on the basis of a single EEG signal. The classification performance obtained using six different EEG signals and various signal processing feature sets is compared using the kappa statistic which has very recently become popular in sleep staging research. A variable duration of the EEG segment (or epoch) to decide on the sleep stage is also analyzed. Spectral-domain, time-domain, linear, and nonlinear features are compared in terms of performance and two types of machine learning approaches (random forests and support vector machines) are assessed. We have determined that frontal EEG signals, with spectral linear features, epoch durations between 18 and 30 seconds, and a random forest classifier lead to optimal classification performance while ensuring real-time online operation.
Original languageUndefined
Title of host publicationProceedings of the 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014
Place of PublicationUSA
PublisherIEEE Computer Society
Pages1876-1880
Number of pages5
ISBN (Print)978-1-4244-7929-0
DOIs
StatePublished - Aug 2014
Event36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014 - Chicago, United States

Publication series

Name
PublisherIEEE Computer Society
ISSN (Print)1557-170X

Conference

Conference36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014
Abbreviated titleEMBC
CountryUnited States
CityChicago
Period26/08/1430/08/14

Fingerprint

Electroencephalography
Sleep research
Learning systems
Signal processing
Learning algorithms
Support vector machines
Classifiers
Statistics

Keywords

  • EWI-25392
  • METIS-309706
  • IR-94098

Cite this

Radha, M., Garcia Molina, G., Poel, M., & Tononi, G. (2014). Comparison of feature and classifier algorithms for online automatic sleep staging based on a single EEG signal. In Proceedings of the 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014 (pp. 1876-1880). USA: IEEE Computer Society. DOI: 10.1109/EMBC.2014.6943976

Radha, Mustafa; Garcia Molina, Gary; Poel, Mannes; Tononi, Giulio / Comparison of feature and classifier algorithms for online automatic sleep staging based on a single EEG signal.

Proceedings of the 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014. USA : IEEE Computer Society, 2014. p. 1876-1880.

Research output: Scientific - peer-reviewConference contribution

@inbook{49e77e3bd7c14868b83077662a6e812e,
title = "Comparison of feature and classifier algorithms for online automatic sleep staging based on a single EEG signal",
abstract = "Automatic sleep staging on an online basis has recently emerged as a research topic motivated by fundamental sleep research. The aim of this paper is to find optimal signal processing methods and machine learning algorithms to achieve online sleep staging on the basis of a single EEG signal. The classification performance obtained using six different EEG signals and various signal processing feature sets is compared using the kappa statistic which has very recently become popular in sleep staging research. A variable duration of the EEG segment (or epoch) to decide on the sleep stage is also analyzed. Spectral-domain, time-domain, linear, and nonlinear features are compared in terms of performance and two types of machine learning approaches (random forests and support vector machines) are assessed. We have determined that frontal EEG signals, with spectral linear features, epoch durations between 18 and 30 seconds, and a random forest classifier lead to optimal classification performance while ensuring real-time online operation.",
keywords = "EWI-25392, METIS-309706, IR-94098",
author = "Mustafa Radha and {Garcia Molina}, Gary and Mannes Poel and Giulio Tononi",
note = "10.1109/EMBC.2014.6943976",
year = "2014",
month = "8",
doi = "10.1109/EMBC.2014.6943976",
isbn = "978-1-4244-7929-0",
publisher = "IEEE Computer Society",
pages = "1876--1880",
booktitle = "Proceedings of the 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014",
address = "United States",

}

Radha, M, Garcia Molina, G, Poel, M & Tononi, G 2014, Comparison of feature and classifier algorithms for online automatic sleep staging based on a single EEG signal. in Proceedings of the 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014. IEEE Computer Society, USA, pp. 1876-1880, 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014, Chicago, United States, 26-30 August. DOI: 10.1109/EMBC.2014.6943976

Comparison of feature and classifier algorithms for online automatic sleep staging based on a single EEG signal. / Radha, Mustafa; Garcia Molina, Gary; Poel, Mannes; Tononi, Giulio.

Proceedings of the 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014. USA : IEEE Computer Society, 2014. p. 1876-1880.

Research output: Scientific - peer-reviewConference contribution

TY - CHAP

T1 - Comparison of feature and classifier algorithms for online automatic sleep staging based on a single EEG signal

AU - Radha,Mustafa

AU - Garcia Molina,Gary

AU - Poel,Mannes

AU - Tononi,Giulio

N1 - 10.1109/EMBC.2014.6943976

PY - 2014/8

Y1 - 2014/8

N2 - Automatic sleep staging on an online basis has recently emerged as a research topic motivated by fundamental sleep research. The aim of this paper is to find optimal signal processing methods and machine learning algorithms to achieve online sleep staging on the basis of a single EEG signal. The classification performance obtained using six different EEG signals and various signal processing feature sets is compared using the kappa statistic which has very recently become popular in sleep staging research. A variable duration of the EEG segment (or epoch) to decide on the sleep stage is also analyzed. Spectral-domain, time-domain, linear, and nonlinear features are compared in terms of performance and two types of machine learning approaches (random forests and support vector machines) are assessed. We have determined that frontal EEG signals, with spectral linear features, epoch durations between 18 and 30 seconds, and a random forest classifier lead to optimal classification performance while ensuring real-time online operation.

AB - Automatic sleep staging on an online basis has recently emerged as a research topic motivated by fundamental sleep research. The aim of this paper is to find optimal signal processing methods and machine learning algorithms to achieve online sleep staging on the basis of a single EEG signal. The classification performance obtained using six different EEG signals and various signal processing feature sets is compared using the kappa statistic which has very recently become popular in sleep staging research. A variable duration of the EEG segment (or epoch) to decide on the sleep stage is also analyzed. Spectral-domain, time-domain, linear, and nonlinear features are compared in terms of performance and two types of machine learning approaches (random forests and support vector machines) are assessed. We have determined that frontal EEG signals, with spectral linear features, epoch durations between 18 and 30 seconds, and a random forest classifier lead to optimal classification performance while ensuring real-time online operation.

KW - EWI-25392

KW - METIS-309706

KW - IR-94098

U2 - 10.1109/EMBC.2014.6943976

DO - 10.1109/EMBC.2014.6943976

M3 - Conference contribution

SN - 978-1-4244-7929-0

SP - 1876

EP - 1880

BT - Proceedings of the 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014

PB - IEEE Computer Society

ER -

Radha M, Garcia Molina G, Poel M, Tononi G. Comparison of feature and classifier algorithms for online automatic sleep staging based on a single EEG signal. In Proceedings of the 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014. USA: IEEE Computer Society. 2014. p. 1876-1880. Available from, DOI: 10.1109/EMBC.2014.6943976