Multichannel sleep spindle detection using sparse low-rank optimization

Ankit Parekh, Ivan Selesnick, Ricardo S. Osorio, Andrew W. Varga, David M. Rapoport, Indu Ayappa

Research output: Contribution to journalArticle

Abstract

Background Automated single-channel spindle detectors, for human sleep EEG, are blind to the presence of spindles in other recorded channels unlike visual annotation by a human expert. New method We propose a multichannel spindle detection method that aims to detect global and local spindle activity in human sleep EEG. Using a non-linear signal model, which assumes the input EEG to be the sum of a transient and an oscillatory component, we propose a multichannel transient separation algorithm. Consecutive overlapping blocks of the multichannel oscillatory component are assumed to be low-rank whereas the transient component is assumed to be piecewise constant with a zero baseline. The estimated oscillatory component is used in conjunction with a bandpass filter and the Teager operator for detecting sleep spindles. Results and comparison with other methods The proposed method is applied to two publicly available databases and compared with 7 existing single-channel automated detectors. F1 scores for the proposed spindle detection method averaged 0.66 (0.02) and 0.62 (0.06) for the two databases, respectively. For an overnight 6 channel EEG signal, the proposed algorithm takes about 4 min to detect sleep spindles simultaneously across all channels with a single setting of corresponding algorithmic parameters. Conclusions The proposed method attempts to mimic and utilize, for better spindle detection, a particular human expert behavior where the decision to mark a spindle event may be subconsciously influenced by the presence of a spindle in EEG channels other than the central channel visible on a digital screen.

Original languageEnglish (US)
Pages (from-to)1-16
Number of pages16
JournalJournal of Neuroscience Methods
Volume288
DOIs
StatePublished - Aug 15 2017

Fingerprint

Sleep
Electroencephalography
Databases
Nonlinear Dynamics
Human Activities

Keywords

  • Convex optimization
  • Multichannel signal processing
  • Sleep EEG
  • Sparse signal
  • Spindle detection

ASJC Scopus subject areas

  • Neuroscience(all)

Cite this

Multichannel sleep spindle detection using sparse low-rank optimization. / Parekh, Ankit; Selesnick, Ivan; Osorio, Ricardo S.; Varga, Andrew W.; Rapoport, David M.; Ayappa, Indu.

In: Journal of Neuroscience Methods, Vol. 288, 15.08.2017, p. 1-16.

Research output: Contribution to journalArticle

Parekh, Ankit ; Selesnick, Ivan ; Osorio, Ricardo S. ; Varga, Andrew W. ; Rapoport, David M. ; Ayappa, Indu. / Multichannel sleep spindle detection using sparse low-rank optimization. In: Journal of Neuroscience Methods. 2017 ; Vol. 288. pp. 1-16.
@article{e6ccc357994542089f72c9cf25063c06,
title = "Multichannel sleep spindle detection using sparse low-rank optimization",
abstract = "Background Automated single-channel spindle detectors, for human sleep EEG, are blind to the presence of spindles in other recorded channels unlike visual annotation by a human expert. New method We propose a multichannel spindle detection method that aims to detect global and local spindle activity in human sleep EEG. Using a non-linear signal model, which assumes the input EEG to be the sum of a transient and an oscillatory component, we propose a multichannel transient separation algorithm. Consecutive overlapping blocks of the multichannel oscillatory component are assumed to be low-rank whereas the transient component is assumed to be piecewise constant with a zero baseline. The estimated oscillatory component is used in conjunction with a bandpass filter and the Teager operator for detecting sleep spindles. Results and comparison with other methods The proposed method is applied to two publicly available databases and compared with 7 existing single-channel automated detectors. F1 scores for the proposed spindle detection method averaged 0.66 (0.02) and 0.62 (0.06) for the two databases, respectively. For an overnight 6 channel EEG signal, the proposed algorithm takes about 4 min to detect sleep spindles simultaneously across all channels with a single setting of corresponding algorithmic parameters. Conclusions The proposed method attempts to mimic and utilize, for better spindle detection, a particular human expert behavior where the decision to mark a spindle event may be subconsciously influenced by the presence of a spindle in EEG channels other than the central channel visible on a digital screen.",
keywords = "Convex optimization, Multichannel signal processing, Sleep EEG, Sparse signal, Spindle detection",
author = "Ankit Parekh and Ivan Selesnick and Osorio, {Ricardo S.} and Varga, {Andrew W.} and Rapoport, {David M.} and Indu Ayappa",
year = "2017",
month = "8",
day = "15",
doi = "10.1016/j.jneumeth.2017.06.004",
language = "English (US)",
volume = "288",
pages = "1--16",
journal = "Journal of Neuroscience Methods",
issn = "0165-0270",
publisher = "Elsevier",

}

TY - JOUR

T1 - Multichannel sleep spindle detection using sparse low-rank optimization

AU - Parekh, Ankit

AU - Selesnick, Ivan

AU - Osorio, Ricardo S.

AU - Varga, Andrew W.

AU - Rapoport, David M.

AU - Ayappa, Indu

PY - 2017/8/15

Y1 - 2017/8/15

N2 - Background Automated single-channel spindle detectors, for human sleep EEG, are blind to the presence of spindles in other recorded channels unlike visual annotation by a human expert. New method We propose a multichannel spindle detection method that aims to detect global and local spindle activity in human sleep EEG. Using a non-linear signal model, which assumes the input EEG to be the sum of a transient and an oscillatory component, we propose a multichannel transient separation algorithm. Consecutive overlapping blocks of the multichannel oscillatory component are assumed to be low-rank whereas the transient component is assumed to be piecewise constant with a zero baseline. The estimated oscillatory component is used in conjunction with a bandpass filter and the Teager operator for detecting sleep spindles. Results and comparison with other methods The proposed method is applied to two publicly available databases and compared with 7 existing single-channel automated detectors. F1 scores for the proposed spindle detection method averaged 0.66 (0.02) and 0.62 (0.06) for the two databases, respectively. For an overnight 6 channel EEG signal, the proposed algorithm takes about 4 min to detect sleep spindles simultaneously across all channels with a single setting of corresponding algorithmic parameters. Conclusions The proposed method attempts to mimic and utilize, for better spindle detection, a particular human expert behavior where the decision to mark a spindle event may be subconsciously influenced by the presence of a spindle in EEG channels other than the central channel visible on a digital screen.

AB - Background Automated single-channel spindle detectors, for human sleep EEG, are blind to the presence of spindles in other recorded channels unlike visual annotation by a human expert. New method We propose a multichannel spindle detection method that aims to detect global and local spindle activity in human sleep EEG. Using a non-linear signal model, which assumes the input EEG to be the sum of a transient and an oscillatory component, we propose a multichannel transient separation algorithm. Consecutive overlapping blocks of the multichannel oscillatory component are assumed to be low-rank whereas the transient component is assumed to be piecewise constant with a zero baseline. The estimated oscillatory component is used in conjunction with a bandpass filter and the Teager operator for detecting sleep spindles. Results and comparison with other methods The proposed method is applied to two publicly available databases and compared with 7 existing single-channel automated detectors. F1 scores for the proposed spindle detection method averaged 0.66 (0.02) and 0.62 (0.06) for the two databases, respectively. For an overnight 6 channel EEG signal, the proposed algorithm takes about 4 min to detect sleep spindles simultaneously across all channels with a single setting of corresponding algorithmic parameters. Conclusions The proposed method attempts to mimic and utilize, for better spindle detection, a particular human expert behavior where the decision to mark a spindle event may be subconsciously influenced by the presence of a spindle in EEG channels other than the central channel visible on a digital screen.

KW - Convex optimization

KW - Multichannel signal processing

KW - Sleep EEG

KW - Sparse signal

KW - Spindle detection

UR - http://www.scopus.com/inward/record.url?scp=85021162263&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021162263&partnerID=8YFLogxK

U2 - 10.1016/j.jneumeth.2017.06.004

DO - 10.1016/j.jneumeth.2017.06.004

M3 - Article

VL - 288

SP - 1

EP - 16

JO - Journal of Neuroscience Methods

JF - Journal of Neuroscience Methods

SN - 0165-0270

ER -