Using convolutional neural networks to automatically detect eye-blink artifacts in magnetoencephalography without resorting to electrooculography

Prabhat Garg, Elizabeth Davenport, Gowtham Murugesan, Ben Wagner, Christopher Whitlow, Joseph A Maldjian, Albert Montillo

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Magnetoencephelography (MEG) is a functional neuroimaging tool that records the magnetic fields induced by neuronal activity; however, signal from muscle activity often corrupts the data. Eye-blinks are one of the most common types of muscle artifact. They can be recorded by affixing eye proximal electrodes, as in electrooculography (EOG), however this complicates patient preparation and decreases comfort. Moreover, it can induce further muscular artifacts from facial twitching. We propose an EOG free, data driven approach. We begin with Independent Component Analysis (ICA), a well-known preprocessing approach that factors observed signal into statistically independent components. When applied to MEG, ICA can help separate neuronal components from non-neuronal ones, however, the components are randomly ordered. Thus, we develop a method to assign one of two labels, non-eye-blink or eye-blink, to each component. Our contributions are two-fold. First, we develop a 10-layer Convolutional Neural Network (CNN), which directly labels eye-blink artifacts. Second, we visualize the learned spatial features using attention mapping, to reveal what it has learned and bolster confidence in the method’s ability to generalize to unseen data. We acquired 8-min, eyes open, resting state MEG from 44 subjects. We trained our method on the spatial maps from ICA of 14 subjects selected randomly with expertly labeled ground truth. We then tested on the remaining 30 subjects. Our approach achieves a test classification accuracy of 99.67%, sensitivity: 97.62%, specificity: 99.77%, and ROC AUC: 98.69%. We also show the learned spatial features correspond to those human experts typically use which corroborates our model’s validity. This work (1) facilitates creation of fully automated processing pipelines in MEG that need to remove motion artifacts related to eye blinks, and (2) potentially obviates the use of additional EOG electrodes for the recording of eye-blinks in MEG studies.

Original languageEnglish (US)
Title of host publicationMedical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings
PublisherSpringer Verlag
Pages374-381
Number of pages8
Volume10435 LNCS
ISBN (Print)9783319661780
DOIs
StatePublished - Jan 1 2017
Event20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017 - Quebec City, Canada
Duration: Sep 11 2017Sep 13 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10435 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017
CountryCanada
CityQuebec City
Period9/11/179/13/17

Fingerprint

Electrooculography
Magnetoencephalography
Independent component analysis
Independent Component Analysis
Neural Networks
Neural networks
Muscle
Labels
Functional neuroimaging
Electrode
Electrodes
Neuroimaging
Pipelines
Data-driven
Magnetic fields
Specificity
Confidence
Preprocessing
Assign
Preparation

Keywords

  • Artifact
  • Automatic
  • CNN
  • Deep learning
  • EOG
  • Eye-Blink
  • MEG

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Garg, P., Davenport, E., Murugesan, G., Wagner, B., Whitlow, C., Maldjian, J. A., & Montillo, A. (2017). Using convolutional neural networks to automatically detect eye-blink artifacts in magnetoencephalography without resorting to electrooculography. In Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings (Vol. 10435 LNCS, pp. 374-381). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10435 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-66179-7_43

Using convolutional neural networks to automatically detect eye-blink artifacts in magnetoencephalography without resorting to electrooculography. / Garg, Prabhat; Davenport, Elizabeth; Murugesan, Gowtham; Wagner, Ben; Whitlow, Christopher; Maldjian, Joseph A; Montillo, Albert.

Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Vol. 10435 LNCS Springer Verlag, 2017. p. 374-381 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10435 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Garg, P, Davenport, E, Murugesan, G, Wagner, B, Whitlow, C, Maldjian, JA & Montillo, A 2017, Using convolutional neural networks to automatically detect eye-blink artifacts in magnetoencephalography without resorting to electrooculography. in Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. vol. 10435 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10435 LNCS, Springer Verlag, pp. 374-381, 20th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2017, Quebec City, Canada, 9/11/17. https://doi.org/10.1007/978-3-319-66179-7_43
Garg P, Davenport E, Murugesan G, Wagner B, Whitlow C, Maldjian JA et al. Using convolutional neural networks to automatically detect eye-blink artifacts in magnetoencephalography without resorting to electrooculography. In Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Vol. 10435 LNCS. Springer Verlag. 2017. p. 374-381. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-66179-7_43
Garg, Prabhat ; Davenport, Elizabeth ; Murugesan, Gowtham ; Wagner, Ben ; Whitlow, Christopher ; Maldjian, Joseph A ; Montillo, Albert. / Using convolutional neural networks to automatically detect eye-blink artifacts in magnetoencephalography without resorting to electrooculography. Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings. Vol. 10435 LNCS Springer Verlag, 2017. pp. 374-381 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{e9ba637e6ba347448118e862726a7b6f,
title = "Using convolutional neural networks to automatically detect eye-blink artifacts in magnetoencephalography without resorting to electrooculography",
abstract = "Magnetoencephelography (MEG) is a functional neuroimaging tool that records the magnetic fields induced by neuronal activity; however, signal from muscle activity often corrupts the data. Eye-blinks are one of the most common types of muscle artifact. They can be recorded by affixing eye proximal electrodes, as in electrooculography (EOG), however this complicates patient preparation and decreases comfort. Moreover, it can induce further muscular artifacts from facial twitching. We propose an EOG free, data driven approach. We begin with Independent Component Analysis (ICA), a well-known preprocessing approach that factors observed signal into statistically independent components. When applied to MEG, ICA can help separate neuronal components from non-neuronal ones, however, the components are randomly ordered. Thus, we develop a method to assign one of two labels, non-eye-blink or eye-blink, to each component. Our contributions are two-fold. First, we develop a 10-layer Convolutional Neural Network (CNN), which directly labels eye-blink artifacts. Second, we visualize the learned spatial features using attention mapping, to reveal what it has learned and bolster confidence in the method’s ability to generalize to unseen data. We acquired 8-min, eyes open, resting state MEG from 44 subjects. We trained our method on the spatial maps from ICA of 14 subjects selected randomly with expertly labeled ground truth. We then tested on the remaining 30 subjects. Our approach achieves a test classification accuracy of 99.67{\%}, sensitivity: 97.62{\%}, specificity: 99.77{\%}, and ROC AUC: 98.69{\%}. We also show the learned spatial features correspond to those human experts typically use which corroborates our model’s validity. This work (1) facilitates creation of fully automated processing pipelines in MEG that need to remove motion artifacts related to eye blinks, and (2) potentially obviates the use of additional EOG electrodes for the recording of eye-blinks in MEG studies.",
keywords = "Artifact, Automatic, CNN, Deep learning, EOG, Eye-Blink, MEG",
author = "Prabhat Garg and Elizabeth Davenport and Gowtham Murugesan and Ben Wagner and Christopher Whitlow and Maldjian, {Joseph A} and Albert Montillo",
year = "2017",
month = "1",
day = "1",
doi = "10.1007/978-3-319-66179-7_43",
language = "English (US)",
isbn = "9783319661780",
volume = "10435 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "374--381",
booktitle = "Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings",

}

TY - GEN

T1 - Using convolutional neural networks to automatically detect eye-blink artifacts in magnetoencephalography without resorting to electrooculography

AU - Garg, Prabhat

AU - Davenport, Elizabeth

AU - Murugesan, Gowtham

AU - Wagner, Ben

AU - Whitlow, Christopher

AU - Maldjian, Joseph A

AU - Montillo, Albert

PY - 2017/1/1

Y1 - 2017/1/1

N2 - Magnetoencephelography (MEG) is a functional neuroimaging tool that records the magnetic fields induced by neuronal activity; however, signal from muscle activity often corrupts the data. Eye-blinks are one of the most common types of muscle artifact. They can be recorded by affixing eye proximal electrodes, as in electrooculography (EOG), however this complicates patient preparation and decreases comfort. Moreover, it can induce further muscular artifacts from facial twitching. We propose an EOG free, data driven approach. We begin with Independent Component Analysis (ICA), a well-known preprocessing approach that factors observed signal into statistically independent components. When applied to MEG, ICA can help separate neuronal components from non-neuronal ones, however, the components are randomly ordered. Thus, we develop a method to assign one of two labels, non-eye-blink or eye-blink, to each component. Our contributions are two-fold. First, we develop a 10-layer Convolutional Neural Network (CNN), which directly labels eye-blink artifacts. Second, we visualize the learned spatial features using attention mapping, to reveal what it has learned and bolster confidence in the method’s ability to generalize to unseen data. We acquired 8-min, eyes open, resting state MEG from 44 subjects. We trained our method on the spatial maps from ICA of 14 subjects selected randomly with expertly labeled ground truth. We then tested on the remaining 30 subjects. Our approach achieves a test classification accuracy of 99.67%, sensitivity: 97.62%, specificity: 99.77%, and ROC AUC: 98.69%. We also show the learned spatial features correspond to those human experts typically use which corroborates our model’s validity. This work (1) facilitates creation of fully automated processing pipelines in MEG that need to remove motion artifacts related to eye blinks, and (2) potentially obviates the use of additional EOG electrodes for the recording of eye-blinks in MEG studies.

AB - Magnetoencephelography (MEG) is a functional neuroimaging tool that records the magnetic fields induced by neuronal activity; however, signal from muscle activity often corrupts the data. Eye-blinks are one of the most common types of muscle artifact. They can be recorded by affixing eye proximal electrodes, as in electrooculography (EOG), however this complicates patient preparation and decreases comfort. Moreover, it can induce further muscular artifacts from facial twitching. We propose an EOG free, data driven approach. We begin with Independent Component Analysis (ICA), a well-known preprocessing approach that factors observed signal into statistically independent components. When applied to MEG, ICA can help separate neuronal components from non-neuronal ones, however, the components are randomly ordered. Thus, we develop a method to assign one of two labels, non-eye-blink or eye-blink, to each component. Our contributions are two-fold. First, we develop a 10-layer Convolutional Neural Network (CNN), which directly labels eye-blink artifacts. Second, we visualize the learned spatial features using attention mapping, to reveal what it has learned and bolster confidence in the method’s ability to generalize to unseen data. We acquired 8-min, eyes open, resting state MEG from 44 subjects. We trained our method on the spatial maps from ICA of 14 subjects selected randomly with expertly labeled ground truth. We then tested on the remaining 30 subjects. Our approach achieves a test classification accuracy of 99.67%, sensitivity: 97.62%, specificity: 99.77%, and ROC AUC: 98.69%. We also show the learned spatial features correspond to those human experts typically use which corroborates our model’s validity. This work (1) facilitates creation of fully automated processing pipelines in MEG that need to remove motion artifacts related to eye blinks, and (2) potentially obviates the use of additional EOG electrodes for the recording of eye-blinks in MEG studies.

KW - Artifact

KW - Automatic

KW - CNN

KW - Deep learning

KW - EOG

KW - Eye-Blink

KW - MEG

UR - http://www.scopus.com/inward/record.url?scp=85029499206&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029499206&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-66179-7_43

DO - 10.1007/978-3-319-66179-7_43

M3 - Conference contribution

C2 - 31656959

AN - SCOPUS:85029499206

SN - 9783319661780

VL - 10435 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 374

EP - 381

BT - Medical Image Computing and Computer Assisted Intervention − MICCAI 2017 - 20th International Conference, Proceedings

PB - Springer Verlag

ER -