Facial muscle activity contaminates EEG signal at rest: Evidence from frontalis and temporalis motor units


Yilmaz G., Budan A. S., Ungan P., Topkara B., Türker K. S.

Journal of Neural Engineering, vol.16, no.6, 2019 (SCI-Expanded) identifier identifier identifier

  • Publication Type: Article / Article
  • Volume: 16 Issue: 6
  • Publication Date: 2019
  • Doi Number: 10.1088/1741-2552/ab3235
  • Journal Name: Journal of Neural Engineering
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Keywords: electromyography, electroencephalography, single motor unit, coherence
  • Istanbul Gelisim University Affiliated: No

Abstract

Objective. In order to reach electroencephalography (EEG) electrodes on the scalp, synchronized activity of neurons needs to pass thorough several tissue layers, including the skull and muscles covering the scalp. The contamination of EEG signal by temporalis and frontalis muscles has been well documented for voluntary muscle contraction even at low contraction levels. The extent of myogenic contamination during postural and/or rest activity of the temporalis and frontalis remains an impediment for EEG research. Approach. In this study, we first aimed to observe involuntary, continuous motor unit activity of the frontalis muscle at rest and evaluate motor unit level frontalis interference on the EEG electrodes. Second, we compared motor unit interference from the frontalis before and after artefact pruning via an independent component analysis (ICA) algorithm. Main results. We demonstrated that motor unit activity of the frontalis muscle produces interference potentials on the frontal electrodes at rest and the interference was significantly reduced after ICA on the frontal electrodes, but not completely eliminated. Likewise, the temporalis interference at rest was significantly smaller after ICA on the fronto-Temporal electrodes, but not completely removed. Significance. We documented the existence of resting involuntary activity of the temporalis and frontalis muscles underneath EEG electrodes and the removal of the EEG signal from their contiguous interference is not possible even after the use of ICA technology. We recommend that EEG researchers readdress the definition of 'rest' for EEG recordings and the ICA experts should extend their electromyography removal strategies to motor unit level interference.