Removing artefacts from TMS-EEG recordings using independent component analysis: Importance for assessing prefrontal and motor cortex network properties
Nigel C. Rogasch,Richard H. Thomson,Faranak Farzan,Bernadette M. Fitzgibbon,Neil W. Bailey,Julio C. Hernandez-Pavon,Zafiris J. Daskalakis,Paul B. Fitzgerald
DOI: https://doi.org/10.1016/j.neuroimage.2014.07.037
IF: 5.7
2014-11-01
NeuroImage
Abstract:INTRODUCTION: The combination of transcranial magnetic stimulation and electroencephalography (TMS-EEG) is emerging as a powerful tool for causally investigating cortical mechanisms and networks. However, various artefacts contaminate TMS-EEG recordings, particularly over regions such as the dorsolateral prefrontal cortex (DLPFC). The aim of this study was to substantiate removal of artefacts from TMS-EEG recordings following stimulation of the DLPFC and motor cortex using independent component analysis (ICA).METHODS: 36 healthy volunteers (30.8 ± 9 years, 9 female) received 75 single TMS pulses to the left DLPFC or left motor cortex while EEG was recorded from 57 electrodes. A subset of 9 volunteers also received 50 sham pulses. The large TMS artefact and early muscle activity (-2 to ~15 ms) were removed using interpolation and the remaining EEG signal was processed in two separate ICA runs using the FastICA algorithm. Five sub-types of TMS-related artefacts were manually identified: remaining muscle artefacts, decay artefacts, blink artefacts, auditory-evoked potentials and other noise-related artefacts. The cause of proposed blink and auditory-evoked potentials was assessed by concatenating known artefacts (i.e. voluntary blinks or auditory-evoked potentials resulting from sham TMS) to the TMS trials before ICA and evaluating grouping of resultant independent components (ICs). Finally, we assessed the effect of removing specific artefact types on TMS-evoked potentials (TEPs) and TMS-evoked oscillations.RESULTS: Over DLPFC, ICs from proposed muscle and decay artefacts correlated with TMS-evoked muscle activity size, whereas proposed TMS-evoked blink ICs combined with voluntary blinks and auditory ICs with auditory-evoked potentials from sham TMS. Individual artefact sub-types characteristically distorted each measure of DLPFC function across the scalp. When free of artefact, TEPs and TMS-evoked oscillations could be measured following DLPFC stimulation. Importantly, characteristic TEPs following motor cortex stimulation (N15, P30, N45, P60, N100) could be recovered from artefactual data, corroborating the reliability of ICA-based artefact correction.CONCLUSIONS: Various different artefacts contaminate TMS-EEG recordings over the DLPFC and motor cortex. However, these artefacts can be removed with apparent minimal impact on neural activity using ICA, allowing the study of TMS-evoked cortical network properties.
radiology, nuclear medicine & medical imaging,neurosciences,neuroimaging