Exploring the advantages of multiband fMRI with simultaneous EEG to investigate coupling between gamma frequency neural activity and the BOLD response in humans

Abstract

We established an optimal combination of EEG recording during sparse multiband (MB) fMRI that preserves high-resolution, whole-brain fMRI coverage while enabling broad-band EEG recordings which are uncorrupted by MRI gradient artefacts (GAs). We first determined the safety of simultaneous EEG recording during MB fMRI. Application of MB factor = 4 produced <1°C peak heating of electrode/hardware during 20 min of GE-EPI data acquisition. However, higher SAR sequences require specific safety testing, with greater heating observed using PCASL with MB factor = 4. Heating was greatest in the electrocardiogram channel, likely due to it possessing longest lead length. We investigated the effect of MB factor on the temporal signal-to-noise ratio for a range of GE-EPI sequences (varying MB factor and temporal interval between slice acquisitions). We found that, for our experimental purpose, the optimal acquisition was achieved with MB factor = 3, 3mm isotropic voxels, and 33 slices providing whole head coverage. This sequence afforded a 2.25 s duration quiet period (without GAs) in every 3 s TR. Using this sequence, we demonstrated the ability to record gamma frequency (55-80 Hz) EEG oscillations, in response to right index finger abduction, that are usually obscured by GAs during continuous fMRI data acquisition. In this novel application of EEG-MB fMRI to a motor task, we observed a positive correlation between gamma and BOLD responses in bilateral motor regions. These findings support and extend previous work regarding coupling between neural and hemodynamic measures of brain activity in humans and showcase the utility of EEG-MB fMRI for future investigations.

The relationship between negative BOLD responses and ERS and ERD of alpha/beta oscillations in visual and motor cortex

“…Therefore the source of this discrepancy between EEG and fMRI response polarities remains unresolved, however it is likely that the stronger contralateral α and β ERD are associated with the widely reported contralateral gamma ERS (Muthukumaraswamy 2010) and tentative reports suggest a gamma ERD occurring ipsilateral to finger movements (Huo, Xiang et al 2010) would appear a plausible correlate of ipsilateral NBR in line with rodent work (Boorman, Kennerley et al 2010, Boorman, Harris et al 2015. A few previous studies have measured gamma EEG responses using sparse fMRI sequences (Mulert, Leicht et al 2010, Scheeringa, Fries et al 2011, including a recent demonstration that gamma ERS during motor abductions is positively correlated with IM motor PBR (Uji, Wilson et al 2018), however the implementation of the DABS sequence employed here was not designed to be used in that manner.…”

Dissociating neural activity associated with the subjective phenomenology of monocular stereopsis: an EEG study

“…These limiting factors likely negatively impacted on the sensitivity and specificity with which we were able to identifying the optimal gamma VE location. We are therefore more confident of the derived alpha VE locations particularly because previous studies that have had better SNRs (number of trials) and/or use of individual anatomical brain images indicate that alpha and gamma VE locations are spatially co-located (Ball et al, 2008; Cheyne et al, 2008; Cheyne and Ferrari, 2013; Cheyne, 2013; Donner and Siegel, 2011; Hoogenboom et al, 2006; Muthukumaraswamy, 2013, 2010; Muthukumaraswamy and Singh, 2013; Scheeringa et al, 2011; Uji et al, 2018).…”

“…While the alpha power change is sustained over longer periods of time and typically observable in EEG with robust SNR, gamma power change is more short lived and has low SNR in the EEG signal (Darvas et al, 2010; Miller et al, 2007). Our study design precluded recording a large number of trials within a workable experimental session duration, and thus we may not have achieved the same level of SNRs required for gamma band source localization that has been achieved in previous studies which typically record in excess of 100 trials per condition (Ball et al, 2008; Cheyne et al, 2008; Cheyne and Ferrari, 2013; Cheyne, 2013; Hoogenboom et al, 2006; Muthukumaraswamy, 2013, 2010; Muthukumaraswamy and Singh, 2013; Scheeringa et al, 2011; Uji et al, 2018); more trials leads to better SNRs (e.g., 120 trials using individual BEM head model, see details Uji et al, 2018). Furthermore, in this study, we used the MNI standard anatomical image and standard EEG electrode locations to construct the BEM head model estimating the neural activities in the brain instead of using individual anatomical brain images and digitizing the EEG electrodes locations (see Uji et al, 2018).…”

“…The significant difference in gamma synchronization between 3D and 2D images that we observed under monocular aperture viewing was obtained at the alpha band virtual electrode (VE) location. Alpha and gamma band activities are thought to reflect different but related processes (Buschman and Miller, 2007; Buzsaki and Draguhn, 2004; Colgin et al, 2009; Fries, 2009; Jensen et al, 2014; Klimesch, 1999; Pfurtscheller et al, 1996; Pfurtscheller and Lopes da Silva, 1999; Singer and Gray, 1995), and the alpha VE and gamma VE locations are typically co-located (Ball et al, 2008; Cheyne et al, 2008; Cheyne and Ferrari, 2013; Cheyne, 2013; Donner and Siegel, 2011; Hoogenboom et al, 2006; Muthukumaraswamy, 2013, 2010; Muthukumaraswamy and Singh, 2013; Scheeringa et al, 2011; Uji et al, 2018). One question regarding our results might be why we found no significant effects of gamma ERS at the parietal gamma VE location itself.…”

Linking visual gamma to task‐related brain networks—a simultaneous EEG‐fMRI study

“…Residuals from MR gradient artifact correction force experimenters to implement fMRI recordings in the interleaved fashion, allowing for gradient‐free EEG periods. Consequently, obtaining high‐frequency EEG activity is compromised by shortening the volume acquisition (Uji, Wilson, Francis, Mullinger, & Mayhew, ) and often selecting only part of the brain volume (Scheeringa et al, , ). In such cases, only voxel‐wise analyses can be used to account for BOLD variance.…”