Cerebral Cortex, May 2018;28: 1597–1609

doi: 10.1093/cercor/bhx055
Advance Access Publication Date: 10 March 2017
Original Article

ORIGINAL ARTICLE

tDCS Modulates Visual Gamma Oscillations

and Basal Alpha Activity in Occipital Cortices:
Evidence from MEG
Tony W. Wilson1,2,3, Timothy J. McDermott3, Mackenzie S. Mills3,
Nathan M. Coolidge3 and Elizabeth Heinrichs-Graham1,3
1
Department of Neurological Sciences, University of Nebraska Medical Center (UNMC), Omaha, NE, USA,
2
Department of Pharmacology and Experimental Neurosciences, UNMC, Omaha, NE, USA and 3Center for
Magnetoencephalography, UNMC, Omaha, NE 68198, USA
Address correspondence to Tony W. Wilson, Center for Magnetoencephalography, University of Nebraska Medical Center, 988422 Nebraska Medical
Center, Omaha, NE 68198, USA. Email: tony.w.wilson@gmail.com

Abstract
Transcranial direct-current stimulation (tDCS) is now a widely used method for modulating the human brain, but the
resulting physiological effects are not understood. Recent studies have combined magnetoencephalography (MEG) with
simultaneous tDCS to evaluate online changes in occipital alpha and gamma oscillations, but no study to date has
quantified the offline (i.e., after tDCS) alterations in these responses. Thirty-five healthy adults received active or sham
anodal tDCS to the occipital cortices, and then completed a visual stimulation paradigm during MEG that is known to elicit
robust gamma and alpha oscillations. The resulting MEG data were imaged and peak voxel time series were extracted to
evaluate tDCS effects. We found that tDCS to the occipital increased the amplitude of local gamma oscillations, and basal
alpha levels during the baseline. tDCS was also associated with network-level effects, including increased gamma
oscillations in the prefrontal cortex, parietal, and other visual attention regions. Finally, although tDCS did not modulate
peak gamma frequency, this variable was inversely correlated with gamma amplitude, which is consistent with a GABA-
gamma link. In conclusion, tDCS alters gamma oscillations and basal alpha levels. The net offline effects on gamma activity
are consistent with the view that anodal tDCS decreases local GABA.

Key words: direct-current stimulation, ERD, ERSGABA, magnetoencephalography

Introduction of stimulated neurons (Filmer et al. 2014; Fertonani and

Transcranial direct-current stimulation (tDCS) is a neuromodu-
latory method that involves delivering low amplitude, direct-
current to specific areas of the brain using sponges or small
electrodes. The stimulation is typically bidirectional with one
electrode acting as the anode and the other as the cathode, and
this arrangement or “montage” is thought to create a semi
current-loop through the brain running from the anode to the
cathode. Of note, this stimulation is not strong enough to elicit
action potentials, but is believed to alter the response threshold
Miniussi 2016). In the most simplistic dosing model, anodal
stimulation increases the excitability of the underlying brain
area by decreasing GABAergic activity, while cathodal stimula-
tion decreases neuronal excitability in the underlying tissues
by decreasing glutamate (Nitsche and Paulus 2000, 2001;
Nitsche et al. 2003; Coffman et al. 2014; Filmer et al. 2014;
Fertonani and Miniussi 2016). Data supporting this general
framework has come from early animal studies (Bindman et al.
1964; Purpura and McMurtry 1965), human pharmacological

© The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com
 1598 | Cerebral Cortex, 2018, Vol. 28, No. 5
studies (Liebetanz et al. 2002; Nitsche et al. 2004; many others,
reviewed in Stagg and Nitsche 2011; Medeiros et al. 2012), tran-
scranial magnetic stimulation (TMS; Di Lazzaro et al. 2005;
Nitsche et al. 2005), and magnetic resonance spectroscopy
(MRS; Stagg et al. 2009a, 2011; Kim et al. 2014; Bachtiar et al.
2015). However, there are other views on the source of tDCS-
induced excitability changes (Jackson et al. 2016; Lafon et al.
2017), and knowledge of the cellular and molecular neurobiol-
ogy of tDCS continues to evolve.
Numerous behavioral studies have evaluated how tDCS
affects visual processing in healthy participants. These studies
have shown that applying anodal tDCS to the occipital cortices
improves visual detection thresholds (Kraft et al. 2010; Olma
et al. 2011), reduces phosphine detection levels (Antal et al.
2003; Lang et al. 2007), enhances face and object perception
(Renzi et al. 2015; Barbieri et al. 2016), and facilitates visuo-
motor coordination (Antal et al. 2004). The physiological altera-
tions that underlie these behavioral effects are largely
unknown, although several offline functional magnetic reson-
ance imaging (fMRI) studies have shown that tDCS generally
alters activity in regions near the anode, as well as in distant
brain structures that are putatively connected to the region tar-
geted for anodal stimulation (Baudewig et al. 2001; Jang et al.
2009; Stagg et al. 2009b, 2014; Keeser et al. 2011; Polania et al.
2011a, 2011b; Pena-Gomez et al. 2012; Amadi et al. 2013; Weber
et al. 2014; Hunter et al. 2015; Krishnamurthy et al. 2015;
Hilgenstock et al. 2016). Findings from online fMRI studies have
been less clear, but far fewer studies have been conducted to
date (Antal et al. 2011). Moreover, several electrophysiological
studies have recently examined the neural oscillatory changes
induced by online tDCS, although their findings have been
complicated and much work remains to be done. For example,
Marshall et al. (2016) used magnetoencephalography (MEG) to
evaluate the impact of an Oz-Cz tDCS montage on occipital
alpha and gamma oscillations. In their study, they used a vis-
ual annulus stimulus that is known to elicit robust occipital
gamma and alpha oscillations (Hoogenboom et al. 2010;
Muthukumaraswamy and Singh 2013), and recorded MEG
concurrently with both anodal and cathodal occipital tDCS.
Surprisingly, they found no reliable alpha or gamma oscillatory
changes during tDCS, despite many exploratory analyses.
Notably, they did report that cathodal stimulation over occipi-
tal cortices decreased basal alpha and gamma activity, but
these changes were not affected by the visual stimulus
(Marshall et al. 2016). In a closely related study, Hanley et al.
(2016) recorded MEG immediately before, during, and after tDCS
using 2 different montages and a visual grating stimulus that
is also known to elicit robust occipital gamma responses
(Muthukumaraswamy et al. 2009, 2010; Muthukumaraswamy
and Singh 2013; Swettenham et al. 2009; Perry et al. 2013). The 2
montages were used in different blocks and were designed to
target occipital (Oz-Cz) and left motor cortices (C3-Fp2) with
anodal stimulation. Across all trials, participants were
instructed to fixate on the visual stimulus and push a button at
stimulus offset, thus eliciting both visual gamma and motor-
related beta oscillations (Wilson et al. 2010, 2011, 2013, 2014a;
Heinrichs-Graham and Wilson 2015, 2016; Heinrichs-Graham
et al. 2016). As with the study by Marshall et al. (2016), the
results were surprising in that neither occipital gamma nor
motor beta oscillations were modulated by the regionally tar-
geted tDCS montage. In fact, paradoxically, neither tDCS mon-
tage modulated motor-related beta activity, and only the motor
montage altered occipital gamma (Hanley et al. 2016). Of note, 2
other MEG/tDCS studies were recently published, although both
of these studies focused on the feasibility of source imaging
using MEG data that was collected concurrently (online) with
tDCS (Soekader et al. 2013; Garcia-Cossio et al. 2016). Thus, the
impact of tDCS on oscillatory activity in visual cortices remains
to be determined, with the available data suggesting that any
potential online effect is small and/or inconsistent.
In this study, our goals were to further investigate whether
tDCS modulates oscillatory alpha and/or gamma activity during
visual processing, and to probe tDCS-induced changes in basal
activity levels in the same occipital cortices. Like previous stud-
ies (Hanley et al. 2016; Marshall et al. 2016), we used MEG and a
visual stimulation paradigm that is known to elicit robust alpha
and gamma oscillations. However, unlike these studies, we
used a different stimulation montage and recorded MEG sub-
stantially after the termination of tDCS (i.e., offline). Based on
previous literature (Kuo et al. 2013; Filmer et al. 2014; Fertonani
and Miniussi 2016), these differences in the tDCS protocol
should have a major impact on our results. Our central hypoth-
eses were that anodal tDCS over the occipital cortices would
modulate offline basal activity levels locally, increase gamma
but not alpha oscillations in occipital cortices during visual pro-
cessing, and reduce the peak gamma frequency. Observing
such changes in basal levels would be consistent with a prior
study (Marshall et al. 2016), while increases in gamma ampli-
tude with reductions in peak gamma frequency would be
expected if anodal tDCS does indeed decrease local GABA activ-
ity (Filmer et al. 2014; Fertonani and Miniussi 2016). Finally, we
examined the exploratory hypothesis that anodal tDCS to the
occipital cortices would induce oscillatory gamma and alpha
changes in brain regions distant from the site of stimulation.
Materials and Methods
Participant Selection
We studied 35 healthy participants (16 females; 3 left-handed),
all of whom were recruited from the community. The mean
age was 24.23, with a range of 20–32 years old. Exclusionary cri-
teria included any medical illness affecting the CNS (e.g., HIV/
AIDS), neurological or psychiatric disorder, history of head trau-
ma, current substance abuse, and the MEG Laboratory’s stand-
ard exclusion criteria (e.g., ferromagnetic implants). After a full
description of the study was given to participants, written
informed consent was obtained following the guidelines of the
University of Nebraska Medical Center’s Institutional Review
Board, which approved the study protocol.
Once participants were consented, they were randomly
assigned to sham (16 participants, 8 female) and active (19 parti-
cipants, 8 female) stimulation groups following a double-blinded
design. The 2 groups were matched on age (sham: 24.0 years;
active: 24.4 years), sex, handedness, ethnicity, and educational
level. With the exception of stimulation (active or sham), all par-
ticipants completed the same experimental protocol.
Transcranial Direct-Current Stimulation
The 5 × 7 cm anode pad was positioned over midline occipital
cortex near the calcarine fissure, while the 5 × 7 cm cathode
pad was positioned over the right prefrontal cortices near the
area generally referred to as supraorbital in most tDCS studies.
Each tDCS sponge was soaked in saline solution and positioned
on the head using the International 10/20 system (Jasper 1958),
which is commonly employed in electroencephalography (EEG),
fNIRS, and tDCS studies (e.g., Wilson et al. 2014b). First, the
 tDCS Modulates Occipital Gamma Oscillations Wilson et al.
distance from the nasion to the inion was measured and a
mark was made on the scalp at the 50% point, then the dis-
tance between the left and right periauricles was measured
and the midpoint was marked. The intersection of the inion/
nasion plane and the periaruicle plane is, by definition, Cz. In
our experiment, the anode was positioned on the midline and
centered about 12.5% above the inion, which corresponds to
~2.5% superior to Oz. The cathode was centered directly lateral
to Fp2 by ~7.5%, which is one of the most common areas for
cathodal placement (Filmer et al. 2014). Importantly, Okamoto
et al. (2004) and Okamoto and Dan (2005) have developed a
method for transforming the scalp-based International 10–20
coordinate system to Montreal Neurological Institute (MNI)-
based coordinates. Briefly, using data from a large sample of
healthy adults, they developed a probabilistic distribution of
the cortical projection points in MNI space that corresponds
to input coordinates from the International 10–20 system
(Okamoto et al. 2004). Based on their data, coordinates in the
International 10–20 system (i.e., scalp-based) can be estimated
in MNI space with an average standard deviation of 8 mm
(Okamoto et al. 2004), which is almost negligible given the size
of our 5 × 7 cm tDCS sponges. Thus, we computed the coordi-
nates of each of our sponges in the International 10–20 sys-
tem, and then used the transformation methods provided by
Okamoto et al. (2004) to obtain the MNI coordinates that corre-
sponded to these scalp-based locations. These data indicated
that the anode was near the calcarine fissure, while the cath-
ode was over right prefrontal cortices (Fig. 1).
Participants in the active stimulation group underwent
20 min of 2.0 mA DCS, plus ~30 s ramp-up and ramp-down peri-
ods, while passively viewing an animated movie. The sham
group received the same passive visual experience for 20 min,
but no stimulation outside of the ~30 s ramp periods. A Soterix
Medical (New York, NY, USA) tDCS system was used for stimu-
lation. Following active/sham stimulation, participants were
prepared for MEG recording and seated with their head posi-
tioned within the MEG helmet. The overall setup took about
35 min from the stop of stimulation to the initiation of the MEG
session, which was by design given the findings of Kuo et al.
(2013). Briefly, this study found that the level of cortical excit-
ability in the motor cortex peaks about 20 min after the ces-
sation of tDCS, and then slowly dissipates over the next
70–90 min. Of note, Kuo et al. (2013) used a montage targeting
the motor cortex and a different size of cathode electrode, both
of which differ from the current study. Thus, we aligned our
Figure 1. Task design and tDCS montage. (A) Participants were seated in a cus-
tom MEG chair and were instructed to fixate on a red square presented cen-
trally. After 0.5 s, a series of visual gratings appeared around the red square for
0.5 s duration, then the gratings shifted by one bar and stayed on the screen for
an additional 0.5 s. Participants completed about 120 trials. (B) The anode was
positioned near the calcarine fissure and the cathode was over the right
prefrontal.
| 1599
MEG recording session to coincide with their period of maximal
excitability, but acknowledge that there were important differ-
ences in the montage and targeting between the 2 studies.
Furthermore, there is no data supporting or denying that other
brain areas will follow the same excitability time course as the
motor cortex.
MEG Experimental Protocol
Throughout the experiment, participants were seated in a cus-
tom chair within the magnetically shielded room (MSR) with
their head positioned in the helmet-shaped MEG sensor array.
Ambient lighting in the MSR was slightly dimmed, but equal
throughout the study. Participants were instructed to remain
still and fixate on a small square presented centrally on a gray
background. After 0.5 s, a series of vertical, stationary, square-
wave gratings (3 cycles per degree) appeared around the fix-
ation box (Fig. 1). These gratings remained on the screen for
0.5 s duration, then shifted by one bar and remained on the
screen for another 0.5 s before disappearing and leaving only
the fixation box. The interval between the offset of the second
series of gratings and onset of the first series (in the next trial)
randomly varied between 2.2 and 2.6 s, and each participant
completed about 120 trials following active or sham tDCS.
Of note, such square gratings are known to elicit robust gamma
and alpha oscillations in occipital cortices (Muthukumaraswamy
et al. 2009, 2010; Muthukumaraswamy and Singh 2013;
Swettenham et al. 2009; Perry et al. 2013), although the typical
paradigm does not generally include a shift in visual space. We
implemented the stimulus shift in this study to evaluate
whether tDCS modulated the neural response to such a visual
scene change occurring during the occipital gamma response.
MEG and MRI Data Acquisition and Coregistration
All recordings were conducted in a one-layer MSR with active
shielding engaged. With an acquisition bandwidth of 0.1–330 Hz,
neuromagnetic responses were sampled continuously at 1 kHz
using an Elekta system with 306 magnetic sensors, including 204
planar gradiometers and 102 magnetometers (Elekta, Helsinki,
Finland). MEG data from each participant were individually cor-
rected for head motion and subjected to noise reduction using
the signal-space separation method with a temporal extension
(tSSS; Taulu et al. 2005; Taulu and Simola 2006).
Prior to MEG measurement, 4 coils were attached to the par-
ticipant’s head and the locations of these coils, together with
the 3 fiducial points and scalp surface, were determined with a
3D digitizer (Fastrak 3SF0002, Polhemus Navigator Sciences,
Colchester, VT, USA). Once the participant was positioned for
MEG recording, an electric current with a unique frequency
label (e.g., 322 Hz) was fed to each of the coils. This induced a
measurable magnetic field and allowed each coil to be localized
in reference to the sensors throughout the recording session.
Since coil locations were also known in head coordinates, all
MEG measurements could be transformed into a common
coordinate system. With this coordinate system (including the
scalp surface points), each participant’s MEG data were coregis-
tered with T1-weighted structural MRI (sMRI) prior to source
space analyses using BESA MRI (Version 2.0; BESA GmbH,
Gräfelfing, Germany). The sMRI data were acquired with a
Philips Achieva 3 T X-series scanner using an 8-channel head
coil (TR: 8.09 ms; TE: 3.7 ms; field of view: 240 mm; slice thick-
ness: 1 mm; no gap; in-plane resolution: 1.0 × 1.0 mm). All sMRI
data were aligned parallel to the anterior and posterior
 1600 | Cerebral Cortex, 2018, Vol. 28, No. 5
commissures and transformed into standardized space, along
with the functional images, after beamforming (see below).
MEG Preprocessing, Time–Frequency Transformation,
and Statistics
Cardiac artifacts were removed from the data using signal-
space projection and the projection operator was accounted for
during source reconstruction (Uusitalo and Imoniemi 1997).
Artifact rejection was based on a fixed threshold method, sup-
plemented with visual inspection. Epochs were of 1.8 s duration
(−0.4 to 1.4 s), with 0.0 s defined as the onset of the gratings and
the baseline being the −0.4 to −0.1 s window.
Artifact-free epochs were transformed into the time–frequency
domain using complex demodulation (resolution: 2.0 Hz, 0.025 s),
and the resulting spectral amplitude estimations per sensor
were averaged over trials to generate time–frequency plots of
mean spectral density. These data were normalized by dividing
the amplitude value of each time–frequency bin by the respect-
ive bin’s baseline amplitude, which was calculated as the mean
amplitude during the −0.4 to −0.1 s window. The resulting quo-
tient was then multiplied by 100%.
The specific time–frequency windows used for imaging were
determined by statistical analysis of the spectrograms corre-
sponding to each of the 204 gradiometer-type sensors across all
participants. Each data point in the spectrogram was initially
evaluated using a mass univariate approach based on the gen-
eral linear model. To reduce the risk of false positive results
while maintaining reasonable sensitivity, a 2 stage procedure
was followed to control for Type 1 error. In the first stage, one-
sample t-tests were conducted on each data point and the out-
put spectrogram of t-values was thresholded at P < 0.05 to
define time–frequency bins containing potentially significant
oscillatory deviations across all participants (sham + active). In
stage 2, time–frequency bins that survived the threshold were
clustered with temporally and/or spectrally adjacent bins that
were also above the (P < 0.05) threshold, and a cluster value
was derived by summing all of the t-values of all data points in
the cluster. Nonparametric permutation testing was then used
to derive a distribution of cluster values and the significance
level of the observed clusters (from stage 1) was tested directly
using this distribution (Ernst 2004; Maris and Oostenveld 2007).
For each comparison, at least 10,000 permutations were com-
puted to build a distribution of cluster values. Based on these
analyses, the time–frequency windows that contained signifi-
cant oscillatory events across all participants were derived and
subjected to the beamforming analysis. We defined the precise
time–frequency parameters using the sensor with the highest
t-value, but the results would have been identical had we used
any of the gradiometers near the peak sensor.
MEG Imaging, Voxel Time Series Extraction, and
Statistics
Using the time–frequency windows determined by the analysis
described above, cortical networks were imaged through an
extension of the linearly constrained minimum variance vector
beamformer (Van Veen et al. 1997; Gross et al. 2001; Hillebrand
et al. 2005), which employs spatial filters in the frequency
domain to calculate source power for the entire brain volume.
The single images were derived from the cross spectral dens-
ities of all combinations of MEG sensors averaged over the
time–frequency range of interest, and the solution of the for-
ward problem for each location on a grid specified by input
voxel space. Following convention, the source power in these
images was normalized per participant using a separately aver-
aged prestimulus noise period of equal duration and bandwidth
(Hillebrand et al. 2005). In principle, the beamformer operator
generates a spatial filter for each grid point, which passes sig-
nals without attenuation from the given neural region while
minimizing interference from activity in all other brain areas.
The properties of these filters are determined from the MEG
covariance matrix and the forward solution for each grid point
in the image space, which are used to allocate sensitivity
weights to each sensor in the array for each voxel in the brain.
MEG preprocessing and imaging used Brain Electrical Source
Analysis software (BESA version 6.1; BESA GmbH, Gräfelfing,
Germany).
Normalized source power was computed for the selected
frequency bands over the entire brain volume per participant at
4.0 × 4.0 × 4.0 mm resolution. Prior to statistical analysis, each
participant’s functional images, which were coregistered to
native space anatomical MRI prior to beamforming, were trans-
formed into standardized space using the transform previously
applied to the structural MRI volume and spatially resampled.
The resulting 3D maps of brain activity were averaged across
all participants to assess the neuroanatomical basis of signifi-
cant oscillatory responses identified through the sensor-level
analysis, and to allow identification of the peak voxels per
oscillatory response.
Voxel time series data (“virtual sensors”) corresponding to
the peak voxels in the averaged gamma and alpha images (see
below) were extracted from the left and right occipital cortices.
Briefly, we averaged the gamma images across all participants
and identified the peak voxel in the left and right occipital
region. The time series corresponding to each of these voxels
were then extracted from the participant’s data individually.
The same procedure was followed for the alpha images. To
compute the voxel time series, we applied the sensor weighting
matrix derived through the forward computation to the prepro-
cessed signal vector, which yielded a time series for the specific
coordinate in source space. Note that this virtual sensor extrac-
tion was done per participant, once the coordinates of interest
(i.e., one per cluster) were known. Once these time series were
extracted, absolute and relative activity values were computed
and subjected to statistical analyses, which consisted of two
(relative and absolute) 2 × 2 × 2 mixed-model ANOVAs. In each
ANOVA, frequency range (alpha and gamma) and hemisphere
(left and right) were within subjects factors, and group (sham and
active stimulation) was a between-subjects factor. Significant
effects were followed up with two-tailed independent-sample
t-tests. Finally, an exploratory, whole-brain analysis was con-
ducted using a mass univariate approach based on the general
linear model to identify tDCS-induced alterations. Since this ana-
lysis was exploratory, the resulting group difference maps were
thresholded at P < 0.01, uncorrected.
Results
All participants completed the task successfully. Two sham and
3 active participants were excluded from all statistical analyses
due to excessive artifacts in their MEG data, no discernable gam-
ma response to the visual gratings, or for exhibiting response
amplitudes >2.5 standard deviation (SD) above/below the mean
for any neural event of interest. There were no statistical differ-
ences in age (P = 0.90) or sex for the remaining 14 sham (mean
age: 24.07 years; 7 females) and 16 active (mean age: 23.94 years;
6 females) stimulation participants. Following artifact rejection,
 tDCS Modulates Occipital Gamma Oscillations Wilson et al.
the average number of trials used in the analysis was 101.88
(SD: 6.08) for the active and 102.36 (SD: 5.02) for the sham group.
This difference was not significant, t(28) = 0.24, P = 0.82.
MEG Sensor-Based Time–Frequency Analyses
Sensor-level time–frequency spectrograms indicated the typical
response pattern of increased gamma and decreased alpha
activity shortly after stimulus onset in all participants. These
spectrograms were statistically examined using one-sample t-
tests of the sensor-level plots from all participants to derive the
precise time–frequency bins for subsequent beamforming ana-
lyses. Significant gamma increases were found in many occipi-
tal sensors in the 40–56 Hz range from about 0.025–1.225 s after
the onset of the first series of gratings (P < 0.001, corrected;
Fig. 2). In addition, there was a significant alpha decrease in the
10–16 Hz range in a large number of sensors that extended
from about 0.175 to 0.425 and 0.675 to 0.925 s after the onset of
the first series of gratings (P < 0.001, corrected; Fig. 2). In order
to identify the cortical dynamics of these responses, we imaged
a 0.3 s window surrounding the peak gamma response
(0.075–0.375 s) and a 0.25 s window that included the peak alpha
response (0.175–0.425 s), and extracted virtual sensor data from
the peak voxels in each frequency range. Note that the gamma
window for beamforming was limited to 0.3 s duration because
this was the maximum duration allowable given the duration
of our baseline.
MEG Voxel-Based Analyses
The square-wave gratings elicited a sharp increase or event-
related synchronization (ERS) in the 40–56 Hz gamma range,
Figure 2. Sensor-level time–frequency analyses. In (A), a group-averaged spectro-
gram from an occipital MEG sensor is shown with time (s) denoted on the x-axis
(0.0 s corresponds to stimulus onset) and frequency (Hz) on the y-axis. All signal
amplitude data are expressed as a percent difference from baseline, with the col-
or legend shown to the far right. The baseline was defined as −0.4 to −0.1 s
before stimulus onset. Data represent a grand averaged (active + sham) gradiom-
eter sensor near posterior and inferior occipital cortices (the same sensor was
selected in each participant). As can be discerned, there was a strong gamma
synchronization (ERS) that emerged shortly after stimulus onset (~0.025 s) and
was sustained until after 1.2 s, as well as a strong alpha desynchronization (ERD)
that began slightly later (~0.175 s) and extended until after 1.2 s. Note that the
clear decrease in alpha ERD activity between 0.55 and 0.65 s reflects the shift in
the visual gratings stimulus. The dotted white line indicates stimulus onset. In
(B), 2D maps of the sensor array are shown to illustrate the sensors where sig-
nificant gamma (top) and alpha (bottom) responses were detected. The sensors
marked for gamma had significant increases in 40–56 Hz activity from 0.025 to
1.225 s, whereas those for alpha had significant 10–16 Hz activity from 0.175 to
0.425 and 0.675 to 0.925 s (P < 0.001, corrected). Each square in the array corre-
sponds to 2 orthogonal gradiometers with both having significant activity.
| 1601
and a sharp decrease or event-related desynchronization (ERD)
in the 10–16 Hz alpha range in bilateral occipital cortices.
Images from each band were separately averaged across all
participants and the peak voxels in the left and right occipital
cortices were identified for gamma (Fig. 3A; peak MNI coordi-
nates: −18, −96, 1, and 18, −96, 9) and alpha (Fig. 3B; peak MNI
coordinates: −13, −80, −3, and 10, −80, −3) activity. The absolute
and relative time series data corresponding to these voxels
were then extracted. For clarity, “relative” responses are typic-
ally referred to as oscillatory neural responses and are com-
puted by dividing the amplitude of each poststimulus data
point by the mean baseline amplitude (−0.4 to −0.1 s), and then
multiplying the resulting quotient by 100%. Absolute responses
are simply the amplitude of the current, irrespective of the
baseline, and allow activity levels during the baseline to be
quantified. Both types of responses reflect the same underlying
neural data; the baseline normalization is the only difference.
The relative voxel time series data indicated strong gamma
ERS and alpha ERD responses in bilateral occipital regions
(Fig. 4). To identify tDCS-induced alterations, we first com-
puted the mean amplitude of gamma ERS and alpha ERD activ-
ity during visual processing. Then, a 2 × 2 × 2 mixed-model
ANOVA was conducted on these relative mean amplitude
values, with frequency (alpha/gamma), hemisphere (right/left),
and group (sham/active) as factors. This model indicated a sig-
nificant three-way interaction F(1, 28) = 5.14 (P < 0.05), and
main effects of hemisphere F(1, 28) = 7.25 (P < 0.05) and fre-
quency F(1, 28) = 200.49 (P < 0.05). The group effect and the
two-way interactions were not significant. Follow-up t-tests of
the three-way interaction effect indicated significantly stron-
ger gamma activity in the right occipital cortices of the active
relative to the sham stimulation group, t(28) = 2.1 (P < 0.05).
There was also a trend toward the same finding for gamma
activity in the left hemisphere. Relative alpha activity did not
statistically differ between groups.
A similar mixed-model ANOVA was conducted on the
mean absolute amplitude values from the baseline period in
the same voxels as were used for the relative analysis detailed
above. These analyses indicated a significant two-way group-
by-frequency interaction F(1, 28) = 5.60 (P < 0.05), and main
effects of group F(1, 28) = 5.15 (P < 0.05) and frequency F(1, 28) =
202.55 (P < 0.05). No other main effects or interactions were sig-
nificant. Follow-up t-tests indicated that the absolute alpha
amplitude during the baseline period was significantly stronger
in the active relative to the sham group in the left, t(28) = 2.24
(P < 0.05), and right occipital cortices, t(28) = 2.51 (P < 0.05;
Fig. 5). There were no statistical differences in absolute gamma
activity.
Of note, we evaluated whether the input data to the para-
metric statistical tests described above (i.e., the mean absolute
amplitude during the baseline and the relative amplitude dur-
ing visual processing for both alpha and gamma) were normally
distributed using Shapiro–Wilk tests of Normality. These tests
indicated that 2 variables violated the normality assumption.
Thus, we applied a constant to all data to eliminate negative
numbers, then performed a log transform and reran all statis-
tical tests. These tests yielded the exact same findings, with
slightly smaller P values on average, and no new findings.
Given this, we have used the real values in all statistical tests
and figures reported in this paper, as they are much more
straightforward to interpret.
We also computed the peak gamma frequency for each par-
ticipant using the voxel time series data. Contrary to our hypoth-
eses, these analyses showed that tDCS did not significantly
 1602 | Cerebral Cortex, 2018, Vol. 28, No. 5

Figure 3. Group mean beamformer images (pseudo-t) of the gamma synchronization (ERS) and alpha desynchronization (ERD) collapsed across all participants. In (A),
the gamma ERS was imaged in each participant using 0.3 s window surrounding the peak gamma response (0.075–0.375 s). The output images were then averaged
across all participants and are shown in pseudo-t units (scale is below coronal slice). The same procedure was followed for (B), except that the alpha ERD was imaged
in each participant using a 0.25 s window that included the peak alpha response (0.175–0.425 s). The output images were again averaged and the scale is shown in
blue in the bottom right corner. These average images were computed to visualize the brain areas generating each oscillatory response that was detected in MEG sen-
sor space, and to enable identification of a peak voxel per response in each hemisphere. Time series data were extracted from these peak voxels and used to assess
tDCS-induced alterations in local neuronal activity.

Figure 4. Using the images shown in Figure 3, voxel time series were extracted from the peak voxels of each hemisphere for the gamma ERS and alpha ERD responses
in each participant. These time series were then normalized using the prestimulus baseline of each participant and plotted separately for each frequency bin and
hemisphere. In each plot, the active stimulation group is shown in red and the sham appears in blue. Time (in s, stimulus onset = 0.0 s) is denoted on the x-axis, while
relative amplitude (%) is shown on the y-axis. As shown, gamma oscillatory activity (ERS) was significantly stronger during visual processing (0.025–1.225 s) in the
active tDCS group relative to the sham group (top panel; P < 0.05), whereas anodal occipital stimulation appeared to have no effect on the local alpha ERD response.
The shaded area around each line denotes the SEM.
 tDCS Modulates Occipital Gamma Oscillations Wilson et al. | 1603

Figure 5. Absolute voxel time series were extracted from the peak voxels of the left and right gamma ERS responses (top panel) and alpha ERD responses (bottom
panel) to identify basal shifts in neural activity in each frequency range. In each plot, the absolute amplitude time series is shown for the active (red line) and sham
groups (blue line). Time (in s, stimulus onset = 0.0 s) is denoted on the x-axis, while amplitude (nAm) is shown on the y-axis. Note that these responses are from the
same voxels as Figure 4 and show the same data, except that these data have not been normalized using the baseline period. As shown, active anodal tDCS did not
significantly affect basal gamma activity in occipital cortices (top), but did result in significantly elevated basal alpha activity during the prestimulus baseline, and
this largely held throughout the time course (bottom panel). The shaded area around each line denotes the SEM.

modulate peak gamma frequency (P = 0.28). To further evaluate
the relationship between peak gamma frequency and amplitude,
we conducted a canonical correlation using the amplitude and
peak frequency of each hemisphere. Briefly, a canonical correl-
ation (Rc) is a multivariate method that can be used to determine
the strength of a relationship between 2 sets of variables. First, a
composite (latent) variable is computed for each set of vari-
ables based on a linear combination of each variable in the set.
This process is repeated on the residual variance until there
are as many orthogonal latent variables per set as the number
of variables in the smaller of the 2 sets (Sherry and Hensen
2005). In this case, we computed the Rc between (left gamma
power and right gamma power) and (left gamma peak fre-
quency and right gamma peak frequency). We found a signifi-
cant correlation between gamma frequency and gamma power,
Rc = 0.58, Wilks λ = 0.66, F(4, 52) = 2.98, P = 0.027, that accounted
for 33.5% of the variance. Standardized canonical correlation
coefficients for this model were as follows: left power: −0.54,
right power: −0.55; left frequency: 0.85, right frequency: 0.40.
Negative power correlation coefficients coupled with positive
frequency correlation coefficients indicates a negative relation-
ship between the 2 sets of variables. The model of the residuals
was not significant, P = 0.713.
Lastly, we conducted exploratory whole-brain statistical
analyses using a mass univariate approach to identify possible
tDCS-induced alterations distant from the anodal stimulation
site. Beyond occipital cortices, these analyses revealed signifi-
cantly stronger gamma ERS responses in the left frontal eye
fields, left prefrontal cortices, and bilateral superior parietal
cortices in the active compared with the sham stimulation
group (P < 0.01, uncorrected; Fig. 6). In contrast, the sham group
had stronger gamma ERS in the right inferior parietal, as well
as stronger alpha ERD responses in the right cerebellum and
right anterior superior parietal compared with the active stimu-
lation group (P < 0.01, uncorrected; Fig. 6).
Discussion
In this study, we examined whether tDCS modulates oscillatory
alpha and gamma activity during offline visual processing.
Participants in the active group watched a 20 min animated
movie while anodal tDCS was applied to the occipital cortices
and cathodal was applied to the right prefrontal cortex. Sham
participants watched the same movie with the same tDCS elec-
trode configuration, but did not receive real stimulation. About
35 min after the movie and stimulation, all participants
 1604 | Cerebral Cortex, 2018, Vol. 28, No. 5
Figure 6. Exploratory whole-brain analysis. Independent-samples t-tests were
conducted on the voxel-level beamformer data for the gamma ERS and alpha
ERD responses. Beyond differences in occipital cortices, these analyses revealed
stronger gamma ERS responses in the left frontal eye fields (top), left prefrontal
cortices (second panel), and bilateral superior parietal areas (not shown) in
the active relative to the sham stimulation group (P < 0.01, uncorrected).
Conversely, the sham group had stronger gamma ERS in the right inferior par-
ietal area (third panel), as well as stronger alpha ERD activity in the right cere-
bellum and the right anterior superior parietal cortices relative to the active
stimulation group (bottom panel). All maps have been thresholded at (P < 0.01,
uncorrected) and are shown in neurological convention. The color legend
appears at the bottom.
underwent a MEG session during which they viewed a gratings
stimulus that was known to elicit robust gamma and alpha
oscillations. Our primary results indicated that tDCS significantly
increased gamma ERS responses in the right occipital cortices
(marginally in the left). Interestingly, anodal tDCS also shifted
basal alpha levels significantly higher during the baseline period
in bilateral occipital cortices. Finally, we found preliminary evi-
dence that tDCS modulates oscillatory activity in brain regions
distant from the electrodes. Regions where gamma activity was
most strongly modulated included those known to be involved
in visual attention circuits, such as prefrontal and superior par-
ietal cortices, as well as the frontal eye fields. tDCS also modu-
lated alpha oscillations in the cerebellum and lateral parietal
cortices. Below, we discuss the implications of these findings for
understanding the impact of tDCS on neural oscillatory activity
during offline visual processing, and the underlying mechan-
isms that may be involved.
One of our most important findings was that anodal tDCS
increased occipital gamma activity (ERS) during visual process-
ing. Demonstrating a connection between anodal stimulation
and changes in local gamma activity is of critical interest due
to the purported link between GABAergic activity and gamma
oscillations. Essentially, anodal stimulation is thought to
decrease GABAergic activity based on human pharmacological
studies (Nitsche et al. 2004), TMS work (Di Lazzaro et al. 2005;
Nitsche et al. 2005), and MRS (Stagg et al. 2009a, 2011; Kim et al.
2014; Bachtiar et al. 2015), and numerous cellular electrophysi-
ology studies have shown that cortical gamma activity is crit-
ically dependent upon the integrity of local interneuronal
networks, which function as GABA-gated pacemakers for neo-
cortical oscillatory activity (Singer 1999; Bartos et al. 2007; Fries
et al. 2007; Fries 2009, 2015; Uhlhaas et al. 2009; Buzsaki and
Wang 2012; Uhlhaas and Singer 2012; Vinck et al. 2013; Salkoff
et al. 2015). Thus, following this line of reasoning, anodal tDCS
should modulate local gamma activity and thereby provide a
powerful and exciting new tool to probe the relationship
between GABAergic activity and gamma oscillations. Several
studies have already connected MEG gamma oscillations in
humans with local GABA levels using MRS. The most com-
mon finding is likely a positive correlation between peak
gamma frequency and local GABA levels (Edden et al. 2009;
Muthukumaraswamy et al. 2009; Gaetz et al. 2011), but this
finding has not been entirely consistent (Hall et al. 2011;
Muthukumaraswamy et al. 2013). None of these studies found
a relationship between gamma amplitude/power and GABA
levels. Assuming that anodal tDCS decreases local GABA activ-
ity (Stagg et al. 2009a, 2011; Coffman et al. 2014; Filmer et al.
2014; Kim et al. 2014; Bachtiar et al. 2015; Fertonani and
Miniussi 2016), these data would appear to be inconsistent with
our main findings of increased occipital gamma amplitude,
with no change in peak gamma frequency, following anodal
tDCS. However, in agreement with our canonical correlation
findings, more recent studies have suggested that peak gamma
frequency and gamma amplitude may have an inverse rela-
tionship with each other (Campbell et al. 2014; Lozano-
Soldevilla et al. 2014; Kujala et al. 2015). In fact, a multimodal
Flumazenil-PET and MEG study found that GABAA receptor
density in the primary visual cortex was positively correlated
with peak gamma frequency and negatively correlated with
gamma power during a visual task (Kujala et al. 2015). In other
words, stronger gamma power was associated with reduced
GABAA receptor density, which is more closely related to
ongoing inhibitory neurotransmission than total GABA concen-
tration as measured with MRS, and thus a more direct assess-
ment of the relationship between GABA signaling and gamma
oscillations. When combined with the current results, the pat-
tern of findings in the Kujala et al. (2015) study support the
 tDCS Modulates Occipital Gamma Oscillations Wilson et al.
notion that anodal tDCS decreases local GABA activity and that
this is associated with an increase in gamma amplitude in the
target brain region. Their data also support the inverse relation-
ship between peak gamma frequency and gamma amplitude
that we observed (Kujala et al. 2015). Although much remains
to be discovered, and the current literature has many discrep-
ancies, determining the inherent connection between gamma
amplitude/power, peak gamma frequency, and local GABAergic
activity is of major importance, and our data suggests that
tDCS could play a significant role in this effort.
Beyond the current investigation, 2 other tDCS/MEG studies
have recently examined whether anodal stimulation modulates
visual gamma activity using very similar stimuli (Hanley et al.
2016; Marshall et al. 2016). In contrast to our findings, neither of
these studies found that anodal tDCS of the occipital cortices
significantly altered local gamma activity, although there are
several possible explanations for this discrepancy. For one,
both of these studies used an Oz-Cz montage where the 2 elec-
trodes were significantly more proximal than the current inves-
tigation. Such configurations are generally associated with less
current reaching the brain, as shunting through the skin is
more problematic when the electrodes are proximal (Rush and
Driscoll 1968; Datta et al. 2008; Bikson et al. 2010; Moliadze et al.
2010). Secondly, the amount and duration of current signifi-
cantly differed between our study and the Hanley et al. (2016;
1 mA for 600 s) and Marshall et al. (2016; multiple blocks of
2 mA for 120 s, which alternated with cathodal stimulation of
the same amplitude and duration) investigations. We used
2 mA for 20 min, which is obviously very different from these
2 prior studies. Thirdly, we had participants watch a movie out-
side of the MEG room during the stimulation, whereas Marshall
et al. (2016) used concurrent tDCS/MEG and Hanley and collea-
gues recorded MEG before, during, and 5 min after tDCS. Thus,
in both of these studies, tDCS was performed as the partici-
pants completed the target task. Whether the context of
stimulation delivery modulates the effect is not currently
understood, but intuitively it makes sense that the process-
ing state of the brain regions being stimulated should matter.
Finally, and perhaps most importantly, we acquired MEG data
in the current study about 35 min after the termination of tDCS
(i.e., offline), which was design given the findings of Kuo et al.
(2013). This TMS study found that the level of cortical excitabil-
ity in the motor cortex peaked about 20 min after the cessation
of tDCS, and then slowly dissipated over the next 70–90 min
(Kuo et al. 2013). If their findings extend to the occipital cortex,
then the key difference between the current work and previous
tDCS/MEG investigations may be that the MEG recording
occurred substantially “after” the termination of tDCS. Of note,
Kuo et al. (2013) focused on the motor cortex and there is cur-
rently no evidence supporting or denying that other brain
regions (e.g., occipital) will have the same cortical excitability
time course following tDCS. However, in the absence of data to
the contrary, using the findings of Kuo et al. (2013) as a tem-
poral model more generally would appear to be beneficial to
using no model. Additional studies to tease apart the possible
effects of stimulation context, amplitude and duration, mon-
tage, and the window of time between stimulation and record-
ing are direly needed. Lastly, note that we focused our
discussion on only the most relevant condition of the Hanley
et al. (2016) study, which included multiple other conditions
and a different montage that targeted the motor cortices.
Another major finding of the current study was the basal
amplitude shift in occipital alpha following stimulation. Briefly,
active tDCS induced a significant increase in baseline alpha
| 1605
activity in both left and right occipital cortices. This finding was
expected and is partially consistent with data from Marshall
et al. (2016), as they found reduced overall alpha in the occipital
cortices during cathodal stimulation relative to anodal and
sham stimulation. For clarity, they did not find elevated occipi-
tal alpha during anodal stimulation, and their overall decrease
during cathodal stimulation was across the whole time series
and not just the baseline (Marshall et al. 2016), as we are report-
ing; although Figure 5 of the current paper clearly indicates
that overall alpha was also elevated. Interestingly, this increase
in basal alpha in the active stimulation group was of similar
magnitude in the left and right hemisphere and did not appear
to impact the amplitude of the alpha ERD response, as these
responses were remarkably similar between groups. As for the
mechanism, the link between GABA activity and alpha oscilla-
tions is less well established than that for gamma, but several
studies have reported alpha modulation and/or comodulation
of alpha and gamma in the context of GABA altering drugs
(Campbell et al. 2014; Lozano-Soldevilla et al. 2014). However, in
the current study, tDCS affected alpha and gamma very differ-
ently, as only basal alpha (not gamma) and gamma oscillations
(not alpha) were affected. This may suggest that the mechan-
isms are largely distinct, and/or that alterations in one fre-
quency range are a secondary consequence of changes in the
other. For example, decreased GABA may result in increased
gamma activity, and this increase in local gamma could alter
the circuit dynamics, and thereby change the level of basal
activity in the alpha range. Of course, this is speculative and
future studies should probe how these parameters interact in
the context of tDCS and GABA altering drugs.
Finally, we also observed network-level tDCS effects in both
the alpha and gamma range in an exploratory whole-brain ana-
lysis. First, outside of the occipital cortices, we observed signifi-
cantly stronger gamma ERS responses in the superior parietal
bilaterally, left prefrontal cortices, and the left frontal eye fields
of the active tDCS group relative to the sham group.
Interestingly, these regions are known to be involved in visual
attention and to be strongly interconnected (Petersen and
Posner 2012). In this regard, we were surprised by the absence
of differences in the right prefrontal cortices, although the
cathode was located near this region and that may have con-
tributed. A previous tDCS/EEG study also showed changes in
high gamma, but these changes were in local connectivity and
more proximal to the anode (Polania et al. 2011a, 2011b). As for
alpha activity, the active stimulation group had significantly
reduced alpha ERD in the right cerebellum and right anterior
parietal region. The significance of these regions is less clear
and future studies should more closely evaluate network-level
changes following tDCS.
In conclusion, this was the first study to show that anodal
tDCS applied to the occipital cortices modulates visual gamma
oscillations. Anodal tDCS is thought to decrease local GABA,
and our finding of altered gamma activity supports this given
the purported connection between gamma oscillations and
GABAergic activity. Future studies should confirm and extend
this link, as the capacity to modulate GABA activity in a tar-
geted and noninvasive way could lead to significant new
insights. Notably, 2 previous studies used tDCS, MEG, and simi-
lar stimuli and found negative results in regard to gamma
modulation. We propose that these discrepancies are likely
attributable to the amount of time that elapsed between the
termination of tDCS and the MEG recording, differences in
stimulation amplitude and duration, and perhaps other factors
related to the stimulation. Future studies should probe how
 1606 | Cerebral Cortex, 2018, Vol. 28, No. 5
each of these parameters affects the underlying physiology
using MEG and other imaging methods. Beyond gamma, we
found that tDCS elevated basal alpha activity in occipital corti-
ces, which is in partial agreement with other studies (Marshall
et al. 2016), although future work is certainly needed. Lastly, we
also showed that the peak gamma frequency is inversely
related to gamma amplitude in the same voxel, which agrees
with some studies (Campbell et al. 2014; Lozano-Soldevilla
et al. 2014; Kujala et al. 2015), but not others (Edden et al. 2009;
Muthukumaraswamy et al. 2009; Gaetz et al. 2011) in the
ongoing debate about the relationship amongst these variables.
Potentially, future tDCS studies can contribute to sorting out
this relationship. Before closing, it is important to note a few
limitations of the study. First, unlike some tDCS studies, we did
not use a repeated-measures crossover design for administer-
ing the stimulation. Such crossover designs certainly have their
strengths, but are relatively rare in pharmaceutical and related
trials due to the increased attrition and the possibility of carry-
over effects. For these reasons, the current study opted for an
independent-samples approach. A second limitation is that the
current study only included a single stimulation montage,
although this design is very common among studies that use
neuroimaging in conjunction with tDCS, due to the major
expense of repeating the imaging sessions for each montage.
Finally, the current study did not investigate oscillatory activity
in frequency ranges beyond alpha and gamma due to the
nature (i.e., specificity) of the MEG visual task, and future stud-
ies should evaluate how tDCS modulates theta and other
neural rhythms. Despite these limitations, the current study
has provided groundbreaking data on how tDCS modulates
alpha and gamma oscillations in healthy adults, and set the
stage for future studies investigating the link between GABA
and gamma with tDCS.
Funding
National Institutes of Health (NIH) grant R01 MH103220 (TWW),
National Science Foundation (NSF) grant #1539067 (TWW), the
Shoemaker Prize from the University of Nebraska Foundation
(TWW), a Kinman-Oldfield Award for Neurodegenerative Research
from the University of Nebraska Foundation (TWW), and a grant
from the Nebraska Banker’s Association (TWW).
Notes
Conflict of Interest: None declared.
References
Amadi U, Ilie A, Johansen-Berg H, Stagg CJ. 2013. Polarity-
specific effects of motor transcranial direct current stimula-
tion on fMRI resting state networks. Neuroimage. 88:
155–161.
Antal A, Kincses TZ, Nitsche MA, Paulus W. 2003. Modulation of
moving phosphine thresholds by transcranial direct current
stimulation of V1 in human. Neuropsychologia. 41(13):
1802–1807.
Antal A, Nitsche MA, Kincses TZ, Kruse W, Hoffmann KP,
Paulus W. 2004. Facilitation of visuo-motor learning by tran-
scranial direct current stimulation of the motor and extra-
striate visual areas in humans. Eur J Neurosci. 19(10):
2888–2892.
Antal A, Polania R, Schmidt-Samoa C, Dechent P, Paulus W.
2011. Transcranial direct current stimulation over the
primary motor cortex during fMRI. Neuroimage. 55:
590–596.
Bachtiar V, Near J, Johansen-Berg H, Stagg CJ. 2015. Modulation
of GABA and resting state functional connectivity by tran-
scranial direct current stimulation. eLife. 4:e08789.
Barbieri M, Negrini M, Nitsche MA, Rivolta D. 2016. Anodal-tDCS
over the human right occipital cortex enhances the perception
and memory of both faces and objects. Neuropsychologia. 81:
238–244.
Bartos M, Vida I, Jonas P. 2007. Synaptic mechanisms of syn-
chronized gamma oscillations in inhibitory interneuron net-
works. Nat Rev Neurosci. 8:45–56.
Baudewig J, Nitsche MA, Paulus W, Frahm J. 2001. Regional
modulation of BOLD MRI responses to human sensorimotor
activation by transcranial direct current stimulation. Magn
Reson Med. 45:196–201.
Bikson M, Datta A, Rahman A, Scaturro J. 2010. Electrode mon-
tages for tDCS and weak transcranial electrical stimulation:
role of “return” electrode’s position and size. Clin
Neurophysiol. 121(12):1976–1978.
Bindman LJ, Lippold OC, Redfearn JW. 1964. The action of brief
polarizing currents on the cerebral cortex of the rat (1) dur-
ing current flow and (2) in the production of long-lasting
after-effects. J Physiol. 172:369–382.
Buzsaki G, Wang XJ. 2012. Mechanisms of gamma oscillations.
Annu Rev Neurosci. 35:203–225.
Campbell AE, Sumner P, Singh KD, Muthukumaraswamy SD.
2014. Acute effects of alcohol on stimulus-induced gamma
oscillations in human primary visual and motor cortices.
Neuropsychopharmacology. 39(9):2104–2113.
Coffman BA, Clark VP, Parasuraman R. 2014. Battery powered
thought: enhancement of attention, learning, and memory
in healthy adults using transcranial direct current stimula-
tion. Neuroimage. 85(3):895–908.
Datta A, Elwassif M, Battaglia F, Bikson M. 2008. Transcranial
current stimulation focality using disc and ring electrode
configurations: FEM analysis. J Neural Eng. 5(2):163–174.
Di Lazzaro V, Oliviero A, Saturno E, Dileone M, Pilato F,
Nardone R, Ranieri F, Musumeci G, Fiorilla T, Tonali P. 2005.
Effects of lorazepam on short latency afferent inhibition and
short latency intracortical inhibition in humans. J Physiol.
564(Pt 2):661–668.
Edden RA, Muthukumaraswamy SD, Freeman TC, Singh KD.
2009. Orientation discrimination performance is predicted by
GABA concentration and gamma oscillation frequency in
human primary visual cortex. J Neurosci. 29(50):15721–15726.
Ernst MD. 2004. Permutation methods: a basis for exact infer-
ence. Stat Sci. 19:676–685.
Fertonani A, Miniussi C. 2016. Transcranial electrical stimula-
tion: What we know and do not know about mechanisms.
Neuroscientist. doi: 10.1177/1073858416631966.
Filmer HL, Dux PE, Mattingley JB. 2014. Applications of transcra-
nial direct current stimulation for understanding brain func-
tion. Trends Neurosci. 37(12):742–753. Review.
Fries P. 2009. Neuronal gamma-band synchronization as a fun-
damental process in cortical computation. Annu Rev
Neurosci. 32:209–224.
Fries P. 2015. Rhythms for cognition: communication through
coherence. Neuron. 88:220–235.
Fries P, Nikolic D, Singer W. 2007. The gamma cycle. Trends
Neurosci. 30:309–316.
Gaetz W, Edgar JC, Wang DJ, Roberts TP. 2011. Relating MEG mea-
sured motor cortical oscillations to resting γ-aminobutyric
acid (GABA) concentration. Neuroimage. 55(2):616–621.
 tDCS Modulates Occipital Gamma Oscillations Wilson et al.
Garcia-Cossio E, Witkowski M, Robinson SE, Cohen LG,
Birbaumer N, Soekadar SR. 2016. Simultaneous transcranial
direct current stimulation (tDCS) and whole-head magne-
toencephalography (MEG): assessing the impact of tDCS on
slow cortical magnetic fields. Neuroimage. 140:33–40.
Gross J, Kujala J, Hamalainen M, Timmermann L, Schnitzler A,
Salmelin R. 2001. Dynamic imaging of coherent sources:
Studying neural interactions in the human brain. Proc Natl
Acad Sci USA. 98:694–699.
Hall SD, Stanford IM, Yamawaki N, McAllister CJ, Rönnqvist KC,
Woodhall GL, Furlong PL. 2011. The role of GABAergic modu-
lation in motor function related neuronal network activity.
Neuroimage. 56(3):1506–1510.
Hanley CJ, Singh KD, McGonigle DJ. 2016. Transcranial modula-
tion of brain oscillatory responses: a concurrent tDCS-MEG
investigation. Neuroimage. 140:20–32.
Heinrichs-Graham E, Arpin DJ, Wilson TW. 2016. Cue-related
temporal factors modulate movement-related beta oscilla-
tory activity in the human motor circuit. J Cogn Neurosci.
28:1039–1051.
Heinrichs-Graham E, Wilson TW. 2015. Coding complexity in
the human motor circuit. Hum Brain Mapp. 36(12):
5155–5167.
Heinrichs-Graham E, Wilson TW. 2016. Is an absolute level of
cortical beta suppression required for proper movement?
Magnetoencephalographic evidence from healthy aging.
Neuroimage. 134:514–521.
Hilgenstock R, Weiss T, Huonker R, Witte OW. 2016.
Behavioural and neurofunctional impact of transcranial dir-
ect current stimulation on somatosensory learning. Hum
Brain Mapp. 37:1277–1295.
Hillebrand A, Singh KD, Holliday IE, Furlong PL, Barnes GR.
2005. A new approach to neuroimaging with magnetoence-
phalography. Hum Brain Mapp. 25:199–211.
Hoogenboom N, Schoffelen JM, Oostenveld R, Fries P. 2010.
Visually induced gamma-band activity predicts speed of
change detection in humans. Neuroimage. 51(3):1162–1167.
Hunter MA, Coffman BA, Gasparovic C, Calhoun VD, Trumbo
MC, Clark VP. 2015. Baseline effects of transcranial direct cur-
rent stimulation on glutamatergic neurotransmission and
large-scale network connectivity. Brain Res. 1594:92–107.
Jackson MP, Rahman A, Lafon B, Kronberg G, Ling D, Parra LC,
Bikson M. 2016. Animal models of transcranial direct current
stimulation: methods and mechanisms. Clin Neurophysiol.
127(11):3425–3454.
Jang SH, Ahn SH, Byun WM, Kim CS, Lee MY, Kwon YH. 2009.
The effect of transcranial direct current stimulation on the
cortical activation by motor task in the human brain: an
fMRI study. Neurosci Lett. 460:117–120.
Jasper HH. 1958. The ten-twenty electrode system of the Inter-
national Federation. Electroencephalogr Clin Neurophysiol. 10:
371–375.
Keeser D, Meindl T, Bor J, Palm U, Pogarell O, Mulert C, Brunelin J,
Moller HJ, Reiser M, Padberg F. 2011. Prefrontal transcranial
direct current stimulation changes connectivity of resting-
state networks during fMRI. J Neurosci. 31:15284–15293.
Kim S, Stephenson MC, Morris PG, Jackson SR. 2014. tDCS-
induced alterations in GABA concentration within primary
motor cortex predict motor learning and motor memory: a
7 T magnetic resonance spectroscopy study. Neuroimage.
99:237–243.
Kraft A, Roehmel J, Olma MC, Schmidt S, Irlbacher K, Brandt SA.
2010. Transcranial direct current stimulation affects visual
| 1607
perception measured by threshold perimetry. Exp Brain Res.
207(3–4):283–290.
Krishnamurthy V, Gopinath K, Brown GS, Hampstead BM. 2015.
Resting-state fMRI reveals enhanced functional connectivity
in spatial navigation networks after transcranial direct cur-
rent stimulation. Neurosci Lett. 604:80–85.
Kujala J, Jung J, Bouvard S, Lecaignard F, Lothe A, Bouet R,
Ciumas C, Ryvlin P, Jerbi K. 2015. Gamma oscillations in V1
are correlated with GABA(A) receptor density: a multi-modal
MEG and Flumazenil-PET study. Sci Rep. 5:16347.
Kuo HI, Bikson M, Datta A, Minhas P, Paulus W, Kuo MF,
Nitsche MA. 2013. Comparing cortical plasticity induced by
conventional and high-definition 4 × 1 ring tDCS: a neuro-
physiological study. Brain Stimul. 6(4):644–648.
Lafon B, Rahman A, Bikson M, Parra LC. 2017. Direct current
stimulation alters neuronal input/output function. Brain
Stimul. 10(1):36–45.
Lang N, Siebner HR, Chadaide Z, Boros K, Nitsche MA, Rothwell
JC, Paulus W, Antal A. 2007. Bidirectional modulation of pri-
mary visual cortex excitability: a combined tDCS and rTMS
study. Invest Ophthalmol Vis Sci. 48(12):5782–5787.
Liebetanz D, Nitsche MA, Tergau F, Paulus W. 2002.
Pharmacological approach to the mechanisms of transcra-
nial DC stimulation-induced after-effects of human motor
cortex excitability. Brain. 125:2238–2247.
Lozano-Soldevilla D, ter Huurne N, Cools R, Jensen O. 2014.
GABAergic modulation of visual gamma and alpha oscilla-
tions and its consequences for working memory perform-
ance. Curr Biol. 24(24):2878–2887.
Maris E, Oostenveld R. 2007. Nonparametric statistical testing
of EEG- and MEG-data. J Neurosci Method. 164:177–190.
Marshall TR, Esterer S, Herring JD, Bergmann TO, Jensen O.
2016. On the relationship between cortical excitability and
visual oscillatory responses – a concurrent tDCS-MEG study.
Neuroimage. 140:41–49.
Medeiros LF, de Souza IC, Vidor LP, de Souza A, Deitos A, Volz
MS, Fregni F, Caumo W, Torres IL. 2012. Neurobiological
effects of transcranial direct current stimulation: a review.
Front Psychiatry. 3:110.
Moliadze V, Antal A, Paulus W. 2010. Electrode-distance
dependent after-effects of transcranial direct and random
noise stimulation with extracephalic reference electrodes.
Clin Neurophysiol. 121(12):2165–2171.
Muthukumaraswamy SD, Edden RA, Jones DK, Swettenham JB,
Singh KD. 2009. Resting GABA concentration predicts peak
gamma frequency and fMRI amplitude in response to visual
stimulation in humans. Proc Natl Acad Sci USA. 106(20):
8356–8361.
Muthukumaraswamy SD, Myers JF, Wilson SJ, Nutt DJ, Lingford-
Hughes A, Singh KD, Hamandi K. 2013. The effects of elevated
endogenous GABA levels on movement-related network
oscillations. Neuroimage. 66:36–41.
Muthukumaraswamy SD, Singh KD. 2013. Visual gamma oscil-
lations: the effects of stimulus type, visual field coverage
and stimulus motion on MEG and EEG recordings.
Neuroimage. 69:223–230.
Muthukumaraswamy SD, Singh KD, Swettenham JB, Jones DK.
2010. Visual gamma oscillations and evoked responses: vari-
ability, repeatability and structural MRI correlates.
Neuroimage. 49(4):3349–3357.
Nitsche MA, Liebetanz D, Schlitterlau A, Henschke U, Fricke K,
Frommann K, Lang N, Henning S, Paulus W, Tergau F. 2004.
GABAergic modulation of DC stimulation-induced motor
 1608 | Cerebral Cortex, 2018, Vol. 28, No. 5
cortex excitability shifts in humans. Eur J Neurosci. 19(10):
2720–2726.
Nitsche MA, Paulus W. 2000. Excitability changes induced in
the human motor cortex by weak transcranial direct current
stimulation. J Physiol. 527:633–639.
Nitsche MA, Paulus W. 2001. Sustained excitability elevations
induced by transcranial DC motor cortex stimulation in
humans. Neurology. 57:1899–1901.
Nitsche MA, Nitsche MS, Klein CC, Tergau F, Rothwell JC,
Paulus W. 2003. Level of action of cathodal DC polarisation
induced inhibition of the human motor cortex. Clin
Neurophysiol. 114:600–604.
Nitsche MA, Seeber A, Frommann K, Klein CC, Rochford C,
Nitsche MS, Fricke K, Liebetanz D, Lang N, Antal A, et al.
2005. Modulating parameters of excitability during and after
transcranial direct current stimulation of the human motor
cortex. J Physiol. 568(Pt 1):291–303.
Okamoto M, Dan I. 2005. Automated cortical projection of head-
surface locations for transcranial functional brain mapping.
Neuroimage. 26:18–28.
Okamoto M, Dan H, Sakamoto K, Takeo K, Shimizu K, Kohno S,
Oda I, Isobe S, Suzuki T, Kohyama K, et al. 2004. Three-
dimensional probabilistic anatomical cranio-cerebral
correlation via the international 10–20 system oriented for
transcranial functional brain mapping. Neuroimage. 21:99–111.
Olma MC, Kraft A, Roehmel J, Irlbacher K, Brandt SA. 2011.
Excitability changes in the visual cortex quantified with sig-
nal detection analysis. Restor Neurol Neurosci. 29(6):
453–461.
Pena-Gomez C, Sala-Lonch R, Junque C, Clemente IC, Vidal D,
Bargallo N, Falcon C, Valls-Sole J, Pascual-Leone A, Bartres-
Faz D. 2012. Modulation of large-scale brain networks by
transcranial direct current stimulation evidenced by
resting-state functional MRI. Brain Stimul. 5:252–263.
Perry G, Hamandi K, Brindley LM, Muthukumaraswamy SD,
Singh KD. 2013. The properties of induced gamma oscilla-
tions in human visual cortex show individual variability
in their dependence on stimulus size. Neuroimage. 68:
83–92.
Petersen SE, Posner MI. 2012. The attention system of the
human brain: 20 years after. Annu Rev Neurosci. 35:73–89.
Polanía R, Nitsche MA, Paulus W. 2011a. Modulating functional
connectivity patterns and topological functional organiza-
tion of the human brain with transcranial direct current
stimulation. Hum Brain Mapp. 32(8):1236–1249.
Polanía R, Paulus W, Antal A, Nitsche MA. 2011b. Introducing
graph theory to track for neuroplastic alterations in the rest-
ing human brain: a transcranial direct current stimulation
study. Neuroimage. 54:2287–2296.
Purpura DP, McMurtry JG. 1965. Intracellular activities and
evoked potential changes during polarization of motor cor-
tex. J Neurophysiol. 28:166–185.
Renzi C, Ferrari C, Schiavi S, Pisoni A, Papagno C, Vecchi T,
Antal A, Cattaneo Z. 2015. The role of the occipital face area
in holistic processing involved in face detection and dis-
crimination: a tDCS study. Neuropsychology. 29(3):409–416.
Rush S, Driscoll DA. 1968. Current distribution in the brain from
surface electrodes. Anesth Analg. 47(6):717–723.
Salkoff DB, Zagha E, Yuzgec O, McCormick DA. 2015. Synaptic
mechanisms of tight spike synchrony at gamma frequency
in cerebral cortex. J Neurosci. 35:10236–10251.
Sherry A, Hensen RK. 2005. Conducting and interpreting canon-
ical correlation analysis in personality research: a user-
friendly primer. J Pers Assess. 84(1):37–48.
Singer W. 1999. Neuronal synchrony: a versatile code for the
definition of relations? Neuron. 24:49–65. , 111–125.
Soekadar SR, Witkowski M, Cossio EG, Birbaumer N, Robinson SE,
Cohen LG. 2013. In vivo assessment of human brain oscilla-
tions during application of transcranial electric currents. Nat
Commun. 4:2032.
Stagg CJ, Bachtiar V, Amadi U, Gudberg CA, Ilie AS, Sampaio-
Baptista C, O’Shea J, Woolrich M, Smith SM, Filippini N, et al.
2014. Local GABA concentration is related to network-level
resting functional connectivity. eLife. 3:e01465.
Stagg CJ, Best JG, Stephenson MC, O’Shea J, Wylezinska M,
Kincses ZT, Morris PG, Matthews PM, Johansen-Berg H. 2009a.
Polarity-sensitive modulation of cortical neurotransmitters
by transcranial stimulation. J Neurosci. 29:5202–5206.
Stagg CJ, Bachtiar V, Johansen-Berg H. 2011. The role of GABA in
human motor learning. Curr Biol. 21:480–484.
Stagg CJ, Nitsche MA. 2011. Physiological basis of transcranial
direct current stimulation. Neuroscientist. 17:37–53.
Stagg CJ, O’Shea J, Kincses ZT, Woolrich M, Matthews PM,
Johansen-Berg H. 2009b. Modulation of movement-
associated cortical activation by transcranial direct current
stimulation. Eur J Neurosci. 30:1412–1423.
Swettenham JB, Muthukumaraswamy SD, Singh KD. 2009.
Spectral properties of induced and evoked gamma oscilla-
tions in human early visual cortex to moving and stationary
stimuli. J Neurophysiol. 102(2):1241–1253.
Taulu S, Simola J. 2006. Spatiotemporal signal space separation
method for rejecting nearby interference in MEG measure-
ments. Phys Med Biol. 51:1759–1768.
Taulu S, Simola J, Kajola M. 2005. Applications of the signal
space separation method (SSS). IEEE Trans Signal Process.
53:3359–3372.
Uhlhaas PJ, Pipa G, Lima B, Melloni L, Neuenschwander S,
Nikolić D, Singer W. 2009. Neural synchrony in cortical net-
works: history, concept and current status. Front Integr
Neurosci. 3:17.
Uhlhaas PJ, Singer W. 2012. Neuronal dynamics and neuro-
psychiatric disorders: toward a translational paradigm
for dysfunctional large-scale networks. Neuron. 75(6):
963–980.
Uusitalo M, Ilmoniemi R. 1997. Signal-space projection method
for separating MEG or EEG into components. Med Biol Eng
Comput. 35:135–140.
van Veen BD, van Drogelen W, Yuchtman M, Suzuki A. 1997.
Localization of brain electrical activity via linearly con-
strained minimum variance spatial filtering. IEEE Trans
Biomed Eng. 44:70–84.
Vinck M, Womelsdorf T, Buffalo EA, Desimone R, Fries P. 2013.
Attentional modulation of cell-class-specific gamma-band
synchronization in awake monkey area v4. Neuron. 80:
1077–1089.
Weber MJ, Messing SB, Rao H, Detre JA, Thompson-Schill SL.
2014. Prefrontal transcranial direct current stimulation
alters activation and connectivity in cortical and subcortical
reward systems: a tDCS-fMRI study. Hum Brain Mapp. 35:
3673–3686.
Wilson TW, Heinrichs-Graham E, Becker KM. 2014a. Circadian
modulation of motor-related beta oscillatory responses.
Neuroimage. 102:531–539.
Wilson TW, Heinrichs-Graham E, Robertson KR, Sandkovsky U,
O’Neill J, Knott NL, Fox HS, Swindells S. 2013. Functional brain
abnormalities during finger-tapping in HIV-infected older
adults: a magnetoencephalography study. J Neuroimmune
Pharmacol. 8:965–974.
 tDCS Modulates Occipital Gamma Oscillations Wilson et al. | 1609

Wilson TW, Kurz MJ, Arpin DJ. 2014b. Functional special-
ization within the supplementary motor area: a fNIRS
study of bimanual coordination. Neuroimage. 85(Pt 1):
445–450.
Wilson TW, Slason E, Asherin R, Kronberg E, Reite ML, Teale PD,
Rojas DC. 2010. An extended motor network generates beta
and gamma oscillatory perturbations during development.
Brain Cogn. 73(2):75–84.
Wilson TW, Slason E, Asherin R, Kronberg E, Teale PD, Reite ML,
Rojas DC. 2011. Abnormal gamma and beta MEG activity
during finger movements in early-onset psychosis. Dev
Neuropsychol. 36(5):596–613.