Facilitated Event-Related Power Modulations during Transcranial Alternating Current Stimulation (tACS) Revealed by Concurrent tACS-MEG

Experimental procedures. A, Time course of the experiment. Blue indicates periods during which the MR task was performed; gray indicates intermittent resting periods. B, Positions of stimulation electrodes (red/blue) and layout of MEG sensors (yellow/green). Stimulation electrodes were placed centered above Cz (7 × 5 cm) and Oz (4 × 4 cm) of the international 10-10 system. MEG was recorded from 102 locations. Each location contains a sensor triplet of one magnetometer and two orthogonal planar gradiometers, resulting in a total of 306 channels. Sensor locations used to determine participants’ individual alpha frequency are marked green. C, Mental rotation task. Each trial started with the presentation of a white fixation cross at the center of the screen. After 3000 ms, a mental rotation stimulus (two objects) was presented and remained on screen for another 7000 ms. During this time participants were required to judge whether the two objects presented were either different (example depicted in 2nd display) or identical (but rotated; 4th display). A and C are adapted from Kasten and Herrmann (2017).

Behavioral results. A, Change in task performance for stimulation and sham group, relative to baseline, pooled over all experimental blocks. Boxes indicate the 25th and 75th percentile of the sample distribution (interquartile length); lines inside the boxes mark the median. Whiskers extend to the most extreme values within 1.5 times the interquartile length. Asterisks code for significance (*, p < 0.05). B, Change in task performance relative to baseline for stimulation and sham group depicted over experimental blocks. The gray area indicates blocks that were performed during tACS or sham stimulation. C, Change in RT for stimulation and sham group relative to baseline pooled over experimental blocks. D, Change in RT for stimulation and sham group relative to baseline depicted over experimental blocks. Gray area indicates blocks that were performed during tACS or sham stimulation.

Event-related alpha-power modulation. A, Region of interest (ROI). Significant cluster (pre- vs. post-stimulus power) in the IAF-band during the first block before tACS or sham stimulation, computed over the whole sample (p_cluster < 0.001). Topographies depict t-values mapped on an MNI standard surface. Statistical maps are thresholded at α < 0.01. The depicted cluster (blue) was used as ROI to extract the time course of alpha-power modulation, relative to baseline, over blocks from the virtual channels. B, Relative alpha-power modulation within ROI depicted for each block. The gray area indicates blocks during tACS or sham stimulation. Shaded areas represent standard error of the mean (SEM). Dashed line depicts baseline level. C, Relative alpha-power modulation during tACS or sham (online) and after stimulation (offline). Error bars represent SEM; asterisks code for significant differences (*, p < 0.05). D, Relative alpha-power modulation during stimulation correlated with stimulation intensity. Each point represents a single subject’s stimulation amplitude and relative alpha-power modulation, averaged over the two stimulation blocks (blocks 2 and 3). Please note that a stimulation intensity was determined for all participants (including sham); however, only participants in the stimulation group had this intensity continuously applied during blocks 2 and 3. E, Relative power modulation in the lower beta-band (IAF + 3 Hz to IAF + 11 Hz) within the ROI for each block. F, Relative power modulation in the higher beta-band (IAF + 12 Hz to IAF + 20 Hz) within the ROI for each block. G, H, Correlation between change in task performance and relative alpha-power modulation during (G) and after (H) tACS. High, albeit nonsignificant, correlations were evident for the sham group, but not the stimulation group.

Normalized, baseline-subtracted TFRs and source topographies. TFRs and source topographies for stimulation (top rows) and sham group (bottom rows). TFRs were aligned at IAF and averaged over subjects in each group. The range from –2.5 to –0.5 before stimulus onset (white bar) served as reference period for baseline subtraction. Spectra were subsequently normalized by the power difference in the alpha-band (IAF ± 2 Hz) during the baseline block (block 1) before stimulation. Normalization was performed such that the data presented resemble data in the statistical analysis. Blocks 2 and 3 (dark gray) represent data acquired during tACS or sham stimulation. All other blocks (light gray) were measured in absence of stimulation. Functional maps were averaged over subjects and projected onto an MNI standard surface. Only activity within the analyzed ROI is depicted. A strong facilitation of event-related power modulation around the IAF can be observed during tACS application (block 2 and 3).

Artifact-to-brain-signal topographies. Topographies depict the average ratio between participants’ pre-stimulus alpha-power, estimated during the baseline block, and residual artifact in the pre-stimulus interval during block 2 (top row) and 3 (bottom row). Results are depicted only for the stimulation group. The ratio is strongest in central areas covered by the stimulation electrodes and cables. Frontal and posterior areas within the ROI seem less affected, with the ratio falling in a physiologically plausible range (<1:4), such that residual artifact and facilitatory effects of the stimulation or spontaneous increase of alpha power cannot be disentangled. Results have to be interpreted in terms of an upper boundary for the size of the residual artifact, as each virtual channel contains a mixture of brain signal of interest and artifact.