Alpha and theta band activity share information relevant to proactive and reactive control during conflict-modulated response inhibition

Abstract

Response inhibition is an important instance of cognitive control and can be complicated by perceptual conflict. The neurophysiological mechanisms underlying these processes are still not understood. Especially the relationship between neural processes directly preceding cognitive control (proactive control) and processes underlying cognitive control (reactive control) has not been examined although there should be close links. In the current study, we investigate these aspects in a sample of N = 50 healthy adults. Time-frequency and beamforming approaches were applied to analyze the interrelation of brain states before (pre-trial) and during (within-trial) cognitive control. The behavioral data replicate a perceptual conflict-dependent modulation of response inhibition. During the pre-trial period, insular, inferior frontal, superior temporal, and precentral alpha activity was positively correlated with theta activity in the same regions and the superior frontal gyrus. Additionally, participants with a stronger pre-trial alpha activity in the primary motor cortex showed a stronger (within-trial) conflict effect in the theta band in the primary motor cortex. This theta conflict effect was further related to a stronger theta conflict effect in the midcingulate cortex until the end of the trial. The temporal cascade of these processes suggests that successful proactive preparation (anticipatory information gating) entails a stronger reactive processing of the conflicting stimulus information likely resulting in a realization of the need to adapt the current action plan. The results indicate that theta and alpha band activity share and transfer aspects of information when it comes to the interrelationship between proactive and reactive control during conflict-modulated motor inhibition.

Keywords: EEG; LCMV beamforming; alpha power; proactive control; response inhibition; theta power; time-frequency analysis.

FIGURE 1

Analysis pipeline. On the left side, each step of the data analysis is depicted. The arrows with a continuous line represent the chronological sequence of the individual steps. The arrows with a dashed line lead to a brief explanation of the respective steps on the right side. Comp., compatible; DBSCAN, Density‐Based Spatial Clustering of Applications with Noise; DICS, Dynamic Imaging of Coherent Sources; FDR, False Discovery Rate; Hz, Hertz; ICA, Independent Component Analysis; incomp., incompatible; LCMV, Linear Constraint Minimum Variance; min. neighbors, minimum number of neighbors; NAI, Neural Activity Index; ROI, Region of Interest; ε, epsilon. Images of eye and ear: ©SvtDesign—Can Stock Photo Inc.

FIGURE 2

Summary of within‐trial results. (a) Time‐frequency representation over the significant fronto‐central electrodes (see topographic plot in the top right corner). Values represent the power difference of the incompatible − compatible condition. The square represents the time frame where the difference between incompatible and compatible Nogo trials was significant in FDR‐corrected t‐tests. The topographic plot shows the power difference in the theta band in the significant timeframe (0 to ~200 ms). The red crosses mark electrodes with significant power differences. (b) Within‐trial theta clusters of the conflict effect, as identified by beamforming and the DBSCAN algorithm (see Section 2) on the theta power difference between the incompatible and compatible Nogo condition. The color of the cluster scales the power difference between the incompatible and compatible Nogo condition. (c) Time‐frequency representation over electrode P7 (see topographic plot). Values represent the power difference between the incompatible and compatible Nogo condition. The square indicates the area of significant differences between the two conditions in the alpha band; the topographic plot shows the distribution of alpha power differences in the significant timeframe (~550 to 700 ms). (d) The within‐trial alpha cluster associated with the conflict effect as defined by beamforming and subsequent clustering with the DBSCAN algorithm. The color of the cluster scales the power difference between the incompatible and compatible Nogo condition.

FIGURE 3

Summary of the pre‐trial clusters. (a) The four clusters of theta band activity in the pre‐trial period as identified by beamforming and subsequent application of the DBSCAN algorithm. (b) The four alpha clusters identified for the pre‐trial period. The color of the clusters in both parts of the figure represents the magnitude of the Neural Activity Index (NAI).

FIGURE 4

Summary of the main findings of the correlations in the within‐trial interval (0 to 0.6 s) after stimulus onset (a and b) and between the pre‐ and within‐trial interval (c). Please note that only significant correlation patterns are displayed. In each section, the top plot visualizes the location of the correlated clusters. Below, the correlation is depicted via a correlation matrix of the activity time courses within the respective clusters. The left plot scales the magnitude of the correlation (r‐value, indicated by color; see color bar at the bottom left of the figure). The right plot visualizes FDR‐corrected q‐values (indicated by color; see color bar at the bottom right of the figure). Black areas indicate no significance (q > .05), while lighter areas visualize significant correlations (q < .05).

FIGURE 5

Summary of the correlations in the pre‐trial interval. The four pre‐trial theta clusters are depicted on the left side while the four pre‐trial alpha clusters are depicted at the top. Each correlation plot shows the correlation between the clusters on the respective locations at the axes. The correlation plot visualizes the magnitude of the correlation coefficient between the two respective time points (r‐value, indicated by color; see color bar at the bottom left of the figure).

FIGURE 6

Schematic illustration of the relationships between proactive (pre‐trial) and reactive (within‐trial) processes during conflict‐modulated motor inhibition. Pre‐trial alpha activity in the primary motor cortex is positively correlated with the within‐trial theta conflict effect in the primary motor cortex (~300 ms after stimulus onset). “Ico > co” refers to the fact that theta power was higher in incompatible Nogo trials than in compatible Nogo trials. This within‐trial theta conflict effect in the primary motor cortex is then positively correlated with the within‐trial theta conflict effect in the midcingulate cortex until the end of the trial. These interrelations suggest that participants with stronger anticipatory information gating in the primary motor cortex (alpha band) show stronger reactive processing of the perceptual conflict in the primary motor cortex (theta band). The results further imply that as a next step, participants with stronger processing of the perceptual conflict in the primary motor cortex (theta band) show a stronger realization of the need for adapting the current response plan in the midcingulate cortex (theta band). In sum, theta and alpha band activity seem to share and transfer aspects of information when it comes to the interrelationship of proactive and reactive control during conflict‐modulated motor inhibition.