Proactive Control: Neural Oscillatory Correlates of Conflict Anticipation and Response Slowing

Abstract

Proactive control allows us to anticipate environmental changes and adjust behavioral strategy. In the laboratory, investigators have used a number of different behavioral paradigms, including the stop-signal task (SST), to examine the neural processes of proactive control. Previous functional MRI studies of the SST have demonstrated regional responses to conflict anticipation-the likelihood of a stop signal or P(stop) as estimated by a Bayesian model-and reaction time (RT) slowing and how these responses are interrelated. Here, in an electrophysiological study, we investigated the time-frequency domain substrates of proactive control. The results showed that conflict anticipation as indexed by P(stop) was positively correlated with the power in low-theta band (3-5 Hz) in the fixation (trial onset)-locked interval, and go-RT was negatively correlated with the power in delta-theta band (2-8 Hz) in the go-locked interval. Stimulus prediction error was positively correlated with the power in the low-beta band (12-22 Hz) in the stop-locked interval. Further, the power of the P(stop) and go-RT clusters was negatively correlated, providing a mechanism relating conflict anticipation to RT slowing in the SST. Source reconstruction with beamformer localized these time-frequency activities close to brain regions as revealed by functional MRI in earlier work. These are the novel results to show oscillatory electrophysiological substrates in support of trial-by-trial behavioral adjustment for proactive control.

Keywords: Bayesian model; Electroencephalogram (EEG); neural oscillation; proactive control; stop-signal task.

Figure 1.

Bayesian model prediction of behavioral performance in the stop-signal task. A, Example sequence of trials. The upper panel shows the sequence of go (blue dots) or stop (red dots) trials and how Bayesian belief about encountering a stop trial [P(stop), black line] increases and decreases, respectively, after each stop and go trial. The lower panel shows the sequence of go-RT in the upper panel. Overall, go-RT tended to be prolonged with a higher P(stop). B, Correlation between P(stop) and RT across all go success trials, with each regression line representing an individual participant. C, Positive correlation between go-RT and P(stop) collapsed over all participants. The plot in the upper panel shows the mean ± SE, the histogram in the lower panel shows the distribution of P(stop), and both were binned at intervals of 0.01. D, Negative correlation between stop error (SE) rate and P(stop), with the same format as in C.

Figure 2.

Trial-by-trial oscillatory power correlates of P(stop), go-RT, and PE at channel FCz. A, Correlation between fixation-locked (onset at 0 ms) power and P(stop) at each time–frequency bin. The color represents the z-value of Spearman correlation, and the black contours represent statistically significant time–frequency clusters in the nonparametric cluster-based permutation test across participants, with clustering threshold at p < 0.02, 0.005, and 0.001 levels (see Materials and Methods for details). It showed a positive correlation in the intervals 3–5 Hz and 0–200 ms. B, No clusters showed a significant correlation between go-locked power and P(stop). C, No clusters showed a significant correlation between fixation-locked power and go-RT. D, Correlation between go-locked power and go-RT, with the permutation test showing a negative correlation in the intervals 2–8 Hz and 200–700 ms. E, The coefficient of trial-by-trial Pearson correlation between the mean power of the clusters identified in A and D of individual participants. Wilcoxon signed rank test showed that the correlation coefficients across participants were significantly below zero. F, G, Correlation between stop-locked power and PE, with the permutation test showing a positive correlation in the intervals 12–22 Hz and 300–400 ms. The topographies of each correlational cluster (p < 0.005) are shown in the bottom panel, where H, I, J, and K each show the cluster of A, D, F, and G, all with the strongest correlations at the midfrontal region. We performed the same analyses at Pz channel (Fig. 2-1).

Figure 3.

Source reconstruction and localization. EEG correlates of conflict anticipation (p < 0.001, uncorrected) were localized to the right SMG and the anterior pre-SMA. B, EEG correlates of RT slowing (p < 0.005, uncorrected) were localized to the middle and inferior frontal gyrus (MFG/IFG) and precentral and postcentral gyrus (PC).