Increased phase synchronization between ACC and FEF might partially reflect an independent phase reset in both areas aligned on stimulus onset. To examine this, we subtracted the averaged stimulus-aligned evoked LFP from the signals before calculating phase synchrony. This procedure resulted in decreased ACC-FEF theta-band phase synchronization during the early delay period (Fig. 3c compared to Supplementary Fig. 3C) but did not affect the increased phase synchronization in the late delay period (400–1,100 ms following the stimulus onset). This finding suggests that the increased phase synchronization in the 400–1,100 ms following the stimulus period is not time locked to the stimulus onset and reflects genuine ACC-FEF synchronization (Supplementary Fig. 3). On the other hand, we did not observe this decrease in the immediate poststimulus beta-band phase synchronization of the average-subtracted data compared to that of the original data (Fig. 3d compared to Supplementary Fig. 3D). This suggests that beta-band ACC-FEF phase synchronization primarily represents an induced oscillatory response that is not evoked by the stimulus onset. We also subtracted the evoked response by applying a different method described by Truccolo
