LFP power also exhibited a contralateral bias, especially for higher frequencies. Gamma power (~40–100 Hz) was significantly elevated relative to baseline during sample object presentation and a ‘‘ramp-up’’ at the end of the delay for both contralateral (Figure 2C) and ipsilateral (Figure 2D) samples (p < 0.01, sign test; summarized in Figure 2F). Gamma power induced by contralateral sample objects was significantly higher than for ipsilateral objects during the sample onset and offset transients and the pre-test ‘‘ramp-up’’ (Figure 2E; p < 0.01, paired t test). Theta power (~3–8 Hz) also showed a contralateral bias during the sample response but not the memory delay (Figures 2E and 2H). In contrast, beta power (~10–32 Hz) showed effects in the opposite direction overall: significant decreases in power from baseline during the sample and late delay periods for both contralateral and ipsilateral sample objects (Figure 2G). Like gamma and theta enhancement, beta suppression was significantly stronger for contralateral than ipsilateral samples (Figure 2E) but, like theta power, only during the sample object response. Beta oscillatory burst rates exhibited a stronger and more sustained
