Fig. 6

Stimulation-induced power across frequency bands. a Stimulation-induced power spectrogram for an example electrode from a single subject (stimulation in left MTL white matter, recording electrode in left inferior parietal cortex). b Post- minus pre-stimulation difference in power from the electrode in a. c Whole-brain stimulation-induced power was measured by computing a t-statistic on the pre- vs. post-stimulation spectral power at each electrode in a subject, and then averaging across electrodes to get an estimate of whole-brain change in power. On average, stimulation elevated whole-brain power in the theta (5–8 Hz), alpha/beta (10–25 Hz), and gamma (30–50 Hz) bands (one-sample t-test, FDR-corrected P < 0.05). Stimulation decreased power in the HFB range (50–200 Hz). d For each subject/stimulation site, electrodes were classified by whether they exhibited a significant (t > 2) change in theta power induced by stimulation (at least 1 theta-responsive electrode was found for 47 of the 72 stimulation sites). The stimulation-induced change at higher frequencies was computed for this subset of electrodes, to determine whether the power response was specific to theta. Across all subjects/stimulation sites, increased power was observed in the alpha/beta and gamma range at theta-responsive electrodes, but no effect was observed in the HFB range (one-sample t-test, t(46) = 0.25). The bar for induced theta power is delineated by a dashed line and shown as a reference only, since theta power was the basis for selecting these electrodes for further analysis. Error bars show ±1 SEM; *corr. P < 0.05; **corr. P < 0.01