Fig. 3. Correlations in long-term measurements of neural oscillatory power across frequency bands.

a Normalized beta power (blue) and theta power (black) were collected concurrently from the left and right STN of one example patient. b Scatter plot of detrended beta power vs detrended theta power shown in a (data were detrended by normalizing values for each day to that day’s mean). For this patient, there was a negative correlation between beta and theta (r = -0.24, p < 0.001). c Median detrended beta (blue) and theta (black) power across the 24-h diurnal cycle, showing different diurnal profiles of theta and beta in this example patient. d Normalized beta power (orange) and theta power (black) were collected concurrently from the left and right STN of a second example patient. e Scatter plot of detrended beta power vs detrended theta power shown in d (data were detrended by normalizing values for each day to that day’s mean). For this patient, there was a positive correlation between beta and theta (r = 0.26, p < 0.001). f Median beta (orange) and theta (black) power across the 24-h diurnal cycle, showing the overlap in the diurnal profiles of beta and theta power in this second example patient. g Pearson’s correlation coefficients for concurrently collected contralateral STN beta and theta power measurements (detrended; 0.18 ± 0.36, n = 5). h Variance explained by time of day in concurrently collected beta and theta power measurements (t(4)=1.1, p = 0.33, 5% confidence intervals (−0.15, 0.35), paired t-test, n = 5). i, j Mean (solid black) and standard deviation (dashed, gray) concurrent diurnal beta (i) and theta (j) power profiles around the diurnal cycle, across the STN of five patients.