Fig. 1: Beta oscillatory power measured with the Percept DBS device shows a consistent diurnal pattern with decreased beta during the night.

a A 10-s LFP recording from the STN of a Parkinson’s disease patient implanted with the Medtronic Percept DBS device, acquired during the BrainSense signal test. b Power spectrum of the LFP segment in a (obtained using Welch’s method) presented on a log-log scale, showing a peak in the beta frequency range. c Mean STN peak-to-peak beta power (µVp, sampled every 10 min) over a 1-month period, from the same example STN (after outlier removal). d Heat map of beta power (detrended by normalizing each day to its median value) across the 24 h of the day for all days in the data collection period, for the same example STN. e Detrended beta power across the 24-h diurnal cycle generated from the data in d. For each day, the median beta power was calculated for each 30-min time bin, and the bar height in the circular bar graph represents the median across days. f Normalized beta power measurements plotted against the time of day the measurements were taken for the same example STN, showing a consistent daily pattern of beta power. The black line represents a linear fit through the mean beta power for the time of day (means obtained by dividing the day into 30-min time bins). g Periodogram of the beta power signal for all STN (normalized; estimated using Welch’s method). The population mean (black) highlights a clear peak at a period of 24 h, which is present in all individual beta power time series (gray lines, n = 9). h The proportion of variance explained by time of day, estimated using fits such as illustrated in d, for all STN in the data set (mean 0.41 ± 0.092; n = 9; p < 0.001 for all-time series, temporal shuffling test). i Mean (solid line) and standard deviation (dashed line) of detrended beta power around the diurnal cycle across all STN time series (n = 9).