Figure 4.

Consecutive phase-consistent stimulus pulses modulate beta oscillation amplitude. Stimulation at the patient-specific suppressing phase for at least three consecutive cycles led to suppression of STN LFP beta oscillations across eight patients. A, Percentage change in beta oscillation amplitude from the median after consecutive cycles of stimuli occurring at a consistent phase is shown for four phase bins in an example subject. In this subject, amplitude suppression was seen after three consecutive cycles of stimuli delivered on the descending phase of the oscillation (blue), whereas amplification was seen after three consecutive cycles of stimuli delivered on the ascending phase (orange). Stimuli delivered at alternative phases (red, green) did not result in modulation of the beta amplitude. B, Suppressing and amplifying phase bins for each patient. C, Phase difference between the amplifying and suppressing phase bin for each patient. D, Median suppressing (blue) and amplifying (orange) effects were grouped across eight patients. The percentage change in beta amplitude was compared with surrogate effects (identical analysis on unstimulated segment of the recording sampled at the stimulus frequency; gray). Beta suppression was dependent on the number of consecutive stimuli delivered at the suppressing phase of the oscillation (χ² = 17.38, p = 0.00160, Kruskal–Wallis test), whereas beta amplification was not (χ² = 6.12, p = 0.190, Kruskal–Wallis test). As six consecutive stimuli were only observed in four of the eight patients at the suppressing phase and two of eight patients at the amplifying phase (indicated by lighter boxes), these was not included in the Kruskal–Wallis test. Horizontal lines with black asterisks indicate differences between groups (post hoc Dunn test to correct for multiple comparisons, p ≤ 0.05). Red asterisks indicate stimulation effects significantly different from surrogates (p ≤ 0.05, Wilcoxon ranked sum test, FDR-corrected).