Objective: Different frequencies within the beta range may have distinct temporal dynamics and an individual assessment should be performed to select the optimal feedback signal for aDBS.

Background: Adaptive deep brain stimulation (aDBS) represents a next generation precision medicine instrument for the treatment of patients with movement disorders such as Parkinson’s disease (PD). In contrast to continuous DBS, aDBS adjusts the delivery of stimulation matched to the temporal dynamics of a pre-defined feedback signal. One of the most promising feedback signals includes beta activity (13-30 Hz) of the subthalamic nucleus (STN) local field potentials (LFP) that has been associated with hypokinetic symptoms in PD.

Method: We recorded LFP from the STN in 15 PD patients OFF-medication while they performed a visual cued joystick reaching task to assess the peak velocity of the movements. The raw signal was decomposed using Morlet Wavelet (width = 10, gwidth = 5) and the impact of burst properties corresponding to the different beta frequencies (from 13 Hz to 30 Hz, 1 Hz resolution) on motor slowing were assessed. In addition, the relationship of the temporal dynamics across frequencies was determined using the percentage burst overlapping metric (%OVL).

Results: The temporal occurrence of beta bursts across the frequency range from 13-30Hz differs and can even be temporally independent. Comparing the envelope of the beta peak frequency with more distant envelopes (±2Hz), the %OVL of bursts reduces to 60% and further to 40% if the envelopes deviate ±4Hz from the beta peak frequency. Note, the change in %OVL also depends on the spectral resolution of the frequency decomposition. Moreover, we found that the individual beta frequency strongest associated with motor slowing, does not always coincide with the individual beta peak and patients may show no clear or two peaks within the beta spectrum. Simulating the aDBS trigger timings across different beta frequencies reveals that deviations from a pre-selected target frequency may cause a relevant drop in the matched trigger timings with the risk of a non-beneficial adaptive stimulation pattern.

Conclusion: The diversity in temporal dynamics of beta envelopes within the beta range can be large. To achieve the greatest possible clinical benefit using aDBS, an individual electro-clinical interrogation should be performed to determine the optimal frequency range of the feedback signal for aDBS.

« Back to 2022 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/frequency-specific-dynamics-of-beta-oscillations-critical-for-tuning-adaptive-dbs/