Objective: To investigate the possibility of tremor detection based on deep brain recordings.

Background: Electrophysiological studies have revealed associations between neuronal oscillations in the subthalamic nucleus (STN) and symptoms of Parkinson’s disease (PD), such as rest tremor. This knowledge can potentially be used to improve therapy, e.g. by using oscillations to inform a closed-loop deep brain stimulator about the presence of tremor.

Methods: We re-analyzed STN local field potential (LFP) recordings from 10 PD patients (12 body sides) with spontaneously fluctuating rest tremor. STN power in several frequency bands was estimated and used as input to Hidden Markov Models (HMMs) which classified short data segments as either tremor-free rest or rest tremor. HMMs were compared to direct threshold application to band-limited power.

Results: Applying a threshold directly to band-limited power was not sufficient for tremor detection. Multi-feature HMMs, in contrast, allowed for tremor detection with high accuracy (mean: 0.84, STD: 0.11), using four power features obtained from a single electrode contact pair. Within-patient training yielded better accuracy than across-patient training (0.84 vs. 0.78, p = 0.03), yet tremor could often be detected accurately with either approach. High frequency oscillations (>200 Hz) were the best performing individual feature.

Conclusions: LFP-based markers of tremor are robust enough to allow for accurate tremor detection in short data segments, provided that appropriate statistical models are used. They could thus be useful control signals for closed-loop deep brain stimulation in tremor patients.

« Back to 2017 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/inferring-the-presence-of-parkinsonian-rest-tremor-from-subthalamic-local-field-potential-recordings/