Evaluating Thalamic Biomarkers for Closed-Loop Deep Brain Stimulation in Essential Tremor

Deep brain stimulation (DBS) and associated local field potential (LFP) recordings offer promising avenues for developing closed-loop stimulation systems for neurological disorders such as Essential Tremor [1]-[3]. In this study, we evaluated the potential of LFP band power and inter-lead coherence as biomarkers for intended movement detection in a patient implanted with dual, contralaterally placed leads using a Medtronic Percept™ RC neurostimulator (with BrainSense™) and SenSight electrode technology. LFPs were recorded from each lead's electrodes in the ventral intermediate (VIM) nucleus and the ventralis-oralis (VO) nuclei of the thalamus while the patient performed two motor tasks: a Drink Cup task, involving grasp-and-hold movements, and a postural holding task. Simultaneously recording accelerometers were used to quantify movement. Generalized linear models (GLMs) were developed to decode movement onset from LFP band power on each lead individually, as well as from combined dual-lead and coherence predictors. Receiver operating characteristic (ROC) curves were generated for each condition over a 15-month follow up period for a 74-year old female patient. Results indicated that, for individual leads, the VIM channel achieved its highest average AUC for the Drink Cup task in the Delta band (0.6009), which was significantly higher than the VO. In contrast, the VO lead demonstrated its best performance in the Beta band, with average AUCs of 0.6444 for the Posture task and 0.6172 for the Drink Cup task, both significantly higher than those from VIM. Importantly, integrating information from both leads across all frequency bands significantly improved detection performance, with overall AUCs exceeding 0.71.Clinical relevance- This study is the first to demonstrate that LFP signals recorded from the ventralis-oralis of the thalamus can serve as a possible biomarker for closed loop DBS for Essential Tremor. Furthermore, by integrating information from both VO and VIM leads, detection performance is significantly enhanced. These novel findings pave the way for more effective, adaptive closed-loop DBS therapies for Essential Tremor, and possibly other movement disorders.