2026 American Academy of Neurology Abstract Website

Objective:

To characterize the prevalence of stimulation-related artifacts in intracranial recordings obtained from deep brain stimulation (DBS) devices, and to determine whether artifact occurrence varies by brain target or electrode contact.

Background:

Deep brain stimulation (DBS) is a therapy for medication-refractory movement disorders such as Parkinson’s disease (PD), essential tremor (ET), and dystonia. Modern devices can record intracranial signals to guide programming, but these recordings may be distorted by stimulation-related artifacts. We aim to quantify how frequently artifacts occur and their potential to falsely elevate electrophysiological activity.

Design/Methods:

We retrospectively reviewed patients with PD, ET, or dystonia who were implanted with sensing-enabled DBS devices (Medtronic Percept) targeting the subthalamic nucleus (STN), globus pallidus internus (GPi), or ventralis intermedius nucleus (VIM). Recordings were obtained using the BrainSense Streaming feature. Artifacts were defined as a linear relationship between stimulation amplitude and spectral power, indicating signals artificially elevated by stimulation rather than true neural activity. Chi-square tests were used to compare the distribution of observations across different scenarios to determine whether significant differences existed between categories.

Results:

742 recordings were analyzed: 85% PD, 9% ET, and 6% dystonia. Targets included GPi (68%), STN (22%) and VIM (10%). Stimulation-related artifacts were detected in 59 of 628 (9%) PD, 2 of 70 (3%) ET, and 1 of 44 (2%) dystonia recordings. Artifact prevalence differed significantly across targets, with GPi showing the highest proportion of artifact-positive recordings at approximately 11% (χ²=13.27, p=0.0013). Within each target, artifact distribution between E01 and E02 contacts was not significantly different (GPI: χ²=1.19, p=0.27; STN: χ²=1.57, p=0.21; VIM: χ²=1.78, p=0.18).

Conclusions:

Stimulation-related artifacts occur infrequently but remain an important consideration in DBS recordings, particularly in those with GPi targets. For future research, these results may help model the relationship between stimulation amplitude and artifact existence, enabling development of normalization strategies to improve interpretation of data.