Analysis of Intraoperative MER Signals in Pediatric Patients with Dystonia Undergoing DBS of the Globus Pallidus Internus

Dystonia is a movement disorder characterized by abnormal postures and repetitive movements, often treated with Deep Brain Stimulation (DBS) in cases resistant to other therapies. Intraoperative MicroElectrode Recordings (MER) are typically used to guide electrode placement during DBS surgery, but this process remains largely subjective. This study explores an automated method to analyze MER data and identify optimal implantation sites, aiming to improve surgical outcomes. MER recordings were collected from five pediatric patients undergoing DBS targeting the Globus Pallidus internus (GPi), with data obtained on three trajectories -anterior, central, and posterior- at multiple depths. Preprocessing involved filtering high-energy interferences and applying an automated spike-sorting pipeline to extract features such as firing rate, firing regularity, and oscillatory activity. Statistical analysis showed that the selected trajectory was associated with the highest neural activity in terms of number of spikes (131.65 ± 20.13 vs. 98.44 ± 9.56, p < 0.05). Analysis of patients with positive outcomes (4 out of the 5 patients of our dataset) at the selected trajectory revealed significantly higher beta frequency features, as mean and peak power. GPi neurons from the negative outcome patient showed higher firing rates (109.43 ± 73.70 Hz vs. 19.90 ± 5.29 Hz, p < 0.001) with respect to the neurons of the GPi of the positive outcome group. These findings show potential for a relationship between features obtained from automatic analysis and the identification of the optimal trajectory and depth for DBS permanent electrode implantation and provide information on possible outcomes.