M1-STN Coherence and Connectivity Analyses Using a Novel, High-density Array

Introduction: The subthalamic nucleus (STN) is a key target for deep brain stimulation (DBS) in treating Parkinson's disease (PD). Understanding cortex-STN interactions can help elucidate PD pathophysiology and how DBS may alter this physiology to improve PD motor symptoms. This study examines cortex-STN interactions in a PD patient during hand movement, focusing on coherence, a measure of signal correlation, and Granger causality, which assesses the direction and strength of information flow between regions.

Methods: A PD patient undergoing DBS lead placement had a 1024-contact cortical surface array in a 2x2cm area placed over the hand knob of M1 and a microelectrode inserted into the dorsolateral STN. The patient performed hand movements while cortical and STN activity were recorded. Coherence calculations were performed between each cortical contact and the STN to evaluate spatial patterns. A Mann-Whitney U test was performed to assess significance across different regions on the array and the Bonferroni correction was used for multiple comparisons correction. Finally, Granger Causality analysis was performed.

Results: Coherence was significantly higher in the anterior portion of the hand knob compared to the medial and posterior portions (p < 0.01, U = 2e7 ) for the high beta (12-35Hz) band during movement. Granger Causality analysis showed that cortical data improved prediction of STN states, just before movement onset and termination (p < 0.05, t = -2.4), with the highest accuracy observed when using cortical data from 20 to 0 ms prior. No significance was observed in the STN to cortex direction around movement.

Conclusion : Our findings reveal specific spatial coherence patterns between the M1 and STN with the anterior hand knob having highest coherence during movement. This spatial specificity may offer valuable insights into the functional connectivity between the STN and M1. Additionally, we saw a higher Granger causal relationship in the M1 to STN direction, just prior to movement onset and termination. By understanding these interaction dynamics, DBS could be more effectively tailored to disrupt maladaptive M1-STN connectivity patterns, thereby improving motor control.