Brain-Machine Interface for Freezing of Gait

Gait problems in Parkinson's disease (PD), especially freezing of gait (FOG), greatly affect quality of life. While deep brain stimulation (DBS) is a highly effective treatment for many motor symptoms of PD, it is less effective for, or can even worsen, gait issues. The primary motor cortex (M1) plays a crucial role in the network that controls gait, particularly in initiating movement. Changes in local field potentials (LFPs) from the subthalamic nucleus (STN) are associated with different aspects of gait. However, detecting abnormal brain activity related to FOG requires a method called electrocorticography (ECoG), which has revealed that during FOG, there is increased beta-gamma phase amplitude coupling (PAC) in the M1.

Brain-machine interfaces (BMIs) have shown promise in understanding motor functions by decoding brain activity. It is believed that BMIs could provide both accurate indicators of FOG and targeted treatments for it in PD.

Our objectives are to use a high-density ECoG-based BMI to both record and stimulate brain activity during real-world gait and FOG in PD patients who are undergoing standard DBS procedures. Our goals are to improve our understanding of the brain's role in FOG and normal gait in PD and to develop new treatments based on cortical stimulation.

Aim 1 - Identify gait biomarkers: brain activity from the M1/SMA cortex during different phases of walking and during FOG episodes, both with and without medication will be recorded. Machine learning will be used to identify the brain patterns linked to FOG.

Aim 2 - Use cortical stimulation to stop FOG: Cortical stimulation and its effects on leg and trunk movements will be studied by measuring muscle activity, movement, and posture during different states, such as resting, standing, walking, and during FOG episodes. The type of stimulation which is most effective at stopping FOG will be identified.