Multiple Objective Particle Swarm Optimization Postural Instability Gait Disorder (MOPSO PIGD)

Sixty patients will be enrolled in this study who are treated for Parkinson's disease (PD) with bilateral deep brain stimulation of subthalamic nucleus (STN) or globus pallidus (GP), who have a pre- operative 7 Tesla MRI including diffusion tensor imaging for tractography and a postoperative head CT for electrode localization, and in whom at least 3 months have passed since activation of their neurostimulators, for stabilization of clinical stimulator settings. Using their MRI and CT, the investigators will construct patient-specific models of electrical current spread to neuroanatomical tar- gets surrounding the electrode. Then applying nonlinear (particle swarm) optimization, patient- specific stimulator settings will be designed to maximally or minimally activate specific path- ways. In STN DBS: pedunculopallidal vs. pallidopeduncular pathways. In GP DBS: pallidopeduncular pathways at its origin in GP pars interna (GPi) vs. inhibitory afferents to GPi (from GP pars externa GPe). All stimulation falls within the the FDA-approved range for DBS for PD.

Patients will undergo on-label DBS for the treatment of Parkinson's disease, at stimulator settings different from their usual settings, but within the FDA-approved range for this indication, based on patient specific models that identify stimulation settings that most selectively target one pathway over other adjacent pathways. In Aim 1 (STN DBS), the comparison will be amongst settings that selectively target pedunculopallidal, pallidopeduncular and subthalamopallidal pathways. In Aim 2 (GP DBS), the comparison will be amongst settings which selectively target the pallidopeduncular pathways at its origin in GP pars interna (GPi), inhibitory afferents to GPi (from GP pars externa GPe), and pallidosubthalamic pathway (from GPe). Total time, including both test stimulation and off-stimulation time, will be about 6 hours for Arm 1, and about 5 hours for Arm 2. Testing will include quantitative measures of gait, balance, bradykinesia, and rigidity.

Patients with implanted STN DBS leads will be allocated to this group for DBS stimulation of the STN, within FDA-approved limits and labeling, for symptoms of PD.

Patients with implanted GP DBS leads will be allocated to this group for DBS stimulation of the GP, within FDA-approved limits and labeling, for symptoms of PD.

Subjects walk for one minute each on a treadmill set to match their individual overground walking speed. A forceplate measures the center of pressure trajectory from which integral software detects time and location of heel-strike and toe off events, and from these, plus the treadmill speed, we compute step lengths.

Subjects stand on the treadmill, with the treadmill initially motionless. The treadmill belt then jerks forward, so that the subject must step to recover balance. Falls are prevented by a safety harness and assisting staff. A rater logs number of steps taken. Aggregated as mean of 5 trials.

Subjects sit with forearm supported, grasping a manipulandum which measures pronation/supination angle, making a pronation/supination motion actively for 30 seconds, instructed as big and fast. Aggregated as root mean square (RMS) of mean-subtracted angle.

Subjects sit with forearm supported, grasping a manipulandum, which is connected to a motor and torque sensor. The motor rotates the forearm through 40 degress of pronation-supination, cyclically, at 1.5 cycle per second for 45 seconds. Aggregated as mean absolute value of torque.

Inclusion Criteria: Diagnosis of PD Bilateral DBS Directional DBS electrode system MRI, at least 7 Tesla, with diffusion tensor imaging, prior to DBS surgery Head CT, at least 1mm cuts, following DBS surgery At least 3 months since initial activation of the neurostimulator Exclusion Criteria: Neurological or musculoskeletal disorder that significantly affects gait, balance, or limb movements Dementia Prior stereotactic neurosurgery other than DBS for movement disorder Pregnancy