Investigation of Spatio-Spectral Dynamics of Local Field Potentials in Parkinson’s Disease

Parkinson’s disease (PD) is characterized by the progressive loss of dopaminergic nigral neurons resulting in motor and non-motor deficits. Deep brain stimulation (DBS) of the subthalamic nucleus (STN) has emerged as an effective neurosurgical treatment for the patients with PD where their motor symptoms cannot be controlled with medications. Accurate localization of STN is an important factor defining the efficacy of DBS. The most common targeting method in DBS surgery is the microelectrode single unit activity recording, which is performed by listening to bursting firing patterns of individual neurons to identify the basal ganglia structures. However, it requires significant expertise and is fraught by potential technical difficulties. On the other hand, local field potentials (LFPs), owing to their oscillatory and robust nature, can overcome these technical issues. In this regard, we recorded LFPs from multitrack microelectrodes and macroelectrode in PD patients who underwent DBS surgery. We demonstrated for the first time that combination of different subband features derived from beta and high frequency oscillations of LFPs can be used to estimate the optimal track for DBS implantation and to identify the dorsal STN border with high accuracy. These results establish the initial evidence that LFPs can be strategically fused with computational intelligence in the operating room to increase the chance of optimal placement of the DBS electrode within the motor sub-territory of the STN, without an appreciable downside. We also investigated the spatio-spectral patterns of LFPs in the most commonly accepted subtypes of PD, tremor dominant (TD) and postural instability and gait difficulty (PIGD). As of today, no underlying neural correlates have been identified. Here we show that activity in the subbands of LFPs recorded with microelectrodes from sub-territories of the STN provide distinguishing neurophysiological information about these phenotypes. We found distinct patterns between TD and PIGD groups in HFOs and their interaction with the beta band in the dorsal and ventral regions of the STN. Our results indicate that the spatio-spectral dynamics of LFPs can be used as an objective method to distinguish the two major subtypes of PD. This observation provides support for distinct pathophysiologic mechanisms underlying these subtypes.