Advanced target identification in STN-DBS with beta power of combined local field potentials and spiking activity

Abstract

Background: In deep brain stimulation of the subthalamic nucleus (STN-DBS) for Parkinson's Disease (PD), often microelectrode recordings (MER) are used for STN identification. However, for advanced target identification of the sensorimotor STN, it may be relevant to use local field potential (LFP) recordings. Then, it is important to assure that the measured oscillations are coming from the close proximity of the electrode.

New method: Through multiple simultaneous recordings of LFP and neuronal spiking, we investigated the temporal relationship between local neuronal spiking and more global LFP. We analyzed the local oscillations in the LFP by calculating power only over specific frequencies that show a significant coherence between LFP and neuronal spiking. Using this 'coherence method', we investigated how well measurements in the sensorimotor STN could be discriminated from measurements elsewhere in the STN.

Results/comparison with existing methods: The 'sensorimotor power index' (SMPI) of beta frequencies, representing the ability to discriminate sensorimotor STN measurements based on the beta power, was significantly larger using the 'coherence method' for LFP spectral analysis compared to other methods where either the complete LFP beta spectrum or only the prominent peaks in the LFP beta spectrum were used to calculate beta power.

Conclusions: The results suggest that due to volume conduction of beta frequency oscillations, proper localization of the sensorimotor STN with only LFP recordings is difficult. However, combining recordings of LFP and neuronal spiking and calculating beta power over the coherent parts of the LFP spectrum can be beneficial in discriminating the sensorimotor STN.

Keywords: Coherence; Deep brain stimulation; Local field potential; Neuronal spiking activity; Parkinson's disease; Subthalamic nucleus.