Uncovering network mechanism underlying thalamic Deep Brain Stimulation using a novel firing rate model

Abstract

ABSTRACTThalamic ventral intermediate nucleus (Vim) is the primary surgical target of deep brain stimulation (DBS) for reducing symptoms of essential tremor. In-vivo single unit recordings of patients with essential tremor revealed that low frequency Vim-DBS (≤50Hz) induces periodic excitatory responses and high-frequency Vim-DBS (≥100Hz) induces a transient excitatory response lasting for ≤600ms followed by a suppressed steady-state. Yet, the neural mechanisms that generate Vim firing rate in response to different DBS frequencies are not fully uncovered. Previously developed models of Vim neurons could not capture the full dynamics of Vim-DBS despite incorporating the dynamics of short-term synaptic plasticity. In this work, we developed a network rate model and a novel parameter optimization method to accurately track the instantaneous firing rate of Vim neurons in response to various DBS frequencies, ranging in low- and high-frequency (5 to 200Hz) Vim-DBS. We showed that the firing rate dynamics during high frequency Vim-DBS are best characterized when incorporating an inhibitory population into the network model. Further, we discovered that the Vim firing rate in response to varying frequencies of DBS pulses can be explained by abalanced amplificationmechanism, in which strong excitation (Vim) is stabilized by equally strong feedback inhibition. As a further validation of this work, we demonstrated similar behavior in a detailed biophysical model consisting of spiking neural networks in which the Vim neural-network can implement balanced amplification and explain in-vivo human Vim-DBS observations.Graphical abstract