Local field potential driven Izhikevich model predicts a subthalamic nucleus neuron activity.

An interesting question has been raised recently regarding the relationship between the local field potentials (LFPs) and the single unit spiking activity. In this study, we investigate whether a linear modification of the LFPs, acquired from microelectrode recordings inside the subthalamic nucleus (STN) of Parkinson's disease patients, can provide input to an appropriately parameterized Izhikevich model to predict the spikes of an STN neuron. We show that the model is able to predict both the exact timing and the rhythm of the recorded spikes with high accuracy in 5 out of 7 intranuclear single neuron recordings. For the rest of the models, one model shows a lower accuracy in predicting the rhythm and the second one shows a lower accuracy in predicting the timing of the spikes. Overall, the results dictate that the LFPs can reliably predict the occurrence of spikes.