Nonlinear model of single hippocampal neurons with dynamical thresholds.

Abstract

Neurons transform a series of presynaptic spikes into a series of postsynaptic spikes through a number of nonlinear mechanisms. A nonlinear model with a dynamical threshold was built using a Volterra Laguerre kernel method to characterize the spike train to spike train transformations of hippocampal CA1 pyramidal neurons. Inputs of the model were broadband Poisson random impulse trains with a 2 Hz mean frequency, and outputs of the model were the corresponding evoked post-synaptic potential (PSP) and spike train data recorded from CA1 cell bodies using a whole-cell recording technique. The model consists of four major components, i.e., feedforward kernels representing the transformation of presynaptic spikes to PSPs; a dynamical threshold kernel determining threshold value based on output inter-spike-intervals (ISIs); a spike detector; and a feedback kernel representing the spike-triggered after-potentials.