A novel density-based neural mass model for simulating neuronal network dynamics with conductance-based synapses and membrane current adaptation (Q6055081)

The aim in this article is to develop a new low-dimensional mean-field model based on adaptive exponential integrate-and-fire (aEIF) neuronal dynamics. The new density-based neural mass model (dNMM) is developed for simulating network dynamics of aEIF neuronal populations with biologically conductance-based synaptic interactions and adaptive behaviors of firing rate responses. In the second section of the article, the spiking cortical network model and the density-based neural mass model are presented. One considers a local cortical network of N = 10000 neurons of excitatory pyramidal (PY) cells -- eighty percent- and inhibitory interneurons (IN) for the rest. Each neuron received recurrent inputs from 400 excitatory PY neurons and 100 inhibitory IN neurons. The dynamics of each neuron was described by the aEIF dynamics and the dynamic equations are shown. The derivation of the dNMM is based on the so-called colored-synapse population density method (csPDM) developed by the authors in a previous article [eNeuro 5, No. 6, Article ID ENEURO.0002-18.2018, 21 p. (2018; \url{doi:10.1523/ENEURO.0002-18.2018})]. csPDM is a dimension-reduced probabilistic description of neuronal population dynamics. The evolution of the new defined density function \(\rho_v\)(V,t) of the distribution of the membrane potentials within the population is described by a partial differential equation. This equation together with some ordinary differential equations describing the dynamics of aEIF neurons build the dynamics of csPDM. These equations are further transformed into the equations building the dNMM dynamics. Simulations of population dynamics trough the dNMM has been compared to numerical simulations by classical NMM. Experiments on Gaussian distributions of membrane potentials under non-diffusive limits of synaptic dynamics , dependence of steady output firing razes on the mean and standard deviation of membrane potentials, performance of dNMM in steady state firing rate responses and in dynamic firing rate responses have been performed. Discussions on obtained results and conclusions can be found in the last two sections of the article.