Analysis of nonlocal neural fields for both general and gamma-distributed connectivities

Axel Hutt and Fatihcan M. Atay

Contact the author: Please use for correspondence this email.
Submission date: 13. Jul. 2005
published in: Physica / D, 203 (2005) 1/2, p. 30-54 
DOI number (of the published article): 10.1016/j.physd.2005.03.002
Bibtex
PACS-Numbers: 02.30.Rz, 87.18.Hf
Keywords and phrases: neuronal populations, synaptic connectivity, bifurcation analysis

Abstract:
This work studies the stability of equilibria in spatially extended neuronal ensembles.We first derive the model equation from statistical properties of the neuron population. The obtained integro-differential equation includes synaptic and space-dependent transmission delay for both general and gamma-distributed synaptic connectivities. The latter connectivity type reveals infinite, finite, and vanishing self-connectivities. The work derives conditions for stationary and nonstationary instabilities for both kernel types. In addition, a nonlinear analysis for general kernels yields the order parameter equation of the Turing instability. To compare the results to findings for partial differential equations (PDEs), two typical PDE-types are derived from the examined model equation, namely the general reaction diffusion equation and the Swift-Hohenberg equation. Hence, the discussed integro-differential equation generalizes these PDEs. In the case of the gamma-distributed kernels, the stability conditions are formulated in terms of the mean excitatory and inhibitory interaction ranges. As a novel finding, we obtain Turing instabilities in fields with local inhibition lateral excitation, while wave instabilities occur in fields with local excitation and lateral inhibition. Numerical simulations support the analytical results.