Abstract: Enhanced Neural Variability and Its Effect on Neural Rhythms

Abstract: Enhanced Neural Variability and Its Effect on Neural Rhythms

As complex systems with many interacting components, neural systems exhibit rhythmic activity on multiple timescales, reflecting emergent coherence due to collective neuronal activity.  Some neurons, e.g., mitral cells in the mammalian olfactory system, exhibit mixed-mode oscillations.  That is, they show small-amplitude oscillations in their membrane voltage in addition to the typical large-amplitude ‘spikes’.  The appearance of the small oscillations affect the neurons’ responses to exogenous, noisy inputs and thus the ability of the spike-driven interactions to support coherent rhythms.  We study these effects in a model system of interacting mitral cells using both direct numerical simulation and a standard reduction to phase oscillators, valid for weak noise and weak interaction.  We assess the validity of the reduced model and, going beyond its limited range, examine the effects of stronger noise and interaction.  We find that the noise exhibits a resonance with the small oscillations, causing the variability of spike times to change dramatically with noise strength.  Taking the enhanced variability into account, we are able to extend the validity of the phase reduction model and reproduce over a much larger parameter range the coherent rhythms that appear in the full simulation.