Investigating the effects of arousal on auditory processing

Animals’ behavioral states vary rapidly over time, and these changes can drastically affect stimulus processing and task performance in different cortical areas. However, the neural mechanisms driving state-dependent effects on sensory processing are unclear. Here, we first present preliminary data from auditory cortex suggesting that tone frequency may be best decoded from population activity during states of moderate arousal. To explain this finding, we then consider a spiking model of a cortical circuit in which cells are arranged in clusters. Stimuli are modeled as cluster-specific-inputs, and state-changes are modeled as perturbations of the baseline external current to the network. We show that perturbations that increase external input heterogeneity modulate stimulus decoding in a way that is consistent with the effect of arousal in auditory cortex, with intermediate perturbation strengths yielding the best performance. We also demonstrate that these variations in stimulus processing arise due to a perturbation-induced shift from a phase in which network activity switches between metastable attractors to a phase with only a single attractor. In this picture, we show that moderate perturbation strengths improve decoding by speeding up transitions between metastable states.