In biological sensory systems, a presence of noise can actually enhance detection of weak signals. This phenomenon is called stochastic resonance (SR). We show that SR can emerge as a collective phenomenon in neural networks. We consider a cortical circuit model composed by stochastic excitatory and inhibitory neurons that form a sparsely connected network. We find that SR appears due to nonlinear dynamics in a region near the critical point of a dynamical phase transition to network oscillations. The critical point is actually an emergent threshold in the collective dynamics. Using the cortical model, we mimic experiments of Gluckman et al. [B. J. Gluckman et al., PRL 77, 4098 (1996)] that observed stochastic resonance in a response of CA1 networks from mammalian brain on periodic electric stimuli. Results of our numerical calculations are in agreement both qualitatively and quantitatively with these experiments.

This content is only available via PDF.
You do not currently have access to this content.