Stochastic resonance in simulated and in vitro hippocampal CA1 cells

Abstract

Stochastic resonance (SR) is a phenomenon whereby the detection of a low-level signal is enhanced in a non-linear system by the introduction of noise, As noise intensity is increased, the SNR increases to a peak value, or resonant point, then decays as the noise intensity reaches high values. Physiological noise in neurons is often ignored or seen as an obstacle in recording, but the discovery of SR in neurons has shown that noise can contribute to neuronal function. A computer model of a hippocampal CA1 cell was developed that introduced synaptic noise in dendrites and enhanced detection of subthreshold synaptic signals using SR. Due to SR effects, synaptic noise reduced the positional dependence of the signal synapse on the dendritic tree. Furthermore, comparison of the noise intensity necessary to produce SR behavior with published noise levels suggests that CA1 cells may reside in a region of the SR curve that maximizes sensitivity to small changes in noise. The model was verified experimentally in rat brain slices.