Reliability, discriminability and stochastic synchronization of olfactory neurons

Combining computer simulations and electrophysiological experiments we have studied the effect of noise on the responses of olfactory neurons. In particular, we first investigated the reliability of mitral cell responses and found, as previously observed in other neural systems, that, in the presence of background noise, mitral cells reliably respond to fast fluctuating inputs but not to constant stimuli. We then investigated a related property, input discriminability and a closely related phenomenon, stochastic synchronization that may account for the synchronous firing of mitral cells leading to network oscillations in the beta/gamma frequency range. We argue that these phenomena: reliability, discriminability and stochastic synchronization are not exclusive to neurons but are rather common to all devices with a resetting threshold. Therefore we suggest that an artificial nose with a hardware implementation of such devices may optimally operate with low signal-to-noise ratios.