Appropriate encoding of information is a major task in the nervous system. Neural mechanisms by which sensory inputs are first encoded and later decoded in the central nervous system have to take into account the presence of considerable uncorrelated background """"""""noise"""""""". To date there is no clear understanding on how the nervous system efficiently improves the signal to noise ratio. A recent development in the field of bistable systems shows that under specific circumstances an increase in the system's input noise can lead to a decrease in the output noise. In this phenomena known as stochastic resonance, the output signal of a noisy bistable system can be modulated in time by applying (additive or multiplicative) a weak external periodic forcing. Preliminary evidence shows that similar phenomena occurs in neuron models. Using this novel theoretical argument this project will examine experimental evidence in support of the hypothesis that noise plays an active role in neuronal encoding. The project will explore several quantitative signatures of stochastic resonance on the response of sensory neurons to periodic stimuli. Estimates of these measures are taken at the cellular and at the system level (in different levels of the neuraxis, from primary afferent fibers to cortex) to clarify the role of stochastic resonance in neuronal encoding of sensory information. Theoretical studies of neurons' generic and ionic models are directed to formulate and characterize measures of stochastic resonance in excitable systems and to explore how rudimentary networks working at the range of stochastic resonance might """"""""decode"""""""" the information embedded in the spike train. These theoretical and experimental results should show that noise in the nervous system is actively controlled and modulated in order to encode sensory information. Overall, the results will provide a better understanding of somatosensory neural encoding.
