Evaluating in-vivo spontaneous firing rate in the brain based on neuronal noise modeling.

Even without external stimuli, neurons produce spontaneous bursts of activities. Theoretical and practical clinical considerations, suggest the importance of determining the in-vivo statistical profile of those spontaneous spikes bursts, however this task has not been accomplished yet. Currently, it is only accepted that the in-vivo value of the mean firing rate (λ) of those spontaneous bursts is below 0.1Hz, without knowing its specific value and its population distribution. Here we propose a framework to evaluate the neurons' λ during rest of a given subject, using stochastic signal processing analysis of in-vivo brain fMRI and EEG. Our main hypothesis is that during rest the input to the neurons is mostly formed by a random neuronal noise, and although it fluctuates with zero mean, it affects the neurons' signal output characteristics. Our results based on in-vivo human fMRI and EEG databases, suggest that different people have different and stable characteristic λ values, and that λ of different functional systems of the same subject correlate in their values. Moreover, we find here that the λ values of subjects correlate with their brain task performances, in particular for tasks which are known to be affected by changes in neuronal noise or neuronal excitability threshold.