The dynamics of recurrent inhibition (Q793653)

The authors present ''a heuristic model for the dynamics of recurrent inhibition''. This last term refers to a well-known phenomenon in neuronal activity, in which an excitation quantified as an action potential E(t) of some cell produces - after some time \(\tau\) (t) - on this same cell an inhibition \(I(t+\tau(t)).\) Basing on some physico-chemical mechanisms and assumptions concerning the delay and also a threshold effect the authors end with an equation of the type: \(dI/dt=-\gamma I(t)+G((E(t-\tau)-I(t-\tau)-\theta)^+)\). Here E(t) plays the role of a forcing term which, for further study, is taken as a constant: \(E(t)=E\); G is a non-monotone saturation function. The theoretical results concern existence and multiplicity of steady states and their local stability, using Hayes' theorems, depending on E. (For some values of E, there are 2 stable steady states). For a more complete study by numerical treatments, a particular neural complex of the hippocampus has been considered, and variations of a certain parameter T (such variations can be performed experimentally by means of penicillin) have been tested. According to the authors they provide ''a surprising sequence of behaviours...'', ''exactly analogous to the patterns of multiple period doubling bifurcation...''.