Multimodality in a Dynamic Feedback Network with Stochastic Input

Different rhythmic activities in CA1 characterize the neuronal correlate of several behavioral states. Recently, in an in vitro preparation of whole hippocampus, generation of spontaneous slow rhythms, similar to in vivo hippocampal EEGs, has been reported. We previously proposed a dynamical feedback mechanism where populations of synchronized interneurons entrain spatially dynamic subpopulations of pyramidal cells. Here, using a stochastic phenomenological model, we show that the rhythm is inherently multimodal and that the number of excitatory cells as well as the excitability of the interneurons allow the network to differentially amplify/suppress the modal structure of the emergent rhythm