The impact of complex network structure on seizure-like activity in networks of excitatory and inhibitory neurons

Meeting Details

For more information about this meeting, contact Carina Curto.

Speaker: Chris Kim, NIH

Abstract: When feedback inhibition fails to control recurrent excitation, the activity of neural networks may result in a pathological state, such as in epileptic seizure. We investigate the transition dynamics from a physiological network state to a pathological state, induced by depolarization block in inhibitory neurons. Through simulations of neural networks and analysis of a mean field equation, we characterize the types of bifurcation and propose a possible mechanism for tonic-to-clonic phase transition. Next, we investigate how introducing network motifs to unstructured Erdos-Renyi networks may alter the network dynamics. We find that increasing the variance of in-degree distribution may suppress or facilitate the onset of seizure-like activity, and changes in the covariance of in- and out-degree distributions can amplify such effects. We perform bifurcation analysis on an extended mean field equation to verify the generality of simulation results.

---

Room Reservation Information

Room Number: 106 McAllister

Date: 11/01/2016

Time: 1:30pm - 2:30pm