Critical Dynamics in Balanced Excitable Networks: Neuronal Avalanches, Dynamic Range, and Ceaseless Activity

When: Tuesday, February 05, 2013 at 3:00 pm
Where: DA 5th fl
Speaker: Daniel Larremore
Organization: Center for Communicable Disease Dynamics, Department of Epidemiology, Harvard School of Public Health
Sponsor: CCNR Seminar

The neuronal network that makes up the mammalian cortex includes a mix of approximately 80% excitatory and 20% inhibitory neurons. The balance between excitation and inhibition is of interest to theoreticians as well as experimentalists, who have conducted fascinating in-vitro experiments usingslices of rat cortex in the presence of various drugs. They find that the cortex appears to operate in state of balanced excitation and inhibition. In this balanced state, the cortex operates at the critical point of a phase transition, which maximizes its dynamic range, and produces avalanches of activity whose size are duration are distributed according to power laws.

We explore the dynamics of this balanced state using a simple model for individual neurons, analyzing the collective dynamics that arise when many such neurons are coupled in a complex network. By characterizing network dynamics using spectral properties of the network adjacency matrix, we examine the effect of network structure on dynamic range and critical avalanches. We then extend this work to explicitly include inhibitory neurons, finding that our model corresponds well to in-vivo recordings from the premotor cortex of awake monkeys. Surprisingly, the inclusion of inhibitory neurons has the counterintuitive effect of guaranteeing ceaseless network activity, providing a novel solution to a longstanding debate about the source of ceaseless activity in the mammalian cortex.

This work is part of an ongoing collaboration with Juan G. Restrepo, Francesco Sorrentino, Woodrow L. Shew, and Ed Ott.