From Structure to Function: Roles of System Level Excitation/Inhibition Balance on Spatio-Temporal Pattern Formation in Neuronal Networks

The brain consists of complex interacting networks of excitatory and inhibitory neurons. The spatio-temporal dynamical patterns in these neural networks are believed to underlie all cognitive functions such as perception, memory formation, etc. Therefore, insights into the mechanisms of the generation of various network-level spatio-temporal dynamics in the context of underlying structural and functional connectivity is essential for understanding how the brain works. These mechanisms themselves depend on global variables characterizing network states, such as relative levels of excitation and inhibition. Here, I combine computational modeling and statistical analysis to investigate changes in neuronal activity patterns and identify the network and cellular mechanisms responsible for those dynamical patterns as a function of changing excitatory/inhibitory (E/I) levels. First, I extend a statistical metric developed in our laboratory, referred to as average minimal distance (AMD), to rapidly quantify the functional connectivity in the network. The metric is able to capture both co-occurrence and causality relationships, providing a more significant correlation value than traditional methods in a much more efficient manner. Combined with a measure of functional network stability (FuNS) which reliably captures the global stability of the functional patterns, I use this AMD-FuNS framework to analyze large-scale in vivo datasets during memory consolidation over extended time periods. Next, I turn to investigate the universal mechanisms underlying the emergence of various functional connectivity patterns as E/I levels are varied in networks composed of excitatory and inhibitory neural populations. I identify multiple E/I balance regimes where E/I balance is similar but network dynamics are heterogeneous. Each balance regime is regulated by the competing interactions between population firing rate and the evolving magnitude of postsynaptic potentials at different synaptic coupling levels. I investigate these patterns E/for networks composed of neurons having different membrane excitability types and find that emerging dynamics and their underlying mechanisms depend critically on neuronal excitability types and network connectivity regimes. Taken together, the theoretical framework presented here provides a mechanistic understanding of the emergence of functional dynamical patterns due to the properties of E/I balance in neural networks.PhDApplied PhysicsUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttps://deepblue.lib.umich.edu/bitstream/2027.42/151598/1/jxwu_1.pd