Author: Yahya Karimipanah

Publisher:

Published: 2016

Total Pages: 256

ISBN-13:

There is mounting evidence that neural networks of the cerebral cortex exhibit scale invariant dynamics. At the larger scale, fMRI recordings have shown evidence for spatiotemporal long range correlations. On the other hand, at the smaller scales this scale invariance is marked by the power law distribution of the size and duration of spontaneous bursts of activity, which are referred as neuronal avalanches. The existence of such avalanches has been confirmed by several studies in vitro and in vivo, among different species and across multiple scales, from spatial scale of MEG and EEG down to single cell resolution. This prevalent scale free nature of cortical activity suggests the hypothesis that the cortex resides at a critical state between two phases of order (short-lasting activity) and disorder (long-lasting activity). In addition, it has been shown, both theoretically and experimentally, that being at criticality brings about certain functional advantages for information processing. However, despite the plenty of evidence and plausibility of the neural criticality hypothesis, still very little is known on how the brain may leverage such criticality to facilitate neural coding. Moreover, the emergent functions that may arise from critical dynamics is poorly understood. In the first part of this thesis, we review several pieces of evidence for the neural criticality hypothesis at different scales, as well as some of the most popular theories of self-organized criticality (SOC). Thereafter, we will focus on the most prominent evidence from small scales, namely neuronal avalanches. We will explore the effect of adaptation and how it can maintain scale free dynamics even at the presence of external stimuli. Using calcium imaging we also experimentally demonstrate the existence of scale free activity at the cellular resolution in vivo. Moreover, by exploring the subsampling issue in neural data, we will find some fundamental constraints of the conventional methods in studying neuronal avalanches. Finally, we show in a computational model that two prevalent features of cortical single-neuron activity, irregular spiking and the decline of response variability at stimulus onset, both are emergent properties of a recurrent network operating near criticality. Our findings establish criticality as a unifying principle for the statistics of single-neuron spiking and the collective behavior of recurrent circuits in cerebral cortex. Moreover, as the observed decline in response variability is regarded as an essential mechanism to enhance response fidelity to stimuli, our discovery of its relation to network criticality offers a starting point toward unraveling the possible roles of critical dynamics in neural coding.