Signatures of critical dynamics in the rat hippocampus

Theoretical studies show that, operating at criticality maximises both the information processing and storage capabilities of neural systems. Experimental indications that the brain might operate in this way come in the form of: (1) power law distributions of avalanches and (2) non-Gaussian fixed points that appear in phenomenological renormalization group analysis. Despite theoretical benefits and experimental findings, it remains highly debated whether neural criticality is functionally relevant. Firstly, it is hard to confirm if avalanches follow true power laws. Secondly, recent studies indicate that experimentally observed properties of avalanches and coarse-grained activity can be explained by a handful of latent factors. To investigate potential links between criticality and behaviour, we analysed the activity of simultaneously recorded hippocampal CA1 neurons from freely moving rats. We monitored avalanches, estimated the branching parameter, and applied the renormalization group approach, in awake periods and sleep / rest. Distributions of avalanche duration and size followed power laws with exponents close to α ≈ 1.75 and τ ≈ 1.55, across all experimental conditions. During sleep / rest periods, avalanche shapes collapsed with a scaling exponent close to the value that characterises the relation between sizes and durations. Applying the phenomenological renormalization group method indicated a scaling of silence periods, activity variance and correlation times, as well as a collapse of activity distributions. However, we also observed behavioural modulation of the branching parameter. In summary, we observed robust measures of criticality in the hippocampal network, indicating stable collective dynamics finely tuned to behavioural demands.