Neural Networks Are Tuned Near Criticality During a Cognitive Task and Distanced from Criticality In a Psychopharmacological Model of Alzheimer’s Disease

Abstract

AbstractDynamical systems that exhibit transitions between ordered and disordered states are described as “critical” when the system is at the borderline between these states. The ability of criticality to explain a variety of brain properties, including optimal information processing, makes it of considerable interest to investigate whether thesein vivonetworks display critical behaviour, and whether some forms of cognitive impairment such as dementia might display altered critical behaviour. To investigate these questions, the activity of several hundred hippocampal CA1 neurons in freely-behaving mice was studied with miniscope widefield calcium imaging during rest, a novel object recognition task, and novel object recognition after administration of the amnesic drug scopolamine which acts as a psychopharmacological model of Alzheimer’s disease. Utilizing rigorous metrics of criticality including the Deviation from Criticality Coefficient and Branching Ratio, the ensemble neural activity in the hippocampus was observed to display evidence of near-critical behaviour during rest periods, but moved significantly closer to a critical state when engaged in a cognitive task. The dynamics were observed to move significantly away from criticality during the cognitive task after scopolamine administration. In contrast to previous theoretical predictions, our results indicate that hippocampus neural networks move closer to criticality under cognitive load, and that critical dynamical regimes produce a near-optimal state for cognitive operations.