By Adrian Ponce Alvarez et al. (Pompeu Fabra University, Barcelona, Spain)
How does the local collective activity of brain regions determine the global brain state related to consciousness? To answer this question, we modeled loss of consciousness using anesthetic agents in macaque monkeys while recording the collective activity of brain regions using functional MRI. We showed that the coupling between each brain region and the rest of the cortex provides an efficient statistic that distinguished the two brain states: wakefulness and anesthesia-induced loss of consciousness. Based on this and other statistics, we estimated maximum entropy models to derive collective, macroscopic properties describing the different brain states. These properties quantify the system’s capabilities to produce work, to contain information and to transmit it, which we found to be all maximized in the awake state. We showed that the differences in these collective properties were consistent with a phase transition from critical dynamics in the awake state to supercritical dynamics in the anesthetized state. Moreover, we used information theory to detect the model parameters whose variations produce the strongest effect on the collective activity. Using this approach, we showed that changes in brain state and in state of consciousness primarily depended on changes in population couplings of insular, cingulate, and parietal cortices. In conclusion, our findings suggest that the phase transition underlying the loss of consciousness is predominantly driven by the uncoupling of specific brain regions from the rest of the network.