The critical brain hypothesis suggests that efficient neural computation can be achieved through critical brain dynamics. However, the relationship between human cognitive performance and scale‐free brain dynamics remains unclear. In this study, we investigated the whole‐brain avalanche activity and its individual variability in the human resting‐state functional magnetic resonance imaging (fMRI) data. We showed that though the group‐level analysis was inaccurate because of individual variability, the subject wise scale‐free avalanche activity was significantly associated with maximal synchronization entropy of their brain activity. Meanwhile, the complexity of functional connectivity, as well as structure–function coupling, is maximized in subjects with maximal synchronization entropy. We also observed order–disorder phase transitions in resting‐state brain dynamics and found that there were longer times spent in the subcritical regime. These results imply that large‐scale brain dynamics favor the slightly subcritical regime of phase transition. Finally, we showed evidence that the neural dynamics of human participants with higher fluid intelligence and working memory scores are closer to criticality. We identified brain regions whose critical dynamics showed significant positive correlations with fluid intelligence performance and found that these regions were located in the prefrontal cortex and inferior parietal cortex, which were believed to be important nodes of brain networks underlying human intelligence. Our results reveal the possible role that avalanche criticality plays in cognitive performance and provide a simple method to identify the critical point and map cortical states on a spectrum of neural dynamics, ranging from subcriticality to supercriticality.

While functional MRI (fMRI) studies have mainly focused on gray matter, recent studies have consistently found that blood-oxygenation-level-dependent (BOLD) signals can be reliably detected in white matter, and functional connectivity (FC) has been organized into distributed networks in white matter. Nevertheless, it remains unclear whether this white matter FC reflects underlying electrophysiological synchronization. To address this question, we employ intracranial stereotactic-electroencephalography (SEEG) and resting-state fMRI data from a group of 16 patients with drug-resistant epilepsy. We find that BOLD FC is correlated with SEEG FC in white matter, and this result is consistent across a wide range of frequency bands for each participant. By including diffusion spectrum imaging data, we also find that white matter FC from both SEEG and fMRI are correlated with white matter structural connectivity, suggesting that anatomical fiber tracts underlie the functional synchronization in white matter. These results provide evidence for the electrophysiological and structural basis of white matter BOLD FC, which could be a potential biomarker for psychiatric and neurological disorders.

The critical brain hypothesis suggests that efficient neural computation could be realized by critical brain dynamics hallmarked by scale-free avalanche activity. However, its further application requires not only accurately identifying the critical point but also depicting the phase transition in brains so that different cognitive states could be mapped on a spectrum. In this work, we mapped individuals' brains onto an inverted-U curve between the mean synchronization and synchronization entropy of blood oxygenation level-dependent (BOLD) signals from resting-state fMRI scans. We found that the critical point lies at the tipping point (i.e., moderate synchrony and maximal variability in synchrony) of this curve, which is consistent with previous findings. We then verified that the complexity of functional connectivity, as well as the similarity between structural and functional networks, was maximized near the critical point, whereas reduction in complexity and structure-function decoupling were found both in the sub- and supercritical regimes. We then observed phase transitions in resting-state brain dynamics and found that the brains showed longer dwell times in the subcritical regime. These results provided strong evidence that the large-scale brain networks were hovering around the critical point. Finally, we found that critical dynamics were associated with high scores in fluid intelligence and working memory tests but not crystallized intelligence scores. Our results revealed the role that avalanche criticality plays in cognitive performance and provide a simple method to identify the critical point and map cortical states on a spectrum of neural dynamics, with a critical point in the domain.

No abstract