Closer to critical resting-state neural dynamics in individuals with higher fluid intelligence

Abstract

According to the critical brain hypothesis, the brain is considered to operate near criticality and realize efficient neural computations. Despite the prior theoretical and empirical evidence in favor of the hypothesis, no direct link has been provided between human cognitive performance and the neural criticality. Here we provide such a key link by analyzing resting-state dynamics of functional magnetic resonance imaging (fMRI) networks at a whole-brain level. We develop a novel data-driven analysis method, inspired from statistical physics theory of spin systems, to map out the whole-brain neural dynamics onto a phase diagram. Using this tool, we show evidence that dynamics of more intelligent human participants are closer to a critical state, i.e., the boundary between the paramagnetic phase and the spin-glass (SG) phase. This result was specific to fluid intelligence as opposed to crystalized intelligence. The present results are also consistent with the notion of "edge-of-chaos" neural computation. Author summaryAccording to the critical brain hypothesis, the brain should be operating near criticality, i.e., a boundary between different states showing qualitatively different dynamical behaviors. Such a critical brain dynamics has been considered to realize efficient neural computations. Here we provide direct neural evidence in favor of this hypothesis by showing that the brain dynamics in more intelligent individuals are closer to the criticality. We reached this conclusion by deploying a novel dataanalysis method based on statistical-physics theory of Ising spin systems to functional magnetic resonance imaging (fMRI) data obtained from human participants. Specifically, our method maps multivariate fMRI data obtained from each participant to a point in a phase diagram akin to that of the Sherrington-Kirkpatrick model of spin-glass systems. The brain dynamics for the participants having high fluid intelligence scores tended to be close to the phase boundary, which marks criticality, between the paramagnetic phase and the spin-glass phase, but not to the boundary between the paramagnetic and ferromagnetic phases.

Recovery of neural dynamics criticality in personalized whole brain models of stroke

“…Ezaki et. al provided empirical support that subjects with higher-IQ have neural dynamics closer to criticality than subjects with lower-IQ participants 78 . Our findings show for the first time that causal manipulations of brain activity, through lesions and recovery, modifies criticality in a significant behavioral manner.…”

Dynamics of Long-Range Temporal Correlations in Broadband EEG During Different Motor Execution and Imagery Tasks

“…The LRTCs have been observed in the alpha, beta, theta oscillation amplitude envelopes (Berthouze et al, 2010), alpha oscillation phase (Botcharova et al, 2014), broadband phase synchrony (Kitzbichler et al, 2009), avalanches (Benayoun et al, 2010;Palva et al, 2013), and energy profile (Parish et al, 2004;Benayoun et al, 2010). It is commonly postulated in the literature that the power-law behavior and LRTCs occur because the brain operates at criticality (Poil et al, 2012;Massobrio et al, 2015), thus optimizing information storage capacity (Kitzbichler et al, 2009) and enabling quick adaptation to the cognitive processing demands (Ezaki et al, 2020;Ouyang et al, 2020;Zimmern, 2020). In the absence of long-range temporal correlations, correlations on shorter timescales lead to reduction in the ability to integrate information (for example, during certain stages of sleep; Meisel et al, 2017).…”

Functional connectivity of EEG is subject-specific, associated with phenotype, and different from fMRI

“…These attempt to capture different time scales ( Bullmore et al, 2004 ), reduce common sources of variance ( Salvador et al, 2005 ), capture delayed correlations ( Kitzbichler et al, 2009 ), or capture causation ( Reid et al, 2019 ). Some of them have also been linked to phenotypic variables such as age ( Meunier et al, 2009 ; Mowinckel et al, 2012 ), fluid intelligence ( Ezaki et al, 2019 ), and schizophrenia ( Fornito et al, 2012 ), and different methods for correlating with phenotype have been benchmarked ( Dadi et al, 2019 ). However, the vast majority of work in this regard is based on Pearson’s correlation, which we will use in the current analysis.…”