Brain Networks May Sit at a Critical Edge That Shapes Thought, Seizures and Consciousness

Neuroscience has long wrestled with a deceptively simple question: how does the brain stay stable enough to keep functioning, yet flexible enough to learn, adapt and generate complex thought? A growing research program argues that the answer may lie in a concept borrowed from physics—criticality, the special dynamical regime that sits at the boundary between order and chaos. A new editorial in Frontiers in Computational Neuroscience, written by Peter Erdi of Kalamazoo College and the HUN-REN Wigner Research Centre for Physics and Tiago L. Ribeiro of the University of Arkansas, gathers eight studies that test this idea across scales, from single neurons to whole brains and even hospital intensive care units.

The brain criticality hypothesis holds that neural systems operate in or near a critical state—a unique dynamical regime separating qualitatively different behaviors such as ordered stability and disordered chaos. According to the editorial, operating near this phase transition is theorized to optimize crucial information processing properties, including sensitivity to inputs, dynamic range, information transmission, and flexible multi-scale coordination.

The idea is not new, but its reach has expanded steadily. The editorial traces the expansion of the hypothesis to early pioneering studies on scale-free neuronal avalanches (Beggs and Plenz, 2003) and theoretical formulations using neuropercolation models (Kozma et al., 2005; Kozma and Freeman, 2006). Since then, the hypothesis has attracted both enthusiasm and skepticism, and the editors are candid about why. What constitutes definitive evidence for criticality in empirical datasets? How much do the statistical signatures of criticality depend on the measurement and analysis techniques used to detect them? And, crucially, how can these principles be leveraged to understand complex brain dynamics across microscopic and macroscopic scales—and can they offer clinically meaningful prognostic utility?

Those questions frame the eight articles collected in the Research Topic, which was curated to unite diverse theoretical, computational, and empirical perspectives and to map how collective neural dynamics emerge from multi-scale interactions to govern robust brain function. The editors describe the collection as an effort to bridge the gap between abstract physics frameworks and functional neurobiology, and the individual contributions tackle that bridge from several directions: multi-scale modeling, network topology, information theory, measurement methodology, biophysics and clinical translation.

One persistent obstacle in computational neuroscience has been the integration of decoupled spatial scales. Addressing this challenge, Hater and colleagues introduce Arbor-TVB, a co-simulation framework that dynamically couples biologically detailed neuronal simulations with whole-brain network dynamics. By executing multi-scale modeling of seizure generation and whole-brain propagation, their work demonstrates how microscopic instabilities can trigger macroscopic dynamics.

Complementing that modeling effort, Cooray brings a physical lens to large-scale brain dynamics by applying wave turbulence theory to evaluate multi-scale spectral dynamics in magnetoencephalography and electroencephalography recordings. The study maps out how power-law spectral features emerge from nonlinear wave interactions, linking continuous energy cascades to transitions between distinct macroscopic cortical states.

To remain functional, networks must actively maintain operations at the edge of chaos without succumbing to runaway activity. Richner and colleagues address this operational requirement from a network perspective, demonstrating that static excitation-inhibition balance alone cannot maintain stability. Instead, they show how dynamic negative-feedback mechanisms, such as spike-frequency adaptation and short-term synaptic depression, continuously modulate effective connectivity to stabilize networks under shifting inputs.

The topological layout of a network dictates its avalanche dynamics, and de Candia and colleagues explore this relationship in a stochastic Wilson–Cowan model featuring a modular network architecture. Their findings reveal two distinct critical lines separating symmetric and symmetry-broken activity states, demonstrating that modular systems generate highly distinct, asymmetric avalanche shapes that mirror experimental observations.

Methodological caution forms another pillar of the collection. Understanding the exact computational benefits of criticality requires advanced mathematical tools, and Rajpal and colleagues investigate information transmission in the visual cortex of awake mice using partial information decomposition. They show that long-range correlations near criticality are driven by an enhancement of synergistic interactions during sensory stimulation, illustrating how synergy and redundancy work in tandem to optimize information sharing across distributed populations. In parallel, Li and colleagues isolate a subtler problem: inferring criticality accurately requires careful navigation of filtering artifacts. By testing phenomenological renormalization group analyses on a network with fixed dynamics, they systematically demonstrate that temporal binning, measurement nonlinearities, and signal deconvolution can heavily distort scaling exponents—providing a cautionary guide for separating actual biological dynamics from measurement artifacts.

The collection closes with its most ambitious reaches. Hunt advances a hypothesis that bridges electromagnetic field interactions, ephaptic coupling, and global neural integration. By framing neural avalanches as field resonance cascades, this work offers a unified biophysical mechanism that simultaneously accounts for critical dynamics and the binding problem of conscious experience. On the clinical side, Newman and colleagues investigate whether criticality-related EEG measures can serve as behavior-independent prognostic markers for sedated pediatric intensive care patients. They establish that higher pre-stimulus signal complexity, entropy, and fractal structures correlate with successful functional recovery, showcasing a diagnostic pipeline derived directly from complex systems theory. Together, the editors argue, these contributions demonstrate that criticality is far more than an abstract physics analogy; it is a framework for linking structural connectivity, information theory and clinical reality under the umbrella of brain dynamics—while the open questions about evidence standards and measurement artifacts remain very much alive.

Subject of Research: Whether neural systems operate near a critical state between order and chaos, and how this shapes brain dynamics, information processing and clinical prognosis

Article Title: Editorial: Exploring the role of criticality in neural systems

Article References: Erdi, P., & Ribeiro, T. L. (2026). Editorial: Exploring the role of criticality in neural systems. Frontiers in Computational Neuroscience, 20, Article 2036015. https://doi.org/10.3389/fncom.2026.2036015

Image Credits: AI Generated

DOI: 10.3389/fncom.2026.2036015

Keywords: brain criticality, neural avalanches, phase transition, whole-brain modeling, EEG biomarkers, neural networks, information theory, homeostatic plasticity, wave turbulence, consciousness, pediatric intensive care, computational neuroscience