Theoretical foundations of studying criticality in the brain

Yang Tian, Zeren Tan, Hedong Hou, Guoqi Li, Aohua Cheng, Yike Qiu, Kangyu Weng, Chun Chen, Pei SunNetwork 2022 · DOI

TL;DR

This paper provides a much-needed theoretical grounding for studying criticality in the brain, clarifying different types of criticality (OC, qC, SOC, SOqC) and the statistical methods used to identify them. By bridging the gap between complex physics theories and neuroscience, it aims to resolve ongoing controversies and improve the reliability of brain criticality research.

What Problem Does This Paper Address?

The "brain criticality hypothesis" suggests that the brain operates at a critical state, which is optimal for information processing. However, this idea is plagued by controversies and misunderstandings. Researchers struggle with:

  • Defining criticality: There are multiple types of criticality (ordinary, self-organized, quasi-critical) that are often conflated.
  • Identifying criticality in data: Statistical methods for detecting criticality in neural data are complex and prone to errors.
  • Misconceptions about benefits: The functional advantages of criticality are often overgeneralized or misunderstood.

These issues stem from the immaturity and complexity of the underlying physical theories and statistical techniques. This paper aims to demystify these foundations, making them more accessible and providing a clearer framework for research.

Methods

The authors present a systematic review and reformulation of the theoretical foundations of brain criticality. Key methods include:

  • Categorization of criticality: They clearly define and differentiate four main types of brain criticality:
  • Ordinary Criticality (OC): Requires external "fine-tuning" to reach a critical state.
  • Quasi-Criticality (qC): Exhibits near-critical behavior, often due to external influences like stimulus intensity.
  • Self-Organized Criticality (SOC): The system naturally evolves to a critical state through internal feedback mechanisms.
  • Self-Organized Quasi-Criticality (SOqC): A combination where internal mechanisms drive the system towards a quasi-critical state, requiring some external fine-tuning to reach true criticality.
  • Accessible explanations of physical theories: They break down complex concepts like directed percolation, conserved directed percolation, and Langevin equations, explaining how they apply to neural dynamics.
  • Step-by-step derivations: The paper provides derivations to characterize neural dynamics as physical systems exhibiting avalanches.
  • Summary of statistical techniques: They review common methods for estimating criticality from empirical data, highlighting their limitations and potential pitfalls.
  • Focus on neural avalanches: A significant portion of the discussion centers on neural avalanches, which are considered key signatures of criticality.

Datasets

Not discussed in the paper. The paper focuses on theoretical foundations and statistical techniques, not on analyzing specific experimental datasets.

Key Insights and Ideas

  • Criticality is not monolithic: The paper clarifies that "criticality" in the brain isn't a single phenomenon but a family of related states (OC, qC, SOC, SOqC), each with distinct underlying mechanisms and properties.
  • Bridging physics and neuroscience: The authors successfully translate complex statistical physics concepts into neuroscience terminology, making them more digestible for researchers in the field.
  • Avalanches are necessary but not sufficient: While neural avalanches are strong indicators of criticality, their mere presence doesn't automatically confirm it. The underlying mechanisms and adherence to scaling relations are crucial.
  • Statistical analysis is error-prone: Estimating exponents, scaling relations, and collapse shapes from empirical data is fraught with challenges, requiring careful application of robust statistical methods.
  • Non-equilibrium is key: The brain, being a biological system, operates far from equilibrium, and understanding non-equilibrium criticality is essential for studying brain dynamics.

Key Results

The paper's primary contribution is not a set of quantitative experimental results, but rather a conceptual and theoretical framework. The key "results" are:

  • A unified classification: A clear framework for understanding and differentiating the four main types of brain criticality.
  • Accessible theoretical foundations: Simplified explanations and derivations of the physical theories underpinning brain criticality.
  • Identification of common pitfalls: A detailed summary of error-prone aspects in current brain criticality analysis and statistical estimation.
  • Proposed solutions and future directions: Recommendations for improving theoretical models and statistical techniques, and for future research avenues.

Limitations and Future Work

The authors acknowledge several limitations and suggest future directions:

  • Theoretical complexity: While they aim for accessibility, the underlying physics theories are inherently complex, and further simplification and systematic reformulation are needed.
  • Experimental verification challenges: Precisely verifying specific universality classes in empirical data remains difficult.
  • Statistical method imperfections: Current statistical techniques for estimating and verifying criticality are still imperfect and error-prone.
  • Synchronous phase transitions: A complete theoretical framework for synchronous phase transitions in the brain is still lacking, requiring more exploration.
  • Integration with biological realism: Future work should aim to integrate more realistic biological details into criticality theories and establish stronger connections with experimental observations.
  • Toolbox development: The authors advocate for open-source toolboxes to standardize and facilitate brain criticality analysis.
  • Data sharing: The creation of open-source, multi-species, and large-scale datasets of neural dynamics is crucial for validating statistical techniques and establishing baselines.
  • Unified criticality theory: Exploring whether a unified criticality theory can characterize both brain and artificial neural networks is an exciting future direction.

Summary updated

Abstract

Criticality is hypothesized as a physical mechanism underlying efficient transitions between cortical states and remarkable information-processing capacities in the brain. While considerable evidence generally supports this hypothesis, nonnegligible controversies persist regarding the ubiquity of criticality in neural dynamics and its role in information processing. Validity issues frequently arise during identifying potential brain criticality from empirical data. Moreover, the functional benefits implied by brain criticality are frequently misconceived or unduly generalized. These problems stem from the nontriviality and immaturity of the physical theories that analytically derive brain criticality and the statistic techniques that estimate brain criticality from empirical data. To help solve these problems, we present a systematic review and reformulate the foundations of studying brain criticality, that is, ordinary criticality (OC), quasi-criticality (qC), self-organized criticality (SOC), and self-organized quasi-criticality (SOqC), using the terminology of neuroscience. We offer accessible explanations of the physical theories and statistical techniques of brain criticality, providing step-by-step derivations to characterize neural dynamics as a physical system with avalanches. We summarize error-prone details and existing limitations in brain criticality analysis and suggest possible solutions. Moreover, we present a forward-looking perspective on how optimizing the foundations of studying brain criticality can deepen our understanding of various neuroscience questions.
BibTeX
@inproceedings{tian2022theoretical,
  author = {Yang Tian and Zeren Tan and Hedong Hou and Guoqi Li and Aohua Cheng and Yike Qiu and Kangyu Weng and Chun Chen and Pei Sun},
  title = {{Theoretical foundations of studying criticality in the brain}},
  booktitle = {Network},
  year = {2022},
  doi = {10.1162/netn\_a\_00269},
}
Figure