May the 4C's be with you: an overview of complexity-inspired frameworks for analysing resting-state neuroimaging data

Fran Hancock, Fernando E. Rosas, Pedro A. M. Mediano, Andrea I. Luppi, Joana Cabral, Ottavia Dipasquale, Federico TurkheimerJournal of The Royal Society Interface 2022 · DOI

TL;DR

This paper argues that complexity science offers a unified framework for understanding spontaneous brain activity, by reviewing four key concepts: connectivity, computation, criticality, and coherence (the "4C's"). It proposes "metastability" as a common underlying principle that links these concepts, providing a more integrated approach to analyzing neuroimaging data than existing domain-specific models.

What Problem Does This Paper Address?

Neuroscience has developed various models to understand how the brain coordinates its activity, especially when it's not actively engaged in a task (resting-state). These models often come from different scientific fields like physics, mathematics, and psychology, leading to a confusing landscape of terminology and potentially conflicting interpretations. This makes it difficult to build a cohesive understanding of brain function. The paper aims to bridge this gap by reviewing and clarifying key concepts from complexity science that are relevant to analyzing resting-state neuroimaging data, particularly fMRI. It seeks to demonstrate how these concepts can provide a more principled and integrated way to describe and understand the brain's spontaneous activity.

Methods

The paper doesn't introduce new experimental methods but rather synthesizes existing research through a complexity science lens. It reviews and clarifies four core concepts:

  • Connectivity: Moving beyond simple functional connectivity (statistical relationships between brain regions), it discusses structural connectivity (physical connections) and effective connectivity (causal influences), highlighting how graph theory can be used to analyze these networks.
  • Computation: It explores "integrated information" (Φ) and "partial information decomposition" (PID) as ways to quantify how systems process information, suggesting that synergistic information flow is crucial for higher cognitive functions.
  • Criticality: The paper discusses both statistical criticality (systems exhibiting power-law statistics) and dynamical criticality (systems near a phase transition), suggesting that the brain might operate near a critical state to optimize information processing.
  • Coherence and Synchronization: It examines various measures of how brain signals align in phase, including phase synchrony and phase-locking, and how these relate to communication between brain regions.

A central theme is the concept of metastability, which the authors propose as a unifying principle that describes the brain's ability to maintain a balance between segregation (specialization) and integration (global coordination) of its activity.

Datasets

The paper analyzes and synthesizes findings from a wide range of existing neuroimaging research, primarily using functional magnetic resonance imaging (fMRI) data from resting-state conditions. It does not introduce new datasets or benchmarks.

Key Insights and Ideas

  • The "4C's" as a Unified Framework: By bringing together connectivity, computation, criticality, and coherence, the paper offers a structured way to think about complex brain dynamics, moving beyond isolated theories.
  • Metastability as the Common Denominator: The authors propose that metastability—the ability of a system to fluctuate between different stable states—is a fundamental property that underpins all four "C's" and is crucial for healthy brain functioning.
  • Complexity Science as a Bridge: Complexity science provides a powerful set of tools and concepts that can bridge the gap between different neuroscientific disciplines, fostering a more integrated understanding of the brain.
  • Beyond Simple Correlation: The review emphasizes the need to move beyond basic statistical associations (like correlation) in functional connectivity to more sophisticated measures that capture dynamic and information-theoretic properties.
  • Brain's "Idle" State is Highly Active: The paper reinforces the idea that resting-state brain activity is not a passive state but a complex, dynamic process crucial for cognition.

Key Results

The paper doesn't present new quantitative results but rather synthesizes findings from numerous studies. Key qualitative findings include:

  • Healthy brain networks exhibit complex properties like small-worldness, rich-club organization, and fractal structures.
  • Synergistic information processing, quantified by measures like ΦR, is linked to higher cognitive functions and is more prominent in evolutionarily expanded brain regions.
  • Evidence suggests that the brain operates near a critical state, which may optimize its responsiveness and information processing capabilities.
  • Metastability is observed across various brain states (rest, sleep, consciousness, psychedelic states) and is correlated with cognitive flexibility and intelligence.
  • The study of phase relationships (synchrony and phase-locking) reveals complementary information about brain dynamics.

Limitations and Future Work

  • Terminology Challenges: The authors acknowledge that the multidisciplinary nature of complexity science can lead to inconsistent terminology and potential misunderstandings within neuroscience.
  • Measuring Metastability: While metastability is proposed as a key concept, its precise physical underpinnings and optimal quantification remain areas for further investigation.
  • Empirical Validation: While the paper synthesizes existing research, it highlights the need for continued empirical validation of these complexity-based frameworks in diverse neuroimaging studies.
  • Bridging Scales: Future work could focus on better understanding how these mesoscale complexity principles relate to microscale neuronal activity and macroscale behavior.
  • Clinical Applications: Further research is needed to explore how these complexity-inspired frameworks can be applied to understand and diagnose neurological and psychiatric disorders.

Summary updated

Abstract

Competing and complementary models of resting-state brain dynamics contribute to our phenomenological and mechanistic understanding of whole-brain coordination and communication, and provide potential evidence for differential brain functioning associated with normal and pathological behaviour. These neuroscientific theories stem from the perspectives of physics, engineering, mathematics and psychology and create a complicated landscape of domain-specific terminology and meaning, which, when used outside of that domain, may lead to incorrect assumptions and conclusions within the neuroscience community. Here, we review and clarify the key concepts of connectivity, computation, criticality and coherence-the 4C's-and outline a potential role for metastability as a common denominator across these propositions. We analyse and synthesize whole-brain neuroimaging research, examined through functional magnetic imaging, to demonstrate that complexity science offers a principled and integrated approach to describe, and potentially understand, macroscale spontaneous brain functioning.
BibTeX
@article{hancock2022may,
  author = {Fran Hancock and Fernando E. Rosas and Pedro A. M. Mediano and Andrea I. Luppi and Joana Cabral and Ottavia Dipasquale and Federico Turkheimer},
  title = {{May the 4C's be with you: an overview of complexity-inspired frameworks for analysing resting-state neuroimaging data}},
  journal = {Journal of The Royal Society Interface},
  year = {2022},
  doi = {10.1098/rsif.2022.0214},
}
Figure