Arousal-driven critical roaming reproduces human functional connectivity dynamics
- 1Laboratoire de Neurosciences Cognitives et Adaptatives (LNCA), Université de Strasbourg, CNRS, UMR 7364, Strasbourg, France
Abstract
Ongoing brain activity displays rich temporal variability associated with efficient cognition, with functional connectivity (FC) continually reconfiguring over time. The resulting functional connectivity dynamics (FCD) specifically show complex, fat-tailed statistics that alternate between persistent epochs and faster reconfiguration transients. While nonlinear whole-brain models tuned nearby a critical point have reproduced some aspects of FCD, they fall short of capturing its full temporal complexity. We propose that slow fluctuations in arousal offer a biologically plausible mechanism for exploring critical regimes in large-scale brain dynamics and thus enrich FCD. Using a connectome-based model of coupled cortical populations, we identified phase boundaries where system dynamics transition between regimes of faster or slower FCD. We then phenomenologically incorporated arousal changes, modeling them as stochastic fluctuations in key parameters such as cortical excitability, input gain, and noise amplitude. This explicitly time-dependent formulation enables the system to roam dynamically across regime boundaries, flexibly tuning its distance from critical transition lines and producing intermittent transitions that mirror the stochastic evolution observed in empirical FCD. Fitting these models to human resting-state fMRI and performing model comparison, we find that arousal-driven models more accurately reproduce the distinctive quantitative features of FCD, with the greatest improvements coming from the previously poorly accounted fat-tailed portions of the distributions. Together, these results suggest that arousal fluctuations—likely mediated by changes in neuromodulatory tone—shape the brain’s attractor landscape over time, expanding the repertoire of accessible functional network states and providing a mechanistic basis for the complexity of spontaneous functional dynamics.
What accounts for the highly dynamic patterns of brain activity at rest? This computational study shows that slow fluctuations in arousal help reproduce key features of brain regional communication dynamics, with implications for our understanding of critical brain states.
Related
- Mapping dynamics in neuronal activation and molecular connectivity with functional PET
- Intrinsic network dynamics catalyze switching in functional connectivity states of the brain
- Spatial organization of neural responses to physical and agentive movement dynamics is reflected in intrinsic functional connectivity
- Physiological arousal as a predominant source of individual differences in functional brain networks
- Population fragmentation and the weak Allee effect: critical transitions caused by increased connectivity
BibTeX
@article{pathak2026arousaldriven,
title = {{Arousal-driven critical roaming reproduces human functional connectivity dynamics}},
author = {Anagh Pathak and Demian Battaglia},
journal = {PLOS Biology},
year = {2026},
month = sep,
doi = {10.1371/journal.pbio.3003916},
url = {https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3003916},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Arousal-driven critical roaming reproduces human functional connectivity dynamics
AU - Anagh Pathak
AU - Demian Battaglia
PY - 2026
DA - 2026/09/01
JO - PLOS Biology
DO - 10.1371/journal.pbio.3003916
UR - https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3003916
AB - Ongoing brain activity displays rich temporal variability associated with efficient cognition, with functional connectivity (FC) continually reconfiguring over time. The resulting functional connectivity dynamics (FCD) specifically show complex, fat-tailed statistics that alternate between persistent epochs and faster reconfiguration transients. While nonlinear whole-brain models tuned nearby a critical point have reproduced some aspects of FCD, they fall short of capturing its full temporal complexity. We propose that slow fluctuations in arousal offer a biologically plausible mechanism for exploring critical regimes in large-scale brain dynamics and thus enrich FCD. Using a connectome-based model of coupled cortical populations, we identified phase boundaries where system dynamics transition between regimes of faster or slower FCD. We then phenomenologically incorporated arousal changes, modeling them as stochastic fluctuations in key parameters such as cortical excitability, input gain, and noise amplitude. This explicitly time-dependent formulation enables the system to roam dynamically across regime boundaries, flexibly tuning its distance from critical transition lines and producing intermittent transitions that mirror the stochastic evolution observed in empirical FCD. Fitting these models to human resting-state fMRI and performing model comparison, we find that arousal-driven models more accurately reproduce the distinctive quantitative features of FCD, with the greatest improvements coming from the previously poorly accounted fat-tailed portions of the distributions. Together, these results suggest that arousal fluctuations—likely mediated by changes in neuromodulatory tone—shape the brain’s attractor landscape over time, expanding the repertoire of accessible functional network states and providing a mechanistic basis for the complexity of spontaneous functional dynamics. What accounts for the highly dynamic patterns of brain activity at rest? This computational study shows that slow fluctuations in arousal help reproduce key features of brain regional communication dynamics, with implications for our understanding of critical brain states.
SN - 1545-7885
ER -