Related Experiment Video
Updated: Dec 22, 2025
06:45
Generation and On-Demand Initiation of Acute Ictal Activity in Rodent and Human Tissue
Published on: January 19, 2019
9.3K
Ghost Attractors in Spontaneous Brain Activity: Recurrent Excursions Into Functionally-Relevant BOLD Phase-Locking
Jakub Vohryzek1,2, Gustavo Deco3,4,5,6, Bruno Cessac7
1Department of Psychiatry, University of Oxford, Oxford, United Kingdom.
Frontiers in Systems Neuroscience
|May 5, 2020
Summary
Resting-state brain networks exhibit dynamic patterns, revealing individual "fingerprints" in brain activity. These patterns, analyzed using Leading Eigenvector Dynamics Analysis (LEiDA), offer insights into brain function and dynamics.
Area of Science:
- Neuroscience
- Dynamical Systems Theory
- Computational Biology
Background:
- Resting-state brain activity involves transient, synchronized patterns within functional networks.
- Understanding the biophysical mechanisms of intrinsic brain activity requires detailed dynamical characterization.
Purpose of the Study:
- To characterize the dynamical features of whole-brain activity using fMRI data.
- To identify and analyze distinct brain activity states and transitions between them.
Main Methods:
- Utilized Leading Eigenvector Dynamics Analysis (LEiDA) on a Human Connectome Project fMRI dataset (100 participants).
- Clustered BOLD phase-locking patterns into k states and analyzed state space trajectories.
- Calculated Fractional Occupancy, Dwell Times, and Transition Probabilities.
Main Results:
- Cluster centroids of brain states corresponded to known functional subsystems.
- High-frequency BOLD signals (>0.1 Hz) maximized within-subject reliability, indicating individual dynamical fingerprints.
- Resting-state networks were characterized as excursions into weakly-stable 'ghost attractor' states.
Conclusions:
- Spontaneous brain activity can be modeled as trajectories within a state space, revealing dynamic network patterns.
- Individual differences in brain dynamics are detectable at high frequencies.
- The 'ghost attractor' model provides a mechanistic explanation for resting-state network dynamics and transitions.
