Interactions Between Large-Scale Functional Brain Networks are Captured by Sparse Coupled HMMs.

Area of Science:

• Neuroscience
• Cognitive Science
• Medical Imaging

Background:

• Resting-state functional magnetic resonance imaging (fMRI) measures spontaneous brain activity, offering insights into brain function.
• Dynamic interactions between brain regions are crucial for understanding brain function, necessitating advanced modeling tools.

Purpose of the Study:

• To develop and validate a novel computational framework for analyzing dynamic functional connectivity in resting-state fMRI data.
• To model temporal evolution of large-scale brain networks and identify cross-network modulatory influences.

Main Methods:

• Utilized an fMRI deconvolution technique to represent resting-state fluctuations.
• Developed a sparse coupled hidden Markov model (SCHMM) to parameterize temporal evolution of network activity.
• Validated the SCHMM on simulated data and applied it to two independent resting-state fMRI datasets.

Main Results:

• The SCHMM accurately estimated network dynamics and revealed precise direct network-to-network modulatory influences, outperforming conventional correlational methods.
• Identified a reproducible set of cross-network couplings in experimental fMRI data across two independent datasets.

Conclusions:

• The proposed SCHMM framework offers a powerful new approach for capturing complex temporal dynamics of brain networks.
• This method provides novel perspectives for studying changes in brain network interactions in both healthy individuals and those with neurological or psychiatric conditions.