Using an ordinary differential equation model to separate rest and task signals in fMRI

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Cortical activity arises from interactions between integrated networks and task-specific sensory-motor processes.
• Differentiating these components in functional magnetic resonance imaging (fMRI) data is a significant challenge.
• Understanding the relationship between rest and task states in brain dynamics is crucial.

Purpose of the Study:

• To develop a network ordinary differential equation (ODE) model to analyze fMRI data from rest and task conditions.
• To investigate the relationship between rest-state and task-state brain activity.
• To establish the Active Cortex Model principle.

Main Methods:

• Developed a network ordinary differential equation (ODE) model.
• Utilized advanced system identification techniques.
• Analyzed fMRI data from rest and task conditions (Human Connectome Project).

Main Results:

• Task-specific ODEs were found to be subsets of rest-specific ODEs.
• The model improved reaction time predictions by 9% (R²).
• Validated predictions for missing trials, accuracy, and task classification.

Conclusions:

• Established the Active Cortex Model: the cortex is always active, with rest encompassing all processes.
• Task activity represents an elevated subset of rest processes for specific computations.
• Provided a framework to link brain activity, connectivity, and behavior using fMRI data.