Distinguishing task-evoked dynamic brain networks from intrinsic activity with tensor component analysis

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Brain Imaging Analysis

Background:

• Task performance reorganizes brain networks, crucial for understanding brain function.
• Functional magnetic resonance imaging (fMRI) data during tasks mixes stimulus responses with intrinsic brain activity.
• Current statistical methods struggle to isolate pure task-evoked brain network activity from intrinsic activity.

Purpose of the Study:

• To introduce Tensor Component Analysis (TCA) for disentangling task-evoked brain network responses from intrinsic brain network (ICN) activity.
• To evaluate TCA's performance in distinguishing dynamic brain networks using simulations and Human Connectome Project (HCP) fMRI data.
• To demonstrate TCA's utility in analyzing Theory of Mind networks in individuals with cannabis use disorder.

Main Methods:

• Tensor Component Analysis (TCA) was proposed to estimate stimulus-evoked brain network responses.
• Numerical simulations were performed to assess TCA's performance.
• In-vivo task and resting-state fMRI data from the Human Connectome Project (HCP) were utilized for validation.

Main Results:

• TCA successfully extracts task-evoked dynamic brain networks distinct from intrinsic brain network activity.
• Simulations and HCP data confirmed TCA's capability to disentangle brain network activities.
• TCA demonstrated effectiveness in evaluating Theory of Mind networks in cannabis use disorder.

Conclusions:

• TCA is a promising tool for analyzing dynamic brain networks during task performance.
• TCA offers a more accurate method than dynamic connectivity analyses for studying brain responses to stimuli.
• This approach provides novel insights into brain-behavior relationships by isolating task-specific network dynamics.