Can fMRI functional connectivity index dynamic neural communication and cognition?

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Neural communication requires flexible yet structured dynamics for cognition and behavior.
• Whole-brain functional connectivity (FC) patterns change over time, but their relevance to cognition is debated.
• It's unclear if observed FC patterns reflect actual neural communication or homeostatic processes.

Purpose of the Study:

• To investigate if time-varying functional connectivity (FC) reflects dynamic, cognition-relevant neural communication.
• To determine if FC meets specific criteria for indexing cognitive processes: temporal complexity, behavioral adaptability, and timescale compatibility.
• To model and characterize time-varying FC using fMRI data.

Main Methods:

• Trained multiple time-varying FC models on Human Connectome Project fMRI data.
• Analyzed data across three conditions: rest, working memory task, and motor task.
• Characterized FC dynamics using Principal Component Analysis.

Main Results:

• Time-varying FC exhibits low-dimensional, multi-dimensional trajectories.
• These FC trajectories can be reliably measured.
• The observed FC dynamics satisfy the proposed requirements for reflecting neural communication relevant to cognition.

Conclusions:

• Time-varying FC demonstrates structured, dynamic trajectories that meet necessary conditions for indexing cognition.
• While not sufficient, these findings suggest that time-varying FC may potentially reflect key aspects of neural communication underlying cognitive function.
• Further research is needed to confirm the sufficiency of time-varying FC as a direct measure of cognitive processes.