Time-resolved functional connectivity during visuomotor graph learning

Area of Science:

• Neuroscience
• Cognitive Science
• Graph Theory

Background:

• Humans build mental models from perceptual patterns for efficient processing.
• Graph learning reveals sensitivity to topological structures, but neural correlates remain unclear.

Purpose of the Study:

• To investigate neural signatures of graph learning using time-resolved network analysis.
• To compare brain responses to modular versus non-modular lattice graph structures.

Main Methods:

• Functional magnetic resonance imaging (fMRI) data acquisition during a visuomotor graph learning task.
• Application of time-resolved network analyses to fMRI data.
• Assessment of neural signatures associated with learning distinct graph topologies.

Main Results:

• Task performance relies on a flexible visual system and stable brain-wide community structure.
• Increased integration between visual and ventral attention systems observed during learning.
• Limbic, default mode, temporoparietal, and subcortical systems show differential connectivity related to modular graph performance.

Conclusions:

• The brain differentially processes statistical patterns based on their underlying graph topology.
• Findings suggest similarities between neural correlates of graph learning and statistical learning.
• Highlights the role of specific brain networks in processing complex relational information.