Time-resolved functional connectivity during visuomotor graph learning

Humans naturally attend to patterns that emerge in our perceptual environments, building mental models that allow for future experiences to be processed more effectively and efficiently. Perceptual events and statistical relations can be represented as nodes and edges in a graph, respectively. Recent work in the field of graph learning has shown that human behavior is sensitive to graph topology, but less is known about how that topology might elicit distinct neural responses during learning. Here, we address this gap in knowledge by applying time-resolved network analyses to fMRI data collected during a visuomotor graph learning task to assess neural signatures of learning modular graphs and non-modular lattice graphs. We found that performance on this task was supported by a highly flexible visual system and otherwise relatively stable brain-wide community structure, cohesiveness within the dorsal attention, limbic, default mode, and subcortical systems, and an increasing degree of integration between the visual and ventral attention systems. Additionally, we found that the time-resolved connectivity of the limbic, default mode, temporoparietal, and subcortical systems was associated with enhanced performance for the modular group but not the lattice group. These findings provide evidence for the differential neural processing of statistical structures with distinct topologies and highlight similarities between the neural correlates of graph learning and statistical learning more broadly.