Couching the brain’s multiple-demand and language-specific systems within a macroscale gradient of cortical connectivity

Abstract

AbstractThe cerebrum comprises a set of specialised systems that tile across the cortical sheet, forming a tapestry-like configuration. For example, the multiple-demand and language-specific systems occupy largely separate neural estates and exhibit disparate functional profiles. Although delimiting the boundary between systems informs where cortical sheet functionally fractionates, it remains unclear why different systems’ topographical placements are spatially configured in typical manners and how a macroscale architecture arises from this topography. Novel approaches have tackled this challenge by condensing the topography into a principal gradient, which represents the workflow of information processing from sensory-motoric to abstract-cognitive. To understand how the multiple-demand and language-specific systems are accommodated in the gradient framework, here we used fMRI to probe cognitive operations in semantic vs. visuospatial domains and projected functional activities onto the principal gradient. We found that the two systems showed distinct trajectories of distribution along gradient tiers, suggesting different roles in the transition from sensation to cognition. Critically, when semantic processing became difficult, the brain recruited a specialised ‘semantic-control’ system that was a functional and anatomical ‘hybrid’ juxtaposed between the multi-demand and language systems. We discuss how the brain’s modular division can be better understood through the lens of a dimensionality-reduced gradient-like architecture.