Macroscale brain states support the control of semantic cognition

Xiuyi Wang,Katya Krieger-Redwood,Yanni Cui,Jonathan Smallwood,Yi Du,Elizabeth Jefferies
DOI: https://doi.org/10.1101/2024.02.29.582250
2024-03-02
Abstract:Understanding how the human brain adapts to varying cognitive demands is crucial in neuroscience. Here, we examined how networks involved in controlled semantic retrieval reconfigure themselves to generate neurocognitive states appropriate to different task contexts. We parametrically varied the demands of two semantic tasks - global association and feature matching judgments - and contrasted these effects of cognitive control with those of non-semantic tasks. We then characterized these effects on the cortical surface and within a whole-brain state space, anchored by the top three dimensions of intrinsic connectivity. Our results revealed that demanding semantic association tasks elicited more activation in the anterior regions of the prefrontal and temporal cortex. In contrast, difficult semantic feature matching tasks produced more posterior activation, aligning closely with regions engaged during multiple demanding non-semantic tasks. In both semantic feature matching and non-semantic contexts, the difficulty effects were situated towards the controlled end of a dimension capturing functional separation between cognitive control and default mode regions. Conversely, in semantic association tasks, the difficulty effects elicited similar responses across both cognitive control and default mode networks. Furthermore, controlled association and non-semantic control were located towards the heteromodal end of a heteromodal- unimodal dimension, while semantic feature matching involved a brain state that was more visual and unimodal. These findings demonstrate that a variety of brain states underpin controlled cognition. Specifically, cognitive control regions with heteromodal semantic knowledge system to identify contextually relevant conceptual overlaps (e.g., associating ’DOG’ with ’BEACH’), and from heteromodal memory regions for modality-specific conceptual overlaps (e.g., connecting ’DALMATIAN’ with ’BLACK AND WHITE’).
Neuroscience
What problem does this paper attempt to address?