Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Involvement of the left angular gyrus and posterior superior temporal sulcus in semantic control
Poster Session D, Saturday, September 13, 5:00 - 6:30 pm, Field House
Yeonwoo Kim1, Ryan Staples1,2, Peter Turkeltaub1,3,4; 1Center for Brain Plasticity and Recovery, Georgetown University, 2Georgetown-Howard Universities Center for Clinical and Translational Science, Georgetown University, 3Center for Aphasia Research and Rehabilitation, Georgetown University, 4Research Division, MedStar National Rehabilitation Hospital
Semantic control refers to the goal-oriented control of semantic representations. As such, it is both a subset of general cognitive control and a crucial component of the controlled semantic cognition (CSC) framework. Previous fMRI meta-analyses suggest the left inferior frontal gyrus (IFG), the left angular gyrus (AG), and left posterior lateral temporal cortex are crucial to semantic control. However, it is unclear whether activity in these regions reflects processes that rely on semantic representations or domain-general control rather than semantic control. Here, we investigated whether activation during semantic judgement correlated with semantic and non-semantic control performance measured in a semantic control task outside the scanner. Participants included 61 neurotypical adults (Age=61.5y (11.1); Gender=32F/29M; Education=17.4y (2.5)). All participants performed a semantic control task requiring rapid selection of targets from foils, where the target and foil objects were semantically related (S) or unrelated (U). In the no-switch conditions (S1, U1), the target remained same throughout the block. In the switch conditions, the target was switched in a repeated sequence of two (S2, U2) or three (S3, U3) stimuli. In this way, we expect the S3 condition to represent high semantic control demand, and U3 to represent high non-semantic control demand. Participants also completed an adaptive semantic decision fMRI task to measure activation during semantic judgement. Mean activation on this task was calculated for each participant in ROIs defined based on group-level activation and Spearman’s correlations related ROI activation with performance on the semantic control task. Linear regressions analyzing S1, S3, U1, and U3 conditions were conducted to assess all ROIs simultaneously as predictors, including age, education, and fMRI task performance as covariates of no interest. Spearman’s correlations showed significant relationships between activations and performance on the semantic control task at the highest switching-demand level (S3) in the left IFG (rs(59)=0.35, p=0.026), left pSTS (rs(59)=0.38, p=0.026), left medial orbitofrontal region (rs(59)=0.32, p=0.033), and left AG (rs(59)=0.32, p=0.033). No correlations were significant for the S1, U3, or U1 conditions. Linear regressions showed activity in the AG significantly predicted both S1 (β=0.038, p=0.035) and S3 (β=0.039, p=0.048) performance. Furthermore, activity in the pSTS was a significant predictor of S3 performance (β=0.073, p=0.0045), while anterior temporal activation predicted S1 performance (β=-0.080, p=0.035) and medial orbitofrontal activation significantly predicted both U3 (β=0.065, p=0.0092) and S1 (β=0.040, p=0.023) performance. Our findings suggest the AG may be involved in semantic processing regardless of switching-demand, implying the AG’s role in the CSC framework may be in semantic representation rather than in semantic control. Contrastingly, activity in the pSTS was not a significant predictor of performance on the S1 condition, suggesting this activity may be modulated by need for greater semantic control. Critically, activation in these ROIs did not predict U3 performance, suggesting their role is specific to semantic demands. Interestingly, we did not replicate previous evidence for the IFG’s role in semantic control. Further work is necessary to analyze the subdivisions of the IFG and investigate whether they may differentially impact semantic control ability.
Topic Areas: Control, Selection, and Executive Processes, Meaning: Lexical Semantics