Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Hierarchical lateralization of the speech production network at rest
Poster Session B, Friday, September 12, 4:30 - 6:00 pm, Field House
E. Susan Duncan1, Andrea Leone-Thide2, Shelby Kayga1, Adam Buchwald2; 1The Graduate Center, CUNY, 2New York University
Introduction: The dorsal stream for language is putatively strongly left-lateralized. We used resting-state fMRI to investigate the lateralization of functional connectivity (FC) in three bilateral pairs of regions engaged in the functional progression from higher-order articulatory planning to direct motor neuron output for speech. Regions included the ventral portions of primary motor cortex (MC, A4), premotor cortex (PMC, A6), and pars opercularis (pOp, A44). Methods: Data came from the publicly available dataset on cognitive aging from the Cambridge Centre for Ageing Neuroscience (Cam-CAN). Participants (n=118) were included if they were right-handed (EHI>40), obtained the maximum score on cognitive screening, and had >6 minutes of usable resting-state data. Data were preprocessed using the established Cam-CAN pipeline followed by analysis with AFNI/FSL. Residual time series were obtained by regressing motion parameters and WM/CSF signal, then band-pass filtering (0.01-0.1 Hz). We extracted the eigenvariate time series from each ROI (identified using the Brainnetome atlas) and performed whole brain correlations across all other brain regions in the atlas (n=274), converting to z-scores to permit inter-subject comparison. We used three t-tests to identify significant FC differences for homologous regions, including covariates (age, sex, education, handedness, duration of usable data) and applying FWE multiple comparison correction (3dClusterSim, voxelwise p=0.001, corrected p=0.01). For each region, we present results indicating which regions were more correlated with LH and RH. Results: L_vMC vs. R_vMC: LH greater FC with L sensory/motor regions (primary auditory, ventral somatosensory, posterior insular cortices), and thalamus, as well as R dorsal angular gyrus and ventral precuneus/posterior cingulate. RH exhibited greater FC with R sensory/motor regions (primary auditory cortex, ventral pOp) and L caudal supramarginal gyrus. L_vPMC vs. R_vPMC: LH greater FC throughout L frontal and parietal lobes, anterior/posterior ITG, fusiform gyrus, precuneus, and posterior cingulate; also noted for R cerebellum. RH significantly greater FC with R sensory/motor regions and additional higher-order association cortices (MFG, IFG, pMTG, STS, IPL, SPL, anterior/posterior cingulate, precuneus); also basal ganglia and thalamus. L_vpOp vs. R_vpOp: LH significantly greater FC with wide swath of bilateral regions (much of PMC, SFG, medial OrbG, anterior/posterior cingulate, precuneus, anterior/posterior MTG and STS, anterior ITG). Additional bilateral connectivity included SMA, amygdala, hippocampus, parahippocampal gyri, and visual regions (including lateral and medial occipital cortices and fusiform gyri). Strict lateralization for IFG, MFG, and IPL, and the R cerebellum. RH significantly greater FC with multiple R regions (dorsolateral SFG, MFG, IFG, lateral OrbG, vPMC, vMC, anterior insula, posterior cingulate, precuneus, IPL, SPL, posterior ITG, entire MTG). Significant left FC present only in SMG and cerebellum. Conclusions: Overall, vMC, vPMC, and vpOp show an expanding pattern of whole-brain FC from modality-specific sensory/motor regions to higher-order cognitive, linguistic, affective, and integrative networks. Left hemisphere regions tend to exhibit broader connectivity patterns than their right homologues, with greater cross-hemispheric FC at the lowest and highest levels. Findings are consistent with the hierarchical roles of these regions in articulatory planning and execution, supporting a central role for the left hemisphere in the speech network even at rest.
Topic Areas: Speech Motor Control, Language Production