Slide Slam I11
The transition from vision to language: distinct patterns of functional connectivity for sub-regions of the visual word form area
Maya Yablonski1, Iliana Karipidis1, Jason Yeatman1; 1Stanford University
Text evokes a response in a region of ventral occipitotemporal cortex, termed the Visual Word Form Area (VWFA; Cohen et al., 2000). Recent findings suggest that this word-selective cortex comprises at least two distinct subregions that process different aspects of written language: the more posterior VWFA-1 is sensitive to visual features, while the more anterior VWFA-2 processes higher level language information like lexical properties (Lerma-Usabiaga et al., 2018; White et al., 2019). Complementary evidence suggests that these sub-regions differ from each other not only functionally, but also in their cortical microstructure and white matter connectivity. Here, we explore whether these two sub-regions exhibit different patterns of functional connectivity, and whether these patterns have relevance for reading ability. A sample of 90 children and adolescents was selected from the HBN database (Alexander et al., 2017; 47 males, mean age 13±3.5y, range 6.5-20y). Selection criteria included the availability of two high quality resting state functional runs and a reading assessment (Wechsler Individual Achievement Test, WIAT). Subjects were included only if each of their functional scans had sufficient quality in terms of both SNR (>1.5) and motion (mean framewise displacement < 0.5mm). For each resting-state run, 375 volumes were collected with the following parameters: TR=820ms, TE=30ms, resolution=2.4mm. Resting-state data were preprocessed with C-PAC (Craddock et al., 2013) and then analyzed using the CONN toolbox (Whitfield-Gabrieli, S. & Nieto-Castanon, A., 2012). The average locations of VWFA-1 and VWFA-2 were defined based on previously reported coordinates (Lerma-Usabiaga et al., 2018). Whole brain functional connectivity analysis revealed that VWFA-1 was more strongly correlated with bilateral visual regions including large portions of ventral occipitotemporal cortex and posterior parietal cortex (FDR corrected at p < 0.05). In contrast, VWFA-2 was more strongly correlated with language regions in the frontal and lateral parietal lobes, in particular the bilateral inferior frontal gyrus (IFG), left superior frontal gyrus (SFG) and bilateral intraparietal sulcus (IPS). Next, for each subject we extracted the connectivity strength (pearson’s r) from each of the clusters that were differentially associated with VWFA-1 and VWFA-2 in the group analysis and examined whether connectivity strength was associated with individual reading skills. This analysis revealed a striking dissociation: Reading ability was associated with connectivity strength between left IFG or left IPS and VWFA-2, but not with connectivity strength between these regions and VWFA-1. Further, connectivity between the same IPS cluster and VWFA-1 was associated with age, but not with reading ability. In sum, our findings support the distinction between sub-regions of the VWFA, showing that adjacent regions are coupled with distinct networks. In addition, we show that the strength of these connectivity patterns is associated with reading ability, such that better readers have stronger connectivity between frontal language regions and VWFA-2, but not VWFA-1. These findings reveal an elegant correspondence between white matter anatomy, functional connectivity, cortical computation and the development of reading skills.