Presentation

Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Poster Slams

White matter associations with performance on a difficult English spelling task

Poster D56 in Poster Session D with Social Hour, Friday, October 7, 5:30 - 7:15 pm EDT, Millennium Hall
Also presenting in Poster Slam D, Friday, October 7, 5:15 - 5:30 pm EDT, Regency Ballroom

Romi Sagi¹, J.S.H. Taylor², Kyriaki Neophytou³, Brenda Rapp³, Kathleen Rastle⁴, Michal Ben-Shachar¹; ¹Bar-Ilan University, ²University College London, ³Johns Hopkins University, ⁴The Royal Holloway

Producing written language is an everyday life skill that supports effective communication. Multiple neuro-cognitive components are involved in the highly complex process of spelling written words. Yet, spelling remains a relatively under-studied field in terms of its underlying white matter substrates. fMRI studies identified a wide, mostly left-lateralized network of frontal, parietal and temporal active sites that are responsive during spelling tasks (Purcell et al., 2011; Planton et al., 2013). A few studies examined white-matter spelling-related tracts in children with dysgraphia (Gebauer et al., 2012; Banfi et al., 2019), but little is known about the neuroanatomical connections that support spelling in typically developing adults. We evaluated the associations between performance on a difficult spelling-to-dictation task and microstructural properties of language-related white-matter pathways, in 73 English-speaking adults (mean age: 21y ± 4.2, 57 females). Participants were scanned on a 3T Siemens scanner, using a diffusion weighted, single-shot EPI sequence (64 diffusion directions at b = 1000 and 1 volume at b = 0 s/mm^2, voxel size: ~2*2*2mm^3). We used constrained spherical deconvolution (CSD) modeling and probabilistic tractography to reconstruct, in each participant, the three branches of the superior longitudinal fasciculus (SLF I, II & III), the frontotemporal segment of the arcuate fasciculus and the inferior longitudinal fasciculus (ILF). All participants completed a wide cognitive assessment including vocabulary, spoonerism, sight word and phonemic decoding efficiency (TOWRE), nonword repetition (CToPP) and rapid automatized naming (RAN). Spelling scores significantly correlated with vocabulary, spoonerism, phonemic decoding efficiency and nonword repetition, while no correlations were found between spelling and sight word efficiency or RAN. Using a stepwise linear regression model, we found that, out of all tracts of interest, only mean fractional anisotropy (FA) in the left ILF and the right SLF III significantly predicted spelling scores. Next, we calculated Spearman's correlations between spelling scores and FA along these tracts. Because spelling performance was bimodally distributed across individuals, we used the local minimum of the bimodal model to divide participants into two groups of low- (N = 41) and high- performing (N = 32) spellers. Interestingly, while high-performing spellers showed a significant positive correlation between spelling scores and FA within the left ILF (r = .53, p < .05, family-wise error corrected), low-performing spellers did not show such an association. Conversely, low-performing spellers showed a negative correlation between spelling scores and FA within the right SLF III (r = -.47, p < .05, family-wise error corrected), while high-performing spellers did not show such an association. These findings demonstrate the complexity of the neurocognitive architecture of the spelling process. Specifically, spelling is shown to be associated with both dorsal, phonologically-related, frontoparietal pathways, and ventral, lexically-related, occipitotemporal pathways. The distinct association patterns detected in low and high performing spellers could point to their reliance on different cognitive components involved in spelling, such as phoneme-to-grapheme conversion, lexical orthographic representations and orthographic working memory.

Topic Areas: Writing and Spelling, Reading