You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster B68, Wednesday, November 8, 3:00 – 4:15 pm, Harborview and Loch Raven Ballrooms

Investigating the functional neural circuitry for spelling using graphical models

Kulpreet Cheema1, Dr. William Hodgetts1,2, Dr. Jacqueline Cummine1;1Faculty of Rehabilitation Medicine, University of Alberta, 2The Institute for Reconstructive Sciences in Medicine

Background: Writing skills are imperative to successful academic and social functioning in today’s literate society. Yet, literature exploring the underlying mechanisms associated with written communication, namely spelling, is limited. Although substantial work has been done to develop neuroanatomical models of reading, the same is not true for spelling. Further, while spelling has been reported to activate a large network of brain areas, there are no studies that examine the connectivity among the brain regions in this distributed network. An understanding of the dynamic nature of the distributed neural systems associated with spelling is critical for the development and advancement of theoretical models of written communication. Objective: This study will examine the functional connectivity patterns, using graphical models (a statistical approach for assessing relationships between brain regions of interest), associated with spelling in skilled adults. Methods: Skilled adults (N = 15) were recruited to take part in a functional magnetic resonance imaging (fMRI) study. After their behavioural testing, participants completed the spelling-based fMRI task called letter probe task (LPT) in MRI. During LPT, the participant first hears the word, then sees a letter on the screen and then is asked to indicate if the letter they just saw was in the spelling of the word that they just heard. Participants completed three conditions of LPT: 1) retrieval of the whole word spelling representations is required (exception words e.g. pint), 2) retrieval of the whole word spelling representation is optional (regular words e.g. hint), and 3) retrieval of the whole word spelling representation is impossible thus they must generate the spelling (nonwords e.g. bint). Left hemisphere brain regions were delineated based on previous work: orthographic processing (fusiform gyrus; FFG), speech input (inferior frontal gyrus; IFG), articulatory processing (supplementary motor area (SMA), putamen, cerebellum, precentral gyrus (PCG)) and phonological processes (superior marginal gyrus (SMG), caudate and superior temporal gyrus (STG)). Graphical modelling was applied to assess the functional connectivity between these regions in general spelling network and separately for each condition. Results: IFG, being the main speech input area, emerged as the hub for the general spelling network, as well as for the other spelling conditions. These functional network patterns changed as a function of word types. During the retrieval of exception words, a more extensive network involving brain areas associated with orthographic, phonological and articulatory processing. Retrieval of regular words activated a comparatively restricted network consisting of brain areas related to phonological and orthographic processing. The nonword spelling condition activated an extensive network, with connections to brain areas associated with articulatory and orthographic processing. Significance: Given the increased reliance on written communications skills in today’s society, advancement of neuroanatomical models for spelling is critical. The results from the proposed work will increase the current state of knowledge regarding the underlying neurobiology of spelling performance.

Topic Area: Writing and Spelling

Back to Poster Schedule