Josiane Mukahirwa¹ (), Qianli Meng¹, Keith Schneider¹; ¹University of Delaware

Introduction:The magnocellular system theory of developmental dyslexia suggests a pervasive dysfunction of the magnocellular system throughout the brain. Therefore, we sought to test the distribution of magnocellular cortical input by measuring brain responses to visual flicker in dyslexic subjects. We used MVPA to discriminate different frequencies and compare the results between dyslexic and normal readers. Methods: Using fMRI at 3T, we imaged whole-brain responses in 8 subjects with dyslexia and 11 normal readers as they observed epochs (3–6 s) of black and white checkerboards, contrast reversing at frequencies from 2–60 Hz. After preprocessing (motion correction, temporal filtering, and smoothing), we performed a trial-wise estimate for individual subjects. The results were used to perform multi-voxel pattern analyses (MVPA). A support vector machine classifier was used for a whole brain searchlight analysis to explore brain areas that displayed significant discrimination probabilities between the lowest and highest frequency bins. Results: The normal readers displayed high discrimination in the visual cortex, in particular, left V5/MT (FWE, p < .05, TFCE), while individuals with dyslexia exhibited discrimination only in the posterior visual cortex and not in V5/MT. The between-group difference was not significant in our small sample. Discussion: The results suggest that the subjects with dyslexia may have weaker magnocellular input to visual cortex. In future analyses we plan to examine between-group classifiers and region of interest analyses.