Abstract
We report a study of motion-coherence and form-coherence in dyslexia using psychophysics and fMRI. Deficits in perceiving coherent motion have been reported in dyslexia, but no differences in form coherence have previously been found. We measured form and motion thresholds for detecting a Glass stimulus of varying coherence in a field of random dots. A coherent visual patch was depicted by dots separated by a rotational transformation in space (form coherence) or space-time (motion coherence). Stimuli were presented for 0.25s to prevent serial search strategies. Coherence was progressively reduced from 1.0 until an error was made, in the manner of a 2-up, 1-down staircase, with thresholds calculated as the mean of 6 reversals, ignoring 2 initial reversals. Motion coherence thresholds were higher in dyslexic participants than controls, replicating previous findings. Form coherence thresholds were also significantly higher in the dyslexic population, however, indicating a parvo- or ventral stream deficit. In a parallel event-related fMRI study in which participants performed the same task in the scanner at fixed coherence levels (0, 0.125, 0.25, 0.5, 1.0), BOLD responses indicated a variable but non-linear relationship with motion coherence in V5/MT for dyslexic participants, compared to the linear response in controls which was consistent with previous studies (Rees et al, 2000 Nature Neuroscience 3 716–723). No significant difference was found between dyslexics and controls in the relationship between BOLD response and form coherence in any of the four occipital regions of interest.