Functional imaging data we have obtained with the method of whole-head Magnetoencephalography (MEG), also referred to as Magnetic Source Imaging or MSI, indicate that the spatiotemporal profiles of brain activation of individual dyslexic children differ dramatically from those of individual non-dyslexic children during the performance of tasks that entail phonological decoding. The main differences in the profiles center on the degree of activation of the posterior superior temporal and supramarginal gyri, the inferior frontal area, and, most notably, the angular gyrus. Accordingly, our first aim in this project will be to assess the reliability of these preliminary findings with adequate samples of 80 non-dyslexic and 160 dyslexic children.
The second aim will be to assess whether, in addition to differences in phonological processing, differences in orthographic and visual form processing, between normal and dyslexic children, contribute to the aberrant activation profile of the latter.
The third aim will be to evaluate the hypothesis that the observed reduction of activity of the left temporoparietal area including the angular gyrus, may be attributed to a more general dysfunction of that region that may underlie many of the cognitive deficits associated with dyslexia. In addition to the above aims, we will investigate that possible contribution of attention deficits to the activation profiles and their differences using additional statistical analyses. Finally, taking advantage of the fact that MEG activation profiles are (a) computed for individual subjects, and (b) are characterized by excellent temporal resolution, we will explore the possibility of uncovering subtle differences in the profiles of normal and dyslexic children through the use of fuzzy clustering techniques. Detailed characterization of the spatiotemporal features of individual profiles may become useful in future studies for assessing the effects of intervention strategies on the functional cerebral reorganization of individual dyslexic children.