Spatiotemporal mapping of visual language processing using magnetoencephalography

The goal of this dissertation was to utilize anatomically constrained magnetoencephalography (MEG) to spatiotemporally map cortical responses supporting word reading. The data demonstrated that visual language tasks evoked a posterior to anterior progression of response, beginning bilaterally in the occipital pole (∼100 ms) then shifting to ventral occipital cortex. Response then lateralized to the left anterior occipital cortex, and spread to ventral and dorsal posterior temporal areas (∼140 ms), followed by anterior temporal (∼200 ms), and finally posteroventral prefrontal cortex (∼250 ms). Subsequently, there was sustained co-activation of all of the above areas, with the exception of posterior occipital cortex. This sustained activation (∼250 ms onward) may permit multiple sources of lexical information (i.e., phonemic, orthographic, semantic) to converge (possibly within inferior prefrontal cortex) and constrain verbal response selection. This response pattern was consistent across a wide range of tasks. However, some functional dissociations were found. Left anterior temporal and inferior prefrontal areas may support semantic processing at ∼250--500 ms poststimulus, while phonological decoding may occur within left temporo-parietal regions. Posterior occipitotemporal areas, particularly within the left hemisphere, may support orthographic processing. Collectively, the results suggest that language is supported by a general, distributed, fronto-temporo-parietal network modulated primarily in the temporal domain.