| dc.description.abstract | Naturalistic language production is integral to daily communication and ubiquitously impaired in aphasia, an acquired language disorder caused by focal brain damage typically due to stroke. Despite its importance, naturalistic language production is an understudied area of aphasia, in part because there are few measurement systems available that are comprehensive, psychometrically validated, and interpretable relative to speech and language domains. This dissertation leveraged auditory-perceptual rating of connected speech in aphasia (APROCSA), a novel measurement system I developed that yields scores on four explanatory dimensions of performance, and applied APROCSA to a large and diverse longitudinal study on aphasia recovery after stroke. In the first study that included 118 participants with aphasia, I found that naturalistic language production has distinct yet overlapping neuroanatomical correlates across key left language regions in acute stroke, and these correlates aligned with both historical and contemporary account on the neurobiology of speech and language. In the second study that included 195 participants with aphasia, I found that naturalistic language production recovery during the first year after stroke is nonlinear, multidimensional, and likely interactive; that brain regions important to long-term outcomes are largely the same as those important in acute stroke; and that brain regions important to recovery are relatively distinct, revealing important information about the capacity of certain regions for reorganization. Collectively, these findings constitute a brain-behavior framework of naturalistic language production in aphasia that could influence clinical management practices, and could support stroke survivors with aphasia in understanding the nature of their everyday speech and language difficulties. | |
