Aphasia is an impairment of language that is a common consequence of stroke and has serious negative effects on health and well-being. Aphasia diagnosis continues to be organized around a 19th century model of the neural basis of language, but cognitive neuroscience research over the last 15-20 years has converged to a very different model of the cognitive and neural organization of spoken language. This contemporary model provides a precise computational account of the sub-systems that support spoken language, but does not explain how those sub-systems produce functional communication ? the outcome that is most important to people with aphasia and to clinicians. The long-term goal of this project is to develop theory-informed, clinically-relevant prognostic tools that combine behavioral and neuroimaging information. The overall objective of this application is to determine the relationships between spoken functional communication impairments of language sub-systems, and neuroanatomical disruption in chronic post-stroke aphasia. The overall project is divided into three specific aims: (1) Determine how spoken functional communication is related to deficits in language sub-systems. We will test how the three key language sub-systems ? semantics, phonology, and sentence planning ? are related to functional communication in a large sample of individuals with post-stroke aphasia. (2) Identify the lesion correlates of spoken functional communication deficits using lesion-symptom mapping. We will conduct the first LSM study of spoken functional communication using multimodal neuroimaging and machine learning tools to discover robust lesion correlates of spoken functional communication. (3) Develop a prediction model of chronic language sub-system and functional communication deficits based on acute lesion data. Routine clinical neuroimaging data collected in the acute stage (48-72 hours after stroke) will be used to build and evaluate a prediction model of chronic deficits in language sub- systems and functional communication. Upon completion of this project, we will have determined how behavioral deficits and lesion patterns are related to functional communication deficits, and developed a prediction model of such deficits based on acute-stage clinical neuroimaging. This integration of psycholinguistics, neuroanatomy, and functional communication will provide theory-informed, clinically-relevant predictions of communication deficits. This project addresses NIDCD Strategic Priority Area 3 (Improving Diagnosis, Treatment, and Prevention) by developing a neural biomarker of objective diagnosis and prognosis for acquired language impairments.
This project will integrate investigate how the cognitive and neural sub-systems that support spoken language work together to allow speakers with language deficits to convey their message. The studies apply machine learning tools to behavioral assessments, neuroimaging, and measures of functional communication in order to reveal how they are related. The long- term goal of this project is to develop theory-informed, clinically-relevant prognostic tools that combine behavioral and neuroimaging information.