The Role of Context in the Neural Processing of Speech in Autism Spectrum Disorder

Project Summary This project aims to advance understanding of brain mechanisms contributing to speech processing differences and related social communication difficulties widely observed in autism spectrum disorder (ASD). In ASD, the most prominent features of speech processing differences include challenges perceiving social and emotional cues from speech signals, and perception of speech in noisy environments. Such difficulties have been theorized to relate to differences in context-dependent auditory processing. That is, autistic individuals appear less adept in using global acoustic and statistical contexts (e.g., the likelihood of one speech sound following a particular sequence of speech sounds or words) to hone in on important aspects of speech signals and suppress irrelevant auditory signals (e.g., noise) in everyday speech environments, with important clinical consequences. Critically, the biological and mechanistic origins of this important functional difference in ASD remain unclear. Recent studies have demonstrated that in typical development, context-dependent auditory processing is mechanistically supported by an interaction between top-down modulation and local representations of speech that fine-tune sensory signals based on expectations computed from auditory history. In this proposal, we test the hypothesis that the homeostasis (i.e., balance) between top-down and local representations of speech signals is disrupted in ASD, which manifests system-wise along the neural auditory pathway. We propose a set of multi-level, advanced and innovative encephalography (EEG) approaches to study how autistic individuals incorporate contextual information to support speech processing in both fundamental encoding of speech sounds and higher order processing of continuous speech samples. To establish the clinical relevance of context-dependent auditory processing, we further test a hypothesis that these brain signatures relate and contribute to key clinical-behavioral speech and language phenotypes related to the social communication domain of ASD. Results of this study will provide important mechanistic insight to help unpack the currently insufficiently understood etiology of functional speech processing difficulties in ASD. As speech processing is crucial in our daily lives (e.g., from the development of friendships through talking and listening, to workplace communication), understanding the causes and underlying mechanisms of speech processing differences in ASD has broad clinical-behavioral implications, essential for the development of effective, mechanistically targeted interventions to improve quality of life. Elicited with innovative and mechanistically apt experimental paradigms that reflect real-world experiences and clinically- translatable EEG techniques with clear biological bases, our hypothesized neural markers of context-dependent auditory processing in ASD has great potential as an objective and robust tool to be implemented in translational research to predict and measure clinical improvements in speech- and language-related interventions.