Oregon Health & Science University's Center for Spoken Language Understanding has an unique research program on algorithms that analyze speech, language, facial expression, eye gaze, and movement in Autism Spectrum Disorder (ASD).
- automated measurement of prosodic patterns indicative of ASD
- using dialog transcripts to assess a child's level of restricted interests
- automated detection of echolalia (echoing the speech a conversation partner)
- detecting inappropriate response to prompts or questions.
Starting with speech and language, we continually expand our work into new behavioral domains such as measuring the degree to which facial expression is coordinated with speech – another clinical feature of ASD.
Our algorithms can serve many purposes: automated at-home screening for ASD, discovering new subtypes of ASD, assisting with the diagnostic process, and providing objective, quantitative treatment response measures.
This work is made possible by close collaboration between computer scientists, applied mathematicians, and clinical/biomedical autism experts.
We gratefully acknowledge funding from the National Institutes of Heath, the National Science Foundation, the Nancy Lurie Marks Family Foundation, Autism Speaks, and the Swigert Foundation.