Uwe Reischl, a College of Health Sciences professor in the Department of Community and Environmental Health, co-authored a scientific poster that was presented at the 2018 International Society for Autism Research (INSAR) conference May 9-12 in Rotterdam, Netherlands. The poster, “Autism Assessment with Artificial Intelligence,” presented the results of an ongoing multi-site National Institutes of Health funded research and development program designed to integrate deep learning (Artificial Intelligence, or AI) into video assessment of typical and atypical behaviors of children with autism.
The presentation summarized research results comparing AI interpretations of autism behaviors captured on video to the interpretations offered by expert clinicians reviewing the same video images independently. Artificial intelligence was able to achieve an 86 percent agreement with the interpretations of the expert clinicians. It is anticipated that future use of AI in video analysis may assist clinicians in performing remote diagnostic assessments of children with autism more efficiently.
Poster co-authors included R. Oberleitner and J. Schwartz of Behavior Imaging Inc., M. Morrier of Emory University, C. Smith of SARRC and B. Martin of the University of Idaho.