Release Date: April 2, 2018
BUFFALO, N.Y. — It’s a difficult decision that many parents face: Should my child be screened for autism spectrum disorder (ASD)?
Starting today, that decision will be much easier due to a new smartphone app — developed by a University at Buffalo-led research team — designed to help detect ASD in children as young as 12 months old.
Called EarlySee, the app is now available for free on this website (under the Updates & News section) to any android smartphone user. Its launch, which coincides with World Autism Awareness Day, gives parents an easy-to-use tool to help identify ASD earlier.
“The decision whether or not to screen for autism is often based on subjective observations made by parents or primary care doctors,” says Wenyao Xu, a computer scientist at UB and the project’s lead investigator. “This often leads to delays for the actual diagnosis, which can be devastating to children and their parents.
“EarlySee addresses the diagnosis delay by giving parents a widely accessible and quantitative tool they can use in their homes to help spot autism earlier. This is critical because the earlier autism is identified, the more effective the benefits of treatment will be,” says Xu, PhD, assistant professor in the UB Department of Computer Science and Engineering.
The app tracks behavioral information such as facial expression and gaze attention to infer the neurocognitive responses of children looking at pictures of social scenes — for example, a conversation among multiple people. The behavioral responses of someone at high risk of ASD are often different from those of a person without autism.
Based on the results of the test, parents will know whether they should seek further medical attention for a conventional diagnosis.
EarlySee is an advanced version of a previous app developed by UB, the John R. Oishei Children’s Hospital of Buffalo and SUNY Buffalo State. The latest version was developed by Xu and PhD students in his lab, along with the SennoTech Group in China.
Director of News Content
Engineering, Computer Science