Voice analysis may help ID autism early

CNN

• share
• email
• print

Analyzing a child's voice may someday be a way to screen children for autism, according to a new study.

"What the study does is apply a technology that is capable of identifying sound differences in children's speech," says Steve Warren, a behavioral psychologist at the University of Kansas and one of the study's authors.

Researchers developed a new method for using an automated voice analysis system called LENA, which appears to pick out voice patterns of children with language disorders related to autism and language delay.

Researchers gave 232 children a small device with a very tiny microphone that fit in the pocket of a child's clothing and could record everything the child said all day long. The recordings were then downloaded into a computer to be analyzed by the automated system. The children in the study were between the ages of 10 months and 4 years old. 77 of these children were known to have autism.

Researchers found the machine accurately separated the voice patterns of typically developing children from children with autism and children with a language delay. Warren is quick to point out that this study is only a proof of concept and that a lot more research needs to be done.

"We have a system that could potentially be a way to screen for autism and language delay – potentially well before diagnosis," says Warren.

He emphasizes that this new algorithm or method for voice analysis is not a screening tool for autism – yet. Warren says more studies involving children who are younger and where researchers know nothing about their development still have to be conducted.

Full article from CNN here