Thursday, July 22, 2010

Autism Has a Unique Vocal Signature, and LENA Can Find It

What if you could reliably detect autism, developmental delays, or speech delays just by listening to a child's babble? LENA can do it.
"GrasGr√ľn"
LENA, which stands for Language ENvironment Analysis, is entirely automated software developed by researchers at the University of Memphis. According to the developers, LENA is 86% accurate in detecting "very young children" with autism. The researchers conducted their research by having 232 children wear special processors in the pocket of their clothes, which recorded their daily speech.

According to the article, "Oller and colleagues found that the most important of the 12 acoustic parameters turned out to be ones that targeted the ability of children to produce well-formed syllables (syllabification), by moving the jaw and tongue rapidly during vocalization," and "the sound samples from the autistic children showed little evidence of development of syllabification, in that the relevant acoustic parameters did not change much as the children got older..."

Beyond just autism, LENA could detect delays in development and in language acquisition, for example in a deaf child who is not acquiring spoken language. It could even be used in therapy at home to gauge a child's progress.

No comments:

Post a Comment

All comments on my blog are moderated, and I reserve the right not to publish any comments for any reason. This blog is set up so that anyone can comment. If you have trouble, email me, or check Blogger's help section.