Speech Patterns May ID Kids With Autism
Study Shows Vocal Analysis May Help Identify Children With Autism Spectrum Disorder
July 19, 2010 -- Researchers have developed a new technology that uses the vocal recordings of children to assess their speech patterns and helps in identifying who has autism spectrum disorder or language delay, according to a report.
The new tool could have an impact on how children are screened for autism, language delay, and other developmental disorders.
The tool is a battery-powered device that attaches to children's clothing. Known as the LENA (Language Environment Analysis) system, it works by recording children's utterances, from infant cooing to preschoolers speaking words, and then analyzes these utterances for speech patterns and characteristics to create a vocal profile.
The LENA system showed that the utterances of very young children with autism are different from those with children who are developing normally -- with 86% accuracy.
The findings are published in the online edition of Proceedings of the National Academy of Sciences.
Researchers led by D. Kimbrough Oller, professor and chair of excellence in audiology and speech language pathology at the University of Memphis in Tennessee, analyzed 1,486 all-day recordings from 232 children ages 10 months to 4 years.
The recordings took place in the children's homes while the children went about their regular daily activities. The researchers collected more than 3.1 million utterances for analysis.
Based on 12 acoustic parameters associated with vocal development, the research team found that a child's ability to produce well-formed syllables while quickly moving the jaw and tongue was a critical parameter that helped confirm those who had diagnosed autism spectrum disorders or language delay and those who did not.
Early Identification of Kids With Autism
Although difficulties with speech are common among those with autism spectrum disorders, vocal characteristics have not been a part of the diagnostic criteria.
"A small number of studies had previously suggested that children with autism have a markedly different vocal signature, but until now, we have been held back from using this knowledge in clinical applications by the lack of measurement technology," said study co-researcher Steven F. Warren, professor of applied behavioral science and vice provost for research at the University of Kansas.
The hardware and software used in the study were developed by the for-profit company Infoture Inc., which was reconstituted as the nonprofit LENA Foundation in 2009. Warren received consultation fees while serving on Infoture's scientific advisory board, and many of the current study's researchers are either LENA Foundation employees or previous employees of Infoture/the LENA Foundation.
It is possible the LENA system could objectively identify children with autism spectrum disorders even earlier, which would allow doctors and families to begin treatment and interventions sooner rather than later, the researchers report. Warren notes that the mean age of an autism spectrum disorder is 5.7 years for children in the U.S., but that the LENA system may be a more cost-effective way to evaluate a child and could produce a diagnosis in children as young as 18 months.
The researchers say more study is needed, but Warren says in a news release that the results are promising. "This technology could help pediatricians screen children for autism spectrum disorders to determine if a referral to a specialist for a full diagnosis is required and get those children into earlier and more effective treatments."