Voice analysis may screen infants for autism

Voice analysis may screen infants for autism

A new automated vocal analysis technology could fundamentally change how autism spectrum disorders (ASD) are identified in young children. The LENATM (Language Environment Analysis) system could also revolutionize the study of some aspects of language itself, according to Steven Warren, one of the authors of the study reported in the July 19 online Proceedings of the National Academy of Sciences. 

“Some studies 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 Warren, professor of applied behavioral science and KU vice provost.

The LENA system, developed at the Boulder, Co. LENA Foundation, automatically counts child vocalizations recorded by a small device placed in a child’s clothing. The recordings for the study were then submitted to an automatic acoustic analysis designed by the researchers that showed that pre-verbal vocalizations of very young children with autism are distinctly different from those of typically developing children. The team was able to identify those children with ASD by their vocal signatures alone with a robust 86 percent accuracy.

The system also differentiated typically developing children and children with autism from children with language delay.

The researchers analyzed 1,486 all-day recordings from 232 children through an algorithm based on 12 acoustic parameters associated with vocal development. The most important of these proved to be those targeting syllabification, the ability of children to produce well-formed syllables with rapid movements of the jaw and tongue during vocalization. These showed little evidence of development in the children with ASD.

The research team, led by D. Kimbrough Oller, professor and chair of excellence in audiology and speech language pathology at the University of Memphis, called the findings a proof of concept that automated analysis of massive samples of vocalizations can now be included in research on vocal development.

Although aberrations in the speech of children with autism spectrum disorders has been examined by researchers and clinicians for more than 20 years, vocal characteristics are not included in standard criteria for the screening or diagnosis of ASD, said Warren.

Warren says that children with ASD can frequently be diagnosed as early as 18 months but that the median age of diagnosis is 5.7 years in the U.S. “This technology could help pediatricians screen children for ASD and determine if referral to a specialist for a full diagnosis is required and get them into earlier and more effective treatments.”