For release: July 21, 2010
For press information, contact Curt Guenther at 901/678-2843
A new automated vocal analysis technology could fundamentally change the study of
language development as well as the screening for autism spectrum disorders and language delay, reports a study in the July 19 online Proceedings of the National Academy of Sciences.
Dr. D. Kimbrough Oller
The LENA™ (Language Environment Analysis) system automatically labeled infant and
child vocalizations from recordings. Following that step, an automatic acoustic analysis designed by Dr. D. Kimbrough Oller, professor and chair
of excellence in audiology and speech-language pathology at the University of Memphis,
showed – with 86 percent accuracy – that pre-verbal vocalizations of very young children
with autism are distinctly different from those of typically-developing children.
The system also differentiated typically-developing children and children with autism
from children with language delay, based on the automated vocal analysis.
The researchers analyzed 1,486 all-day recordings from 232 children (more than 3.1
million automatically identified child utterances) through an algorithm based on the
12 acoustic parameters associated with vocal development, as indicated in Oller’s
research over a 35-year period. The most important of these parameters proved to be
the ones targeting syllabification, the ability of children to produce well-formed
syllables with rapid movements of the jaw and tongue during vocalization. Infants
show voluntary control of syllabification and voice in the first months of life and
refine this skill as they acquire language.
“The autistic sample showed little evidence of development on the parameters, as indicated
by low correlations between the parameter values and the children's ages (from 1 to
4 years),” said Oller. “On the other hand, all 12 parameters showed statistically
significant development for both typically-developing children and those with language
The research team called the findings a proof of concept that automated analysis of
massive samples of vocalizations can now be included in the scientific repertoire
for research on vocal development.
Although aberrations in the speech (or lack of speech) of children with autism spectrum
disorders have been examined by researchers and clinicians for more than 20 years,
vocal characteristics are not included in standard criteria for diagnosis of those
disorders, said Steven F. Warren, professor of applied behavioral science and vice
provost for research at the University of Kansas, who contributed to the study and,
along with Oller, was among the first to see the potential of the technology for autism
spectrum disorders screening.
"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," Warren
Oller also noted that a key feature of this approach is that it is totally objective.
“Previous methods of screening for and diagnosis of autism have relied very heavily
on the observations and judgments of parents, teachers, and professional diagnosticians,”
Oller said, “all of which inevitably include subjectivity.”
The researchers predict that LENA, which allows the inexpensive collection and analysis of magnitudes of data unimagined
in language research before now, could have a significant impact on the screening,
assessment, and treatment of autism and the behavioral sciences in general.
“Even though the study focused on English-learning children only,” said Oller, “there
is reason to believe that the technology should be applicable to children speaking
languages other than English. The reason for this expectation is that the model of
acoustic parameters used in the research is designed to be universal with regard to
languages, focusing on features of voice and articulation of speech or speech-like
sounds that occur in languages everywhere.”
The researchers noted that children with autism spectrum disorders can be diagnosed at 18 months, but the median age of diagnosis in the United States is
Warren said that this technology could help pediatricians screen children for ASD to determine if a referral to a specialist for
a full diagnosis is required and get those children into earlier and more effective treatments.
LENA combines a digital language processor and language analysis software. The processor
fits into the pocket of specially designed children's clothing and records everything
the child vocalizes, but it can reliably distinguish a child’s vocalizations from
its cries and vegetative sounds, other voices, and extraneous environmental sounds.
Recordings with the device have been collected since 2006. Parents responded to public
notices in the media and indicated if their children had been diagnosed with autism
or language delay. A speech-language clinician employed by the project also evaluated
many of the children who had a reported diagnosis of language delay. The parents of
many of the children with language delay and all of the children with autism supplied
documentation from the diagnosing clinicians, whose diagnoses were independent of
The recordings were made by the parents at home and in the other natural environments
of the children, simply by turning the recorder on and placing it in the children's
clothing, where it was worn all day.
The discovery that it was possible to differentiate recordings of the autistic children
from those of the typically developing children by the totally objective method of
automated vocal analysis inspired the researchers to consider the possibility of earlier
screening and diagnosis and earlier intervention for children with autism, Oller pointed out.
"Autism interventions are expensive and arduous. This tool may help us develop cost-effective
treatments and better understand how they work and how to keep them working," said Warren.
LENA could allow parents to continue to supplement language enrichment therapy at
home and assess their own effectiveness for themselves, Warren said. "In this way,
LENA could function similarly to the way a pedometer measures how much exercise a
person gets from walking."
Oller summed up his thoughts on the research like this: “We are now in a new era in
research in vocal development. Never before was it possible to use totally automated
systems to analyze massive quantities of recordings collected in children’s homes,
and thus track typical development and provide the basis for significant clinical
contributions to screening, to diagnosis, and to help in monitoring intervention.
Now, all of that is possible, and the effects of this development will be felt in
research, in education, and in the clinical setting.”