Autism Can Be Diagnosed at 18 Months with New LENA Digital Language Technology
A new, compelling study brings great news for the earlier diagnosis of autism. A breakthrough study introduced a new system called LENA (Language Environment Analysis) – which is a digital language processor utilized with language analysis software.
The study of 232 autistic and non-autistic children was highly technical from an analytical standpoint, but the results were simply astounding. Using 12 acoustic parameters associated with vocal development, the U.S.-led research team was able to show that vocalizations of young autistic children are different than those of typically developing children with 86 percent accuracy.
The researchers realized that a child’s ability to produce well-formed syllables is a key indicator in identifying autism. Infants show a voluntary ability to begin pronouncing syllables in the first months of life, and these skills sharpen as they acquire language. It is well within the bounds of vocalization analysis to determine with confidence whether or not a child may have autism by 18 months of age.
The autistic children in the sample showed little development within many of the 12 study parameters, while typically developing children showed statistically significant development in all 12 parameters. The average age for diagnosis of autism is now 5.7 years, and LENA has the potential to drastically decrease that number. If a child is diagnosed with autism years earlier, it would most likely have dramatic positive effects on their development.
Up until now, vocal characteristics were not included in the standard diagnosis of autism, but this new study shows that automated analysis of vocalizations should become an accepted and accurate way of screening for autism. At the very least, the results can be used to alert parents that their child needs a full clinical diagnosis from a specialist.
Children that take advantage of LENA wear a small, sophisticated recording device in a pocket of their shirts, and the device records all of their verbalizations for an entire day. The device is able to remove outside noise, and the voices of other people. Nearly 1,500 all-day voice recordings were taken and analyzed for the study – about 3.1 million automatically identified child utterances.
The LENA system can also have far-reaching benefits for children with language delay, as the voice software analysis revealed differences for those children as well. It is expected that LENA will also be effective regardless of the language that is analyzed, because the physics of human speech are similar throughout the world, according to the researchers. D. Kimbrough Oller, a professor at the University of Memphis, led the study.