A new technique that identifies early differences in vocal development between children with an autism spectrum disorder or language delay and those developing on a normal trajectory could give pediatricians and other caregivers a tool for earlier detection of autism, and as a result facilitate earlier intervention. To distinguish the vocal patterns of normal infant and child development from those of children with autism or a speech delay, a team of researchers led by D.K. Oller of the School of Audiology and Speech-Language Pathology at the University of Memphis recruited 232 children between the ages of 10 months and four years — including 77 who had been diagnosed with autism, 49 with identified language delay and 106 whose development was characterized as typical. Children were outfitted with small, portable recording devices fastened to their clothing and researchers collected more than 1,400 all-day recordings. After analyzing the recordings and excluding cries or other sounds, the researchers identified a total of 3.1 million individual utterances.
The researchers than analyzed the utterances for 12 different characteristics known to play a role in speech development — including features like pitch and ability, even in baby talk, to break sounds into syllables. The researchers then used a sophisticated algorithm to analyze each individual utterance for these 12 characteristics. They found a clear correlation between children’s vocalizations and individual speech characteristics and overall development. That is, for both the typically developing and language-delayed groups, children consistently showed high levels of development in many of the vocal parameters; in the autistic group, however, children’s vocalizations suggested little development on the 12 speech characteristics used to track vocal development.
The study, published this week in the Proceedings of the National Academy of Sciences suggest that this vocal analysis system, known as LENA (Language Environment Analysis) could help care providers detect signs of autism in very early stages of development. As the authors conclude:
“Based on the results reported here, there appears to be little reason for doubt that totally automated analysis of well selected acoustic features from naturalistic recordings can provide a monitoring system for developmental patterns in vocalization as well as significant differentiation of children with and without disorders such as autism or language delay.”
Though this is just the first study to incorporate the automated vocal analysis tool, researchers are hopeful that subsequent research will add to these findings — and enable them to tweak the system and ultimately enable pediatricians and other care providers to utilize it as a tool for early autism screening.
