Technology has made it possible for many activities to take place in the comfort of your own home: communication, education, even medical treatment. Now a team led by the University of Memphis’ Kimbrough Oller, professor of audiology and speech language pathology, has added diagnosis of disorders such as autism to the list.
The researchers built on existing diagnostic tools, namely the LENA (Language Environment Analysis) system, a device that labels vocal recordings. Oller’s team then added an automatic analysis based on 12 vocal development parameters.
The result? A tool that allows automated analysis of massive quantities of vocal recordings, eliminating countless hours of tedious work.
“A manual analysis would require such critical and exhaustive attention that it would never get done,” Oller says.
Oller’s team studied more than 1,000 recordings from 200 children. They found that the sounds autistic children make before learning to speak are different from those of children with more normal development — and distinct enough to be distinguished and diagnosed through the automatic system.
Diagnosis of autism now depends largely on the assessment of parents and doctors. Adding an objective diagnostic tool is a potential game-changer, Oller says.
“There are all sorts of precursors to this effort that will get folded into the project,” Oller says, noting the possibility of earlier and more widespread diagnosis for diseases with vocal indicators. “What this does is add a completely objective method to supplement other methods.”
Equally important is how easily the tool can be used. The LENA system is small, affordable, and available for use in households.
“How well it’s worked has surprised all of us,” Oller says. “This invention was bound to occur in the future, but it’s incredible that we’re already at a level where it’s practical to use.”
