AI Detects Autism Speech Patterns Across Different Languages

A recent study lead by Northwestern University researchers utilised machine learning, a type of artificial intelligence, to detect speech patterns in children with Autism that were consistent in both English and Cantonese, implying that speech characteristics might be a viable tool for diagnosing.

The research, which was carried out in collaboration with Hong Kong-based collaborators, yielded insights that could help scientists distinguish between genetic and environmental factors that influence people with Autism communication abilities, potentially allowing them to learn more about the condition’s origins and develop new therapies.Youngsters with  frequently speak more slowly than typically developing children and have additional pitch, intonation, and rhythm problems.

However, such variances (which experts refer to as “prosodic differences”) have proven unexpectedly difficult to quantify in a consistent, objective manner, and their origins have remained a mystery for decades.However, supervised machine learning was successfully used by a team of researchers led by Northwestern scientists Molly Losh and Joseph C.Y. Lau, as well as a Hong Kong-based collaborator Patrick Wong and his team, to identify speech differences associated with Autism. The results were published in the journal PLOS One on June 8, 2022.

“When you have languages that are so structurally different, any similarities in speech patterns seen in Autism across both languages are likely to be traits that are strongly influenced by the genetic liability to Autism,” said Losh, who is the Jo Ann G. and Peter F. Dolle Professor of Learning Disabilities at Northwestern. “However, the variability we saw is as intriguing, since it may hint to more flexible characteristics of speech that might be ideal targets for intervention.”