Communications of the ACM
AI Can Spot Signs of Autism in Babies
Researchers from Boston University, Boston Children's Hospital, and the University of San Francisco have developed a method of using artificial intelligence to read electroencephalograms to determine the likelihood that a baby has some degree of autism sp
Credit: Shutterstock
Researchers at Boston University, Boston Children's Hospital, and the University of San Francisco have discovered a way to interpret electroencephalograms (EEGs) using artificial intelligence to determine the chances of a child having autism spectrum disorder (ASD).
The team examined data from the Infant Screening Project, using algorithms created by University of San Francisco professor William Bosl. His research suggests that even an EEG that appears normal holds data about brain function, connectivity patterns, and structure that human practitioners cannot detect.
The algorithms analyzed six EEG frequencies to measure differences in the babies' brains and how they processed information.
In over 95% of cases, the algorithms predicted a clinical diagnosis of ASD with "high specificity, sensitivity, and positive predictive value."
For babies aged three months to nine months, predictive accuracy was almost 100 %.
From Alphr
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA
No entries found