Public Release: 

Detecting autism in infants before symptoms emerge

American Association for the Advancement of Science


IMAGE: Schematics representing brain scan signatures that predicted later autism diagnosis in infants. view more 

Credit: R.W. Emerson et al., Science Translational Medicine (2017)

According to the results of a new study, a brain scan can detect functional changes in babies as young as six months of age that predicts later diagnosis with autism spectrum disorder (ASD). An estimated one in 68 children globally are affected with ASD - a wide-ranging group of neurodevelopmental disorders that often cause ongoing problems with communication, repetitive behaviors, and other symptoms that impair an individual's ability to function socially. Early detection and behavioral interventions could significantly improve quality of life for people with ASD, but the full range of behavioral symptoms typically don't appear until children are two years old or later. Now, Robert Emerson and colleagues linked reliable measures of brain connectivity in six month-old infants to a later ASD diagnosis at two years of age with almost 100% accuracy. The scientists scanned the brains of 59 infants with high familial risk for ASD while they were sleeping, and collected data on 26,335 pairs of functional connections between 230 different brain regions using an imaging technique called functional connectivity magnetic resonance imaging (fcMRI). Of the 59 infants, 11 went on to be diagnosed with ASD at 24 months of age, which enabled the researchers to apply machine-learning algorithms to parse out specific brain patterns that correctly predicted nine of the 11 diagnoses without any false positives. Although future work is needed to determine if the signature applies to infants without high genetic risk, the authors say their findings may be first step towards much-needed early detection measures for ASD.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.