This artilce also appeared in the 2017, volume 3 issue of ARI’s Autism Research Review International newsletter.
A single brain scan can identify autism spectrum disorders (ASD) in six-month-old children with a remarkable degree of accuracy, according to a new study.
The study, by Robert Emerson and colleagues, involved 59 six-month-olds at high risk for ASD because their older siblings had the condition. The researchers used a type of scan called functional connectivity magnetic resonance imaging (fcMRI) to analyze how different brain areas connected with each other.
Focusing on brain connections linked to features of autism—including language, social behavior, and repetitive behaviors—the researchers identified 974 connections in the infants’ brains that could be used to predict an ASD diagnosis by 24 months of age. Using a “machine learning” program to refine their results, they predicted that nine of the infants would develop ASD.
Emerson comments, “When the classifier determined a child had autism, it was always right. But it missed two children. They developed autism but the computer program did not predict it correctly, according to the data we obtained at six month of age.”
The researchers say, “These findings have clinical implications for early risk assessment and the feasibility of developing early preventive interventions for ASD.”
“Functional neuroimaging of high-risk 6-month-old infants predicts a diagnosis of autism at 24 months of age,” R. W. Emerson, C. Adams, T. Nishino, H. C. Hazlett, J. J. Wolff, L. Zwaigenbaum, J. N. Constantino, M. D. Shen, M. R. Swanson, J. T. Elison, S. Kandala, A. M. Estes, K. N. Botteron, L. Collins, S. R. Dager, A. C. Evans, G. Gerig, H. Gu, R. C. McKinstry, S. Paterson, R. T. Schultz, M. Styner, B. L. Schlaggar, J. R. Pruett, Jr., and J. Piven, Science Translational Medicine, Vol. 9, No. 393, June 7, 2017. Address: Robert Emerson, firstname.lastname@example.org.
“Special brain scans may predict autism in high-risk babies,” MedlinePlus, June 7, 2017.
“A single brain scan has been used to accurately predict autism at just 6 months old,” Mike McRae, Science Alert, June 8, 2017.