Brain Scans Predict Autism In Babies Before Symptoms Show Up

This algorithm could predict autism in babies with more than 96 percent accuracy.

ADVERTISEMENT

Brain scans of six-month-old babies can help physicians diagnose Autism Spectrum Disorder with more than 96 percent accuracy, according to a new study in Science Translational Medicine. The findings are especially significant because, until now, researchers have struggled to diagnose autism in children younger than age two, in part because the disorder is defined as a collection of behaviors that even healthy infants do not display.

“This new paper focused on how brain regions are synchronized with each other at one-time point (six months) to predict at an even younger age which babies would develop autism as toddlers,” coauthor on the study Joseph Piven of the University of North Carolina School of Medicine said in a statement. “The more we understand about the brain before symptoms appear, the better prepared we will be to help children and their families,”

Piven and colleagues looked at MRIs from 59 six-month-old babies who had siblings with autism (and were therefore at higher risk for the disorder) obtained from the Infant Brain Imaging Study. Then they measured the so-called “synchronous activity” between 26,335 pairs of brain regions, a metric that is thought to signal which regions of the brain are most strongly connected. Later, when the babies were two-years-old, researchers follow-up with parents and asked them complete questionnaires about each child’s social behaviors, language abilities, and motor skills. Based on these questionnaires, Piven and his team diagnosed 11 out of the 59 children with autism.

The team then entered their brain activity figures and behavioral test scores into a machine-learning algorithm. The algorithm accurately predicted whether specific patterns of brain activity at six months old would be linked with an autism diagnosis at age two, and predicted autism in nine out of the 11 kids diagnosed with the disorder. Although the algorithm missed two children, it also did not mistakenly diagnose any healthy children with autism. The new study is a fitting follow-up to previous work by Piven and colleagues, which identified differences in brain anatomy that can predict autism in toddlers. This new method rounds out the team’s approach to early autism diagnosis.

Although the results are promising, it is important to note that there were only 11 kids involved in the study group (and essentially 48 controls). Meaningful inferences cannot be drawn from such a small sample size and, before experts can gauge what this means for special needs families, the results will need to be replicated on a larger scale.

In any case, “predicting autism as a category isn’t necessarily that useful,” Emily Jones of the Centre for Brain & Cognitive Development at Birkbeck, University of London, who was not involved in the study, told Scientific American. For Jones, the next step must be figuring out what “synchronous activity” means for early brain development, and how specific patterns may forecast future disabilities. “What you want to do is predict which children are going to have more difficulties, or the kinds of difficulties that they might need early intervention for,” she says.

Get Fatherly In Your Inbox


Survey Callout Image