Using MRI to Detect Autism Biomarkers in Infants

Researchers at the University of Washington have used magnetic resonance imaging (MRI) to identify potential biomarkers of autism in the brains of infants whose older siblings have been diagnosed with the condition. Using the brain imaging technique, the researchers showed 80 percent accuracy in predicting which children would be diagnosed with autism at 24 months.

The researchers used a computer algorithm to analyze the MRI scans of infants with no family history of autism (low risk), and those who had at least one sibling with the condition (high risk). The results of the study were published in the journal, Nature.

The study is the first of its kind to show that brain biomarkers could be useful in predicting which infants in the high risk group will be diagnosed with autism spectrum disorder at age two. The findings could aid in the development of a diagnostic to help clinicians diagnose autism even before behavioural symptoms emerge.

“Typically, the earliest we can reliably diagnose autism in a child is age two, when there are consistent behavioral symptoms, and due to health access disparities the average age of diagnosis in the US is actually age four,” said Annette Estes, University of Washington professor of speech and hearing sciences. “But in our study, brain imaging biomarkers at 6 and 12 months were able to identify babies who would be later diagnosed with autism spectrum disorder.”

Three million individuals in the US alone have been diagnosed with autism spectrum disorder, which has noticeable effects on their communication abilities and social behaviours. Autism is believed to occur in as many as 68 births in the US, however children with an autistic older sibling face an even higher risk of developing the condition.

Hundreds of children across multiple study sites underwent MRI scans at six, 12 and 24 months of age. The behaviour and intellectual abilities of the research participants were also assessed at each study visit.

From six to 12 months, infants who went on to develop autism showed a hyper-expansion of brain surfaces, compared to those with a family history of the condition who themselves did not develop it. By imputing the MRI brain volume calculations as well as other criteria into the computer algorithm, the researchers were able to accurately predict which children would be diagnosed with autism spectrum disorder at two years of age.

“By the time ASD is diagnosed at two to four years, often children have already fallen behind their peers in terms of social skills, communication and language,” said Estes. “Once you’ve missed those developmental milestones, catching up is a struggle for many and nearly impossible for some. Our hope is that early intervention—before age two—can change the clinical course of those children whose brain development has gone awry and help them acquire skills that they would otherwise struggle to achieve.”