Autism affects roughly 1 out of 68 children in the US. While early diagnosis and intervention can help children with autism, there’s so far no method for diagnosing it before children start showing symptoms. However, I recently read about one study funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) has discovered that functional connectivity magnetic resonance imaging (fcMRI) could be used to predict autism among infants.
As previous findings suggest that brain-related changes occur in autism before behavioral symptoms emerge, detecting brain differences may help enable physicians to diagnose and ultimately treat autism early. This current study, led by researchers at UNC Chapel Hill and Washington University in St. Louis, was focused on the brain’s functional connectivity, or how different regions of the brain work together. The researchers used fcMRI to scan 59 high-risk 6-month-old infants while they slept. These children were considered “high-risk” since they had older siblings with autism. At 2 years of age, 11 of them were ultimately diagnosed with autism.
For their study, the researchers used a computer-based technology called “machine learning”, which trains itself to look for differences that can separate neuroimaging results into two different groups (autism and non-autism), then predict future diagnoses. One analysis predicted each infant’s future diagnosis by using the other subjects’ data to train the computer program. Through this, it was able to identify 82 percent of the infants who would go on to have autism, and correctly identified all of those who didn’t end up developing autism. Another analysis tested how well the results could apply to other cases, in which case the computer program had an accuracy of 93 percent.
While these findings are still at an early stage, and far from fully developed, this study, according to the researchers, suggests that neuroimaging could be a useful tool in diagnosing autism and helping healthcare providers evaluate a child’s risk. The team overall found 974 functional corrections in the brains of 6-month-olds that were associated with autism-related behaviors. The authors have proposed that one neuroimaging scan can accurately predict autism among high-risk infants. However, their results need to be replicated in a larger group before anything can be considered definite.