It is well known that the earlier a child can be diagnosed with Autism Spectrum Disorder (ASD), the more effective treatment can be. This is not to say autism can be cured through earlier diagnosis, but rather, early interventions can better help those children become more independent adults. For the first time, researchers at the University of North Carolina have used MRIs of six-month-olds to successfully predict which babies were at a high risk of developing autism as toddlers.
ASD symptoms generally don't appear in a child until the age of two, and it's often many years later before official diagnoses are even made. Because doctors don't have any solid biological way to identify the condition they have to rely on behavioral observations for diagnosis. This can mean that many children often go several years before being diagnosed and starting treatments.
GET 30% OFF NEW ATLAS PLUS
Read the site and newsletter without ads. Use the coupon code EOFY before June 30 for 30% off the usual price.BUY NOW
Researchers have been hunting for a biomarker to help identify the condition, and we have seen several early indications of success from saliva tests and blood tests, to even a smartphone app that can read a child's facial expressions.
This new study from UNC researchers follows on from earlier research that looked at how different brain regions synchronized with each other. The previous study measured brain development at both six and 12 months of age, but this new study concentrated on just the six-month time point to allow for earlier diagnosis.
The study examined 59 babies, each with an older sibling with autism, meaning they have a higher chance of developing the condition. The sleeping babies were scanned in an MRI machine and the researchers studied the neural activity across 230 different brain regions.
An illustration of the 200 different connections the researchers studied to predict a baby's onset of autism (Credit: UNC)
The coordinated activity between these regions has been identified as being crucial for cognition, memory and behavior. A machine learning algorithm was then developed to sort through the differences in synchronization across these regions. With 11 out of the 59 babies ultimately developing autism symptoms, the software correctly identified 81 percent of the babies that would go on to develop autism by age two.
"When the classifier determined a child had autism, it was always right," says first author of the study, Robert Emerson. "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 months of age."
Broader studies still need to be done to replicate these results, but this is just one of several diagnostic tools that researchers at the UNC are currently working on. They are also investigating the connection between increased cerebrospinal fluid and autism, as well as other brain network functional connections that could signal ASD behavior.
"I think the most exciting work is yet to come, when instead of using one piece of information to make these predictions, we use all the information together," Emerson says. "I think that will be the future of using biological diagnostics for autism during infancy."
The new study was published in the journal Science Translational Medicine.View gallery - 2 images