AI Predicts Autism From Infant Brain Scans

Share

Researchers from the University of Washington were part of a North American effort led by the University of North Carolina to use MRI to measure the brains of "low-risk" infants, with no family history of autism, and "high-risk" infants who had at least one autistic older sibling.

Ultimately, this algorithm was pretty good at predicting from the six- and twelve-month brain scans of the same group of children if the child would be diagnosed with autism. The study also assessed behavior and intellectual ability at each visit, using criteria developed by Estes and her team.

Using magnetic resonance imaging (MRI), researchers took brain scans at 6 months, 12 months and 24 months of children who were at high risk for autism because their older siblings had the disorder. It supports other research suggesting that overgrowth of certain brain regions is involved in autism. "The earlier an intervention is implemented, the better the outcome for kids with autism". The researchers note that the greater the brain overgrowth, the more severe a child's autistic symptoms tended to be.

The team also performed behavioral evaluations on the children at 24 months, when they were old enough to begin to exhibit the hallmark behaviors of autism, such as lack of social interest, delayed language, and repetitive body movements.

The analysis was most accurate in predicting the high-risk babies that did not develop autism, they added.

As an NIH-funded Autism Center of Excellence, the researchers' data and tools are open-source and will eventually be submitted to the NIH's National Database for Autism Research.

"In the field we are always trying to detect autism at younger ages, so we can start treatment earlier, but we hit a wall around 2 to 3 years of age, because the symptoms don't start showing up until around then", said the study's senior author, Joseph Piven, a professor of psychiatry, psychology and pediatrics at the University of North Carolina, Chapel Hill, in an interview with CNBC.

Despite extensive research, it has been impossible until now to identify these children before the second year of life, when behaviors typical of autism emerge.

Raptors trade Terrence Ross to Magic for Serge Ibaka
As for the Raptors, they've been in need of a game-changing move for weeks now, particularly at the power forward spot. That's why I have chosen Serge Ibaka of the Orlando Magic as the player the Toronto Raptors should pursue in a trade.

The findings could lead to an early test and even therapies that work while the brain is more malleable. "So for children like my Padric who's the youngest of four boys and the oldest has Autism, I mean that would be great if he were an infant and we had something we could go to that 'we will follow your child through these scans and developmental tests, ' and then be able to get you the earliest intervention we can if we see markers for Autism, that's unbelievable". The Center for Autism Research (CAR) at Children's Hospital of Philadelphia (CHOP) was a major study site in the multicenter research project.

The researchers made measurements of cortical surface areas and cortical thickness at 6 and 12 months of age and studied the rate of growth between 6 and 12 months of age.

However, by considering other factors as well including additional brain measurements and the child's sex, the researchers used a statistical approach known as machine learning to assess with near flawless accuracy who would develop autism. And the team was pleased with the prediction accuracy. It is estimated that one out of 68 children develop autism in the United States.

Elison has applied for a grant to conduct a new round of imaging studies in Minnesota, because he said this initial finding needs to be replicated with different children and different MRI techniques.

"This means we potentially can identify infants who will later develop autism, before the symptoms of autism begin to consolidate into a diagnosis", Piven said. The idea would be to then intervene "pre-symptomatically" before the emergence of the defining symptoms of autism.

By this point, however, fundamental developments in the brain have already occurred.

"So we find it very promising".

Share