Scans of brain activity in babies as young as six months old could help physicians predict reliably the incidence of autism in babies the age of two. The study, though small, reveal findings which are considered game-changing for those babies who are at a high risk attributed to an older autistic sibling. This, in forthcoming years, will hopefully pave way for the early and affordable diagnosis of the disease in infants and consequently enable clinicians to subject them to timely behavioral therapy.
The study was published on June 7, 2017 in Science Translational Medicine, an interdisciplinary medical journal. Symptoms of the disease, a developmental disorder, surface in infants at age of two and the disease is diagnosed mostly at the age of four. The findings of the results have to be corroborated with further studies before they prove useful.
Differentiating Patterns Emerge as Early as Six Months
The study was funded by National Institute of Child Health and Human Development. The researchers chose a group of 59 babies at high risk to observe distinct patterns. They were able to correctly predict the prevalence of autism in nine out of 11 infants who later were found with the disease, with a significant accuracy of 97 percent.
The study, a part of the Infant Brain Imaging Study (IBIS), closely tracked the development of over 300 siblings. Joseph Piven, a Professor of Psychiatry at the University of North Carolina at Chapel Hill (UNC) and also one of the lead investigators of the project, along with his colleagues used functional connectivity MRI to detect differentiating brain patterns.
Affordable Way to Diagnose Autism for Worried Parents Soon
Robert Emerson, a researcher in involved in the study, and his colleagues measured synchronous activity across 26,335 pairs of brain regions. Once these children attained the age of two, parents were asked to fill questionnaires furnishing details on repetitive behavioral patterns of their children. The researchers then assessed language capabilities, motor skills, and communication skills in 59 infants and using machine-learning algorithm they were able to flag nine of the 11 children with autism.
The findings once replicated will help trigger early warning signals for the prevalence of the disease which has no single known cause. The high specificity of the algorithm used in diagnosis was a blessing in disguise for parents who are left clueless on the onset of the disease, since the defining features of autism become noticeable in toddlers only after a year. The disease become full-fledged only when the children are of two years of age, leaving them limited treatment options. In future, worried parents will have at their disposal minimally invasive and inexpensive methods to help them opt for early diagnosis and treatment.