Authors: Alice Bough, Future Science Group
A team of researchers from Stanford University School of Medicine (CA, USA) have devised a new method for diagnosing autism in children. The paper, which has been published in PLOS Medicine, utilized machine learning to provide a diagnostic score based on evaluations of behaviors exhibited by children in short home videos. This novel method could decrease the time required to diagnose this behavioral and developmental disorder.
Individuals living with autism often have difficulty forming social connections, exhibit repetitive behaviors and can have restricted interests. It has been established previously that behavioral therapies for autism are most effective when started before the age of 5. Early diagnosis of the disorder is therefore crucial.