Autism Biomarker Candidates
Some groups of researchers believe that at least some cases of ASD are caused by problems in neural connectivity. Along these lines, a group of researchers measured brain activity using electroencephalography (EEG), a technique for non-invasively recording and mapping electrical activity in the brain, in typically developing and high-risk groups of infants.1 Using several machine learning algorithms, the researchers found that infants with high risk for autism were classified with an accuracy of more than 80% at 9 months of age. This study suggests that electrical activity in the brain can be used as a predictor of autism risk at 9 months of age.
In the study, however, the classification of girls as high risk did not reach as high a level of accuracy as the classification of boys, and the outcomes of the high-risk infants were not reported. Future studies should reveal whether infants classified as high-risk by EEG measurements actually go on to develop ASD.
Because EEG is non-invasive and is relatively inexpensive compared with magnetic resonance imaging (MRI), EEG has the potential to be used in more large-scale screenings of infants.