Structural MRI offers insights into autism

Italian researchers are touting the possibility that structural MRI might be able to predict the development of autism spectrum disorders and help guide treatment options.

The conclusion comes after a study, published in the Journal of Neuroimaging, found that subtle changes in the grey matter of preschool children were able to classify the disorders.

“Despite the subtle impact of [autism spectrum disorders] on brain morphology and a limited cohort size, results from sMRI-based classifiers suggest a consistent network of altered brain regions,” wrote Ilaria Gori, of the University of Sassari, and colleagues.

The researchers analyzed structural MRI scans of two cohorts: male children with an autism spectrum disorder and controls matched for age and nonverbal IQ. They leveraged two types of commercial preprocessing packages (SPM and Freesurfer) to extract brain morphometric information, with support vector machines use to classify brain features and localize the regions that most differentiated the two cohorts.

Frontal and temporal areas associated with the theory of mind and empathy were implicated in those with autism spectrum disorders, with classification performance in the gray matter subregions reaching an area under the receiver operating characteristic curve of 74 percent.

"Since [autism spectrum disorder] is a common and lifelong impairing condition, a better knowledge of distinctive neuroanatomical structures underlying symptoms and cognitive profile is likely to improve not only the diagnostic process, but also the current treatment options," wrote the authors.