Multivariate pattern analysis of MRI brain-scan data differentiated children and adolescents with autism from neurotypical children, according to a study published online in the September issue of Biological Psychiatry.

The results showed distinct structural abnormalities in the brains of autistic youth, which could eventually lead to a new process for diagnosing autism spectrum disorders (ASD) as well as the development of early interventions.

The study, which was led by Vinod Menon, PhD, director of the Cognitive and Systems Neuroscience Laboratory, at Stanford University School of Medicine in Stanford, Calif., noted differences in gray matter structure in autistic children compared to typically developing counterparts, particularly in the Default Mode Network (DMN).

The DMN includes systems that discern self-relevant information, allow for self-reflection and regulate emotion. Autism occurs in about one in every 110 children and affects social interactions and self-perception, which makes the DMN a logical place to look for structural abnormalities.

The study analyzed MRI data from 24 autistic children ages 8 to 18 years and compared that with data from 24 age-, gender- and IQ-matched neurotypical children.

Using a multivariate pattern analysis of voxel-based morphometry data obtained from the study participants, Menon et al were able to distinguish between autistic and neurotypical children with approximately 90 percent accuracy.

The analysis method, multivariate searchlight classification, divided the brain with a 3D grid, then examined one cube of the brain at a time, and identified regions in which the pattern of gray matter volume could be used to discriminate between children with autism and typically developing children.

While more research is needed, similar image analysis could eventually become a diagnostic tool.

“One of the major impediments to progress in understanding ASD results from the fact that it is currently diagnosed solely on the basis of behavioral characteristics,” wrote Menon and colleagues. “Findings from the current study and similar efforts integrating other types of neuroimaging data may eventually lead to the identification of robust brain-based biomarkers with the potential to aid in early detection and intervention in children with ASD.”

The advantage of a multivariate pattern analysis over traditional univariate approaches is that it is more sensitive to subtle changes in multiple brain areas affected by autism, according to the authors.

Data from the analysis were also able to provide a gauge of autism symptom severity in addition to simply distinguishing between autistic and typically developing children.

Menon et al acknowledged previous studies that reported gray matter data could be used to classify individuals as autistic, but noted this study is the first to identify the specific loci of gray matter differences in children and adolescents.

The researchers plan to repeat the study in younger children and extend it to larger groups of subjects. If the results are upheld, brain scans might eventually help distinguish autism from other behavioral disorders or might predict whether high-risk children will develop autism.

“Scans would likely be used alongside clinical expertise, giving that extra hint from the brain data,” Lucina Uddin, PhD, an instructor in psychiatry and behavioral sciences at Stanford, said in a statement.