Diagnosing autism with MEG imaging

Children's Hospital of Philadelphia researchers find that a pattern of brain response visible via magnetoencephalography may be the first imaging biomarker for autism.

Elizabeth Armstrong Moore
Elizabeth Armstrong Moore is based in Portland, Oregon, and has written for Wired, The Christian Science Monitor, and public radio. Her semi-obscure hobbies include climbing, billiards, board games that take up a lot of space, and piano.
Elizabeth Armstrong Moore
2 min read

Because children with autism spectrum disorders tend to process sound and language a fraction of a second slower than children without the disorders, researchers have discovered that measuring magnetic signals that mark this kind of delayed response has the potential to become a standardized tool for diagnosing autism.

A child reads instructions on a screen while seated with his head surrounded by the MEG's noninvasive magnetic detectors. The Children's Hospital of Philadelphia

"More work needs to be done before this can become a standard tool, but this pattern of delayed brain response may be refined into the first imaging biomarker for autism," said Dr. Timothy P.L. Roberts, vice chairman of Radiology Research at The Children's Hospital of Philadelphia and head of the study.

As with diagnosing many neurodevelopmental disorders without biomarkers, psychologists and caregivers use clinical, but still subjective, judgments to diagnose a range of childhood disorders affecting as much as 1 percent of U.S. children. Standardized imaging could help diagnose disorders at an earlier age and differentiate between disorders across the autism spectrum.

In its study, Roberts' team used magnetoencephalography, or MEG, to detect magnetic fields in children's brains, much the way electroencephalography, or EEG, detects electrical fields.

Using a helmet that surrounded each child's head, the team played a series of recorded beeps, vowels, and sentences. As each child's brain responds to each sound, noninvasive detectors in the MEG machine analyzed the brain's changing magnetic fields.

The researchers compared 25 children with ASDs to 17 typically developing children; the group's mean age was 10. The children with ASDs responded to sound 11 milliseconds (only about a hundredth of a second) slower than the control children did. Within the group already diagnosed with ASDs, the delays were similar regardless of language impairments.

Roberts said that while an 11-millisecond delay is brief, it suggests that a child with ASD, on hearing the word "elephant," is still processing the sound "el" while other children have moved on: "The delays may cascade as a conversation progresses, and the child may lag behind typically developing peers."

Roberts added that his team will continue to refine their imaging techniques in order to determine whether this biomarker is specific to ASDs, and it will investigate other MEG patterns found in children with ASDs beyond auditory delays.

The findings, collected at The Children's Hospital of Philadelphia, are scheduled to appear Friday in the journal Autism Research.