"This is by no means a definite way to detect the disorder, the researchers say..."
True, not definite, but 97% accuracy sounds pretty close! How effective are other methods of detecting autism?
When I was an undergraduate at the University of Washington, I assisted in an EEG laboratory at the UW Autism Center. In EEG (electroencephalogram) testing a person wears a net cap full of electrodes encased in damp sponges. Imagine trying to place 125 damp sponges at specific skull coordinates on a three-year-old with autism, or any three-year-old for that matter. That requires a lot of gluten-free shortbread cookies and M & M’s.
What EEG is really good at is recording time-specific brain activity in response to a stimulus. It’s not as spatially accurate as an fMRI, but it can tell you exactly when brain activity happens in a rough area. A researcher can watch and record the activity live, as shown in waveforms on a computer screen. Each waveform represents the activity recorded at a different point on the skull.
At our lab we looked at people’s EEG waveforms in response to pictures of things like faces and houses. We were interested in how people with autism have different brain activity reactions to images than typical people. For example, the usual “face response” waveform a person shows in response to a face picture is smaller in people with autism than it is in typical individuals.
Today a team of researchers at Boston Children’s Hospital reports in BMC Medicine that they have identified a pattern of EEG responses that distinguish children with autism from their typically developing peers.
The team conducted EEG recordings of 463 children with autism and 571 “neuro-typical” children. They found that by analyzing a combination of 33 specific brain activity factors from the recordings, they had 97% success rate in correctly diagnosing a child with autism (when looking the data as divided by age ranges).
This is by no means a definite way to detect the disorder, the researchers say, but it does offer one more tool to promote early diagnosis. Michelle Roberts of BBC News reports:
Dr. Frank Duffy who is leading the investigation said the work could [..] point the way to determining if younger siblings of children with autism are likely to develop the same condition themselves.
“It is a great cause of anxiety when an older sibling develops autism. EEG might offer a way to check for the same condition in younger siblings in advance of them having symptoms.”
The technique was found to be useful in detecting autism in children as young as two, which is helpful since earlier intervention has been shown to improve social functioning in children with autism.
I don’t envy the people now responsible for getting that moist sponge cap on even more two-year-olds’ heads.
Photo: SCA Svenska Cellulosa Aktiebolaget/Flickr
