Innovations enable earlier study of autism

Many children find it difficult to cooperate for lengthy neuroimaging protocols. As a result, today?s revolution in brain research could easily bypass children. But two innovations?the Geodesic Sensor Net and scanner preparation?are lowering the practical age limit for certain neuroimaging methods and giving scientists new ways to study some of the youngest people with autism.

The Geodesic Sensor Net, worn over the head, uses dozens of sponged-tipped electrodes to map patterns of the brain?s electrical activity, says psychologist Geraldine Dawson, PhD, at the University of Washington. This type of imaging was previously limited by the need to attach electrodes to the scalp?something youngsters do not tolerate well. The Geodesic Sensor Net uses saline solution as the conductor between scalp and electrodes, making brain studies essentially pain-free. Dawson and her colleagues are using the Geodesic Sensor Net to study social information processing in autistic children as young as age 2. The methodology involves showing the child a photo of his or her mother?s face or the face of an unfamiliar person, as well as showing the child a favorite object or an unfamiliar object. Preliminary results indicate that in autistic children, the system in the brain that allows people to process important social information?distinguishing Mother?s face from the face of a stranger, for example?is not working properly. The part of the brain responsible for recognizing objects, however, remains intact.

While the Geodesic Sensor Net provides a child-friendly technology, scanner preparation helps children become good research participants in studies using magnetic resonance imaging (MRI). Developed at Baltimore?s Kennedy Krieger Institute, this behavioral training technique enables all but the most unruly children to lie still for an MRI scan.

The institute has used scanner preparation with about 250 children as young as age 3 who have autism, profound mental retardation and a variety of developmental delay problems, says behavioral psychologist Michael Cataldo, PhD. Most had already failed an MRI scan?they moved too much to get a usable image?or were predicted to fail. In contrast, 'between 80 percent and 85 percent of the children get successfully scanned if they come for our preparation,' says Cataldo, director of behavioral psychology at the Kennedy Krieger Institute.

The researchers begin the preparation by training the children to lie still in a mock scanner. They attach a motion sensor to a child?s head, which is connected to a computer that monitors head motion and sends feedback to the child. When head movements are small enough, the child hears a tone and earns points that he or she can redeem for a toy. When head movements become too large, a buzzer sounds and the child loses points. Children may need one to 15 of the 30-minute training sessions before they can remain still, but the median number is two, Cataldo says. 'You can take kids who have head movement out of criteria a couple hundred times in the first session, and by the second session they may violate more stringent criteria two or three times.'

Researchers use a fantasy technique to help the children see the MRI machine as a fun, inviting place. To do this, they hide the device behind a plastic-foam facade with a Star Wars theme painted on one side and a Pooh Bear cave on the other, Cataldo says. The lightweight facade can be turned so that either side faces the child, who crawls through a hole in the facade and into a 'spaceship' or 'cave.'

So far, researchers use this training regimen only with structural imaging studies where the subject only has to lie still. With functional MRI studies, where the child must do a task without moving the head, training is more difficult. Cataldo says a training regimen for these protocols is still in pilot studies.

?Hugh McIntosh