From Washington University School of Medicine
New glimpse into the pediatric brain
St. Louis, — Brain imaging works well in children, too, according to new research at Washington University School of Medicine in St. Louis. The results are good news both for scientists studying brain development and for pediatric patients with neurological problems, says principal investigator Bradley L. Schlaggar, M.D., Ph.D., instructor of neurology and pediatrics.
“Most researchers assume that you can’t use imaging techniques to directly compare kids and adults, and that kids won’t sit still long enough to get accurate images,” says Schlaggar. “We took that as more of an empirical question: Can you image kids as readily as adults? We found that yes, we can.”
The ability to image healthy children will allow researchers to create a template of the normal pediatric brain and study how it develops into a healthy adult brain. Such a template also could help scientists understand what goes wrong in diseases that affect children’s brains.
Schlaggar’s team will present three abstracts at 9 a.m. PT on Nov. 13 at the Society for Neuroscience’s 31st Annual Meeting in San Diego. Collectively, these findings provide evidence that it is feasible to use functional magnetic resonance imaging (fMRI) to study the pediatric brain and directly compare it to the adult brain. FMRI is an imaging method developed in the last decade that allows scientists to take pictures of the brain while it is processing information to see which specific brain regions are involved in different tasks.
Experts believed it was difficult if not impossible to compare brain activation between children and adults and to view images from the two populations against the same functional map of the brain. To address this issue, Schlaggar and his colleagues combined several strategies to accommodate children in an imaging study.
First, the researchers asked participants to say their responses out loud. Verbal responses are critical for studying children’s performance compared with that of adults in order to properly compare the two groups’ responses.
“In most studies, participants think the response to themselves, and you have no idea what they’re doing,” says Schlaggar. “Kids are probably even less reliable than adults as they may be more distracted by the scanner. They could be rhyming when you want them to think of an antonym, or they could be thinking about the ice cream sandwich they had at lunch.”
Since responses were verbal, Schlaggar’s team could flag images from children and adults when they responded correctly to a given trial and in roughly the same amount of time. By matching participants based on accuracy and speed, the researchers argue that their comparisons reveal maturational differences and similarities in how the brain processes these tasks as opposed to differences in how well the individual performs the tasks.
“Differences between kids and adults may simply result from the fact that kids don’t do the task as well,” explains Schlaggar. “You have to find ways to get the level of performance equated so you can discern true processing differences.”
Because talking forces the head to move, many scientists believe speaking during fMRI detracts from the clarity of the images. This is the first study that required children to speak while in the imaging scanner and is one of the first studies to require adults to speak aloud while being scanned. This feat was made possible by a second strategy the team employed, a relatively new method of analyzing imaging results called event-related fMRI, developed in part by researchers at Washington University.
Normally during an fMRI study, participants repeat a task several times and the results from the entire set of repetitions are averaged together to produce an image. Any head movements during the set of repetitions will cloud the final image. Since speaking requires significant head movement, speaking tasks rarely are used. Schlaggar’s team used event-related fMRI to analyze each repetition individually.
They reasoned that head movements while speaking occur before the brain activity changes register on an fMRI image. Using this event-related design, the researchers therefore reasoned that they could minimize the problem of head movements in their calculations.
The team tested these strategies on 19 children (9 years old on average) and 21 adults (an average of 25 years old). Each participant either saw or heard a word while lying in an fMRI scanner. At different times during the study, both groups were asked to do one of three things: say a word that rhymes with the one they saw or heard; say a word whose meaning is opposite that of the word they saw or heard; or say a verb whose meaning is related to the word given.
Schlaggar’s team found that these new strategies did in fact yield clear images sufficient for studying activation patterns. In their abstracts, they highlight several of these new observations.
In the first abstract, brain images from 10 children and adults matched for performance looked similar overall, but there were several key differences. Some regions, such as the left dorsal frontal cortex, located toward the front of the brain, were active only in adults, while other regions, such as those involved in vision, were much more active in children. This implies that children and adults use different brain strategies to perform the same task.
In the second abstract, Timothy T. Brown, a graduate student with Schlaggar’s team, describes another intriguing difference between the two groups. A brain region toward the front of the brain called the medial frontal cortex appeared to be comprised of three distinct sub-regions, each with different brain responses in children versus adults doing the same task.
One region became more active during the task in children and was less active than normal in adults; a second region that also was more active in children did not show any response in adults; a third region was less active than normal in both groups, but the result was more drastic in children.
“We’re still trying to identify where the boundaries are between functional areas of the brain,” says Schlaggar. “Tim’s study shows one way that we can use developmental relationships to help clarify the organization of the brain.”
Another problem that has prevented functional comparisons between the two groups has been the assumption that brain shape and variability are too great in children to be directly compared with adults. In the third abstract, H.S.C. Kang, also a graduate student on Schlaggar’s team, found that most brain regions are only a few millimeters apart. This difference is too small for fMRI to detect, making the comparisons possible.
“Once you adjust for these identified pitfalls and use these strategies to address them, you can overcome perceived barriers and get stable, consistent, reliable results that will help us understand how the brain works in children,” says Schlaggar. “But most important, we can start to understand how a child’s brain responds to trauma and disease so we can help pediatric neurology patients recover as much function as possible.”
Schlaggar BL, Brown TT, Lugar HM, Coalson RS, Miezin FM, Petersen SE. Children and adults show systematic differences in dorsal frontal and extrastriate regions during lexical processing. Society for Neuroscience, Nov. 13, 2001.
Brown TT, Lugar HM, Coalson RS, Miezin FM, Petersen SE, Schlaggar BL. Children and adults show systematic differences in brain areas involved in lexical processing: Evidence for three functionally distinct regions in medial frontal cortex. Society for Neuroscience, Nov. 13, 2001.
Kang HSC, Burgund ED, Lugar HM, Petersen SE, Schlaggar BL. Atlas transformation of pediatric brains: Comparison of functional activation foci in children and adults. Society for Neuroscience, Nov. 13, 2001.
Funding from the National Institutes of Health (including the Child Health Research Center in Developmental Biology) and the McDonnell Center for Higher Brain Function supported this research.