Scientists Use Scans to Better Understand Brain Maturity


As children mature, their brains bloom madly with activity — growing, pruning and rewiring. During this process, some connections strengthen, while others die off. The brain is reorganizing itself with experience.

Researchers from the Washington University School of Medicine have created a “virtual machine” to study such brain development and to track young brains as they develop. Their hope is that this method may one day be used to identify neuropsychiatric disorders in children.

The researchers relied on functional MRI scans to study simultaneous activity occurring in different regions of the brain. When groups of nerve cells from different brain areas fire together, this indicates neural connections between these regions. “This is the ‘what fires together wires together’ idea writ large,” says Bradley Schlaggar, a pediatric neurologist and senior investigator of the study, which was published Thursday in the journal Science.

The researchers mapped hundreds of these brain regions in people that ranged in age from 7 to 30 years. Then they fed the data to a computer and looked for patterns in the scans. Based on these patterns, they trained the machine to kick out a measurement of the brain’s maturity level.

Earlier studies have showed that as the brain develops, many of the short-range connections – neurons wired to other neurons in neighboring brain regions – are lost, and replaced with long-range, more faraway brain connections.

“In aggregate, we were able to use the measurements of hundreds of regions in the brain to build a machine to tell the next person, are you a child or are you an adult, and where are you in the maturation curve,” Schlaggar said.

MRI’s have been notoriously bad at detecting disorders like autism and epilepsy in children. But most have relied on structural MRI’s, which image the brain’s architecture, not its function. Functional MRI’s measure blood-oxygen levels, and are better indicators of electrical activity in the brain.

“We wanted to get a good baseline to understand how normal works before we could understand what goes wrong,” says Nico Dosenbach, the study’s lead author. A brain found veering off the course of its age-appropriate maturity level could indicate an underlying problem.

Commonly, when a child is brought to the hospital after having a seizure, a structural MRI will come back normal. Other tests will determine if the child has epilepsy. “The promise of this kind of approach,” Schlaggar says, “is that lurking in that scan is a ton of untapped information that will give us greater insight into the patient that we’re taking care of.” An additional five minutes of data from a functional MRI, he adds, should tell doctors if there’s a pattern of epilepsy, autism or Tourette syndrome or another disorder.

A brain-age metric like this could potentially help doctors distinguish between a neurological delay and a disorder, says Russell Poldrack, professor of psychology and neurobiology at the University of Texas at Austin. “The hope is that we’ll be able to train statistical machines to do this kind of diagnosis.”

But he cautions that more studies need to be done to determine the success of resting-state MRI’s, or scanning people in a resting state, as this study did. “It may well be that not everything we need to know is contained in the brain activity when people are rest. We don’t yet know what the limitations are of this resting state MRI.”

A next step will be to take this functional connectivity data and study how closely related it is to the brain’s structure. But ultimately, the researchers hope to use the baseline they’ve developed to learn more about brain disorders and eventually be able identify them earlier.

“Let’s say you have a family who has one child with autism, and wants to know whether the next child is going to have autism,” Schlaggar says. If a scan could predict the likelihood of autism for that child, he continues, “maybe the interventions that are most effective could be applied before any symptoms develop. Early identification could be really helpful.”