Everyone’s brain is different. Until recently neuroscience has tended to gloss this over by averaging results from many brain scans in trying to elicit general truths about how the organ works. But in a major development within the field researchers have begun documenting how brain activity differs between individuals. Such differences had been largely thought of as transient and uninteresting but studies are starting to show that they are innate properties of people’s brains, and that knowing them better might ultimately help treat neurological disorders.
The latest study, published April 8 in Science, found that the brain activity of individuals who were just biding their time in a brain scanner contained enough information to predict how their brains would function during a range of ordinary activities. The researchers used these at-rest signatures to predict which regions would light up—which groups of brain cells would switch on—during gambling, reading and other tasks they were asked to perform in the scanner. The technique might be used one day to assess whether certain areas of the brains of people who are paralyzed or in a comatose state are still functional, the authors say.
The study capitalizes on a relatively new method of brain imaging that looks at what is going on when a person essentially does nothing. The technique stems from the mid-1990s work of biomedical engineer Bharat Biswal, now at New Jersey Institute of Technology. Biswal noticed that scans he had taken while participants were resting in a functional magnetic resonance imaging (fMRI) scanner displayed orderly, low-frequency oscillations. He had been looking for ways to remove background noise from fMRI signals but quickly realized these oscillations were not noise. His work paved the way for a new approach known as resting-state fMRI.
This type of scan, it turns out, reveals a lot about a particular brain. It analyzes the commonplace slow fluctuations of neural signaling, which form networks of brain cells that fluctuate in synchrony—and these networks often resemble those the brain engages when it is actively doing something. “We’ve known for awhile that the brain networks we pull out of resting-state data look similar to the maps we get from task-induced activity,” says neuroscience doctoral student Emily Finn of Yale University. Finn and her colleagues published a study last October showing that brain networks contain enough information to identify individuals with up to 99 percent accuracy. “This study takes things a step further,” Finn says.
The team behind the new study, led by neuroscientists Ido Tavor and Saad Jbabdi of the University of Oxford, used data collected by the Human Connectome Project (HCP)—a National Institutes of Health collaboration that is trying to map the wiring of the human brain and is led by Washington University in Saint Louis, the University of Minnesota and Oxford University. The team obtained data for 98 healthy young adults, including scans taken while the participants performed tasks involving memory, motor functions, decision-making (gambling), language (reading) and others as well as just resting. They analyzed the relationships between participants’ resting-state brain activity and the oscillations that emerged while they were engaged in various undertakings. They then tried to predict brain activity profiles for a given participant on each of the tasks, using only the individual’s resting-state scan. The predictions matched the brain activity of that person more closely than any of the other participants’ scans. “We extract a set of images that highlight brain areas that fluctuate together during this mind-wandering state,” Jbabdi explains. “Our study shows that these co-fluctuations contain enough information to predict how the brain behaves when it is actually doing something explicit.”
These are only first steps. What other information might be contained in the resting-state scans, and how the relationship between resting and active states might change under some circumstances, remain open questions. “It will be interesting to see if and how this mapping relates to actual performance on the tasks,” Finn says. “And how it changes with factors like age or neuropsychiatric illness.”
Tavor says his group was impelled to do this study by a common problem neuroscientists face. For many studies, researchers need to know exactly which brain areas are chugging along during certain tasks—so (for instance) they can see what happens when they block or enhance that activity. The new technique could allow researchers to predict where these regions are without having to conduct a separate scan for each of the tasks, saving time and money. “It’s a very practical result,” Finn says. “Resting-state could eventually serve as a “one-size-fits-all” scan from which we can glean a lot of information about someone, without actually having them sit though multiple task sessions in the scanner,” she adds.
One of the next endeavors in this research is to determine whether these findings hold not just for the healthy participants used in this study but for patients with various illnesses. “We’re looking at brain tumor patients before surgery,” Tavor says. Knowing what parts of the brain are responsible for sensitive functions, like language, can be crucial information to a neurosurgeon, and tumors can cause shifts in where functions are performed in the brain. “If we can predict this shift, it could affect the surgeon’s strategy of where to enter to remove the tumor,” Tavor explains.
Biswal is also interested in medical implications. “In clinical cases, if there’s a difference in performance, compared to healthy controls, would the resting-state still predict patients’ performance?” he asks. “Or has something mechanistic happened that means the prediction won’t be as good, and might this tell us something about the underlying mechanism of the disease?” Using the technique for diagnostic applications might enable researchers to measure disease severity by examining the accuracy of predictions for brain functions known to be affected by a particular disease.
Whatever the eventual outcome, this work adds to a body of evidence suggesting the resting brain is anything but. “During this so-called resting-state, the brain is not really resting,” Tavor says. “It does everything, all the time.”