A little over a decade ago, neuroscientists began using a new technique to inspect what was going on in the brains of their subjects. Rather than giving their subjects a task to complete and watching their brains to see which parts lit up, they’d tell them to lie back, let their minds wander, and try not to fall asleep for about six minutes.
That technique is called resting state functional magnetic resonance imaging, and it shares a problem with other types of fMRI: It only tracks changes in the blood in the brain, not the neurons sending the signals in the first place. Researchers have recently called fMRI into question for its reliance on possibly-faulty statistics. And things get even less certain when the brain isn’t engaged in any particular task. “These signals are, by definition, random,” says Elizabeth Hillman, a biomedical engineer at Columbia’s Zuckerman Institute. “And when you’re trying to measure something that’s random amidst a whole bunch of noise, it becomes very hard to tell what’s actually random and what isn’t.”
Six years ago, Hillman, along with many others in the field, was deeply skeptical of resting state fMRI’s ability to measure what it promised to. But this week, in a paper in Proceedings of the National Academy of Sciences, she presents compelling evidence to the contrary: a comprehensive visualization of neural activity throughout the entire brain at rest, and evidence that the blood rushing around in your brain is actually a good indicator of what your neurons are doing.
Ever since 1992, when researcher Bharat Biswal first started scanning people who were just sitting around, resting state fMRI has become increasingly popular. Partly, that’s because it’s just way simpler than regular, task-based fMRI. With task fMRI, “if you wanted to assess a schizophrenic,” Hillman says, “you’d have to get them to go into the magnet and do some sort of task. And what task do you design to test if someone’s schizophrenic?” Resting state fMRI is also great for figuring out which parts of the brain are functionally connected—researchers just need to find areas that light up together when the brain is idly noodling around.
Most resting state studies focus on small sections of the brain, primarily because of camera limitations. So no one’s really investigated the way the brain is organized more globally. “It’s kind of like looking at traffic patterns in a city,” says Shella Keilholz, a neuroscientist at Georgia Tech. On the freeway in a car, sitting bumper to bumper, things can seem arbitrary. But take a bird’s-eye view, and the movements of cars make a lot more sense.
A Window to the Brain
It took Hillman’s lab years to devise its methods. They wrestled with only being able to scan anesthetized mice—they needed them to keep still in the scanner, but the researchers would see different things when they applied different anesthesia, which distorted the signals they were trying to study. So they took mice and surgically thinned their skulls so they could see the surface of the brain, then used superglue (“We’re big users of superglue,” Hillman says) to affix a small headplate on the mice—a set of tiny, laser-cut plastic antlers.
After the mice recovered with morphine and treats, the scientists clipped the headplates to a holster to keep the mice in one place, complete with a camera and a couple LEDs. Then they tracked two things: where blood was flowing in the mice’s brains, using the oxygen in the hemoglobin; and which neurons were firing when, thanks to a calcium-sensitive protein that lit up green. When they analyzed the data, they found that the two components matched up really well—turns out, blood flow is a good proxy for what the nerves are doing, and that neural activity is likely driving patterns in the blood flow.
To do that, Hillman and her team pulled out the slow changes underlying the flickering raw data of the neurons, yielding a detailed view of the brain’s activity at rest. The neurons flash rhythmically and symmetrically, like something out of a trippy music video. What’s actually going on with the flashing is an open question, though. “My money is on this being an essential piece of how the brain checks itself,” she says. “But its mechanisms? That’s uncharted territory.”
Though the paper provides one of the first clear views of the brain’s resting state, some parts of the blood flow Hillman’s lab tracked didn’t match up with the neural patterns they found, and the researchers still aren’t sure why. Maybe it was noisy data, or the way they set up their experiment, or other unrelated blood patterns. But more careful, large-scale studies like this one open the way for other scientists to tease those patterns out.