My Brain Is a Walnut

Inside an fMRI machine.

The brain: a tough nut to crack

“If we find any gross abnormalities in your brain, would you like a radiologist to tell you about it?” Tobias Egner asks me. He is about to wheel me into the dark gullet of an fMRI machine at the Functional MRI Research Center at Columbia University, a leading neuroscience lab where he is a research fellow. I say yes to his question and ask if anyone ever says no. “If you answer no, we cannot do the test,” Egner says. He speaks with a soft certainty and a German accent; if this were a movie, he’d be played by Willem Dafoe, a la The Life Aquatic With Steve Zissou. “Ready to roll?”

In the last decade, fMRI (or functional magnetic resonance imaging) has become a premier—and scrutinized—tool of neuroscience. I wanted to see for myself how the technology works. On a visit to Columbia’s lab to view an experiment, I mentioned an interest in participating, and two researchers within earshot accepted my offer. I signed on with Egner to become one of about 20 subjects in a study of how the brain manages conflicting information. To me, this means choosing every night at 7 between the Seinfeld rerun on Fox and the one on TBS. To Egner, it means studying how the amygdala, the brain’s emotional hub, resolves the millions of emotional conflicts people experience every day.

To watch my mind in action, Egner will use the fMRI machine’s magnetic coil to scan my brain every two seconds for blood rushes, which show an increase in neural activity. The more common MRI uses magnetic resonance imaging to scan the body for abnormal tissue, in order to diagnose a tumor or an injury. The added “f” for functional means the machine will look at my brain while I perform mental tasks, so it can see which regions are most active. From a control room facing me, Egner will capture the results on a computer.

Columbia’s fMRI room lies behind several hazard signs and a heavy wooden door. Through the door’s glass window I can see the machine and hear its robotic pulse: zeeew-kip, zeeew-kip, zeeew-kip. The window is lined with a sheet of mesh called a Faraday screen, which keeps radio frequencies from disrupting the machine’s magnetic signal. Before entering, I have to “de-metal” as if I was going through airport security.

Inside, the room has the clean, sharp smell of dry ice. It’s filled by the fMRI machine, a giant white cylinder lying on its side, so large that Columbia had to take out the ceiling to lower it in. A sky-blue rim borders a hole in the center of the machine wide enough for a human body. A padded platform waits to feed me headfirst into the maw.

The test will be loud, Egner has warned, so I stuff two spongy yellow plugs into my ears. Egner wedges my head face-up between two cushions. In a small mirror suspended above my eyes I can see a projector screen beyond my feet, and Egner hands me a device like the control for a video game that I’ll use to answer his questions while I’m in the machine. Egner stretches a piece of tape a quarter-inch above my eyes to remind me not to move my head—if I hit it, I’ve moved too far. “What if I sneeze?” I ask. “I imagine that would cause a bit of movement,” Egner says. The room goes dark.

Egner goes outside and peers back in through another window with a Faraday screen. In the mirror above my eyes I see the projector screen and watch a cursor rifle through files on a computer desktop. As Egner punches buttons from behind the glass, a series of pictures of real human faces appear. They are either scared or smiling, with the word “FEAR” or “HAPPY” written in red across the eyes. I am supposed to press one button on the controller if the face is fearful, another if it’s happy. Sometimes the face and the word show the same emotion. Other times, Egner tries to overload my amygdala with what he calls incongruent tasks: a happy face with the word “FEAR” across its forehead, or a scared face with “HAPPY” across its forehead. Incongruent tasks force the amygdala to ask another area of the brain for help in deciding which emotion the face on the screen is displaying. If Egner sees my neural activity rising in another region while I’m looking at a happy face labeled “FEAR,” his team will have a better idea what parts of the brain aid the amygdala when it’s overloaded with emotions.

A few years ago, researchers at the University of Pennsylvania used fMRI to show that the human brain is more active when lying than when telling the truth, sparking conjecture that the fMRI could replace the polygraph. In a 2001 study, behavioral researchers at Princeton found that fMRI subjects showed greater emotional activity when deciding if they would kill a person with their own hands than when deciding if they would kill indirectly (like by flipping a switch or voting to pass a law). This suggested that emotions play a greater role in moral decisions than previously believed. And in 2005, a team of neuroscientists and economists used fMRI to show how people build trust among potential competitors. These studies are hailed for advancing age-old psychological and neuroscientific arguments. But some scientists worry that fMRI data are by nature imprecise, a limitation that’s often ignored when the press hypes test results. Functional images accurately present brain activity in tiny three-dimensional squares called voxels, but each voxel represents thousands of neurons, so the imaging can’t pinpoint exactly where brain activity takes place. And while functional data provide a good idea of neural activity when averaged among all subjects in an experiment, conclusions can’t be drawn based on individual tests.

That’s why researchers like Egner need to feed groups of subjects like me into their uncomfortable machines. On the inside, the blare is ceaseless; it sounds like a modem connecting and a telephone ringing at once. At one point I hear a staccato burst, as though a woodpecker is pecking its way through the magnet and into my brain.

Black-and-white pictures of real people—some smiling, others scared—continue to flash on the projection screen. After a while I begin to recognize the same characters—the man with crooked lips, the used car salesman, the over-actor. The images repeat themselves though the words change. The only blond girl is always smiling.

The tasks aren’t overly difficult, but the faces grow tiresome to watch. After 300 of them in 40 minutes, my eyes feel heavy. Egner’s voice bursts through the intercom. “Are we falling asleep?” he asks. I wonder if he knows this from watching my brain or from experience with other research subjects. The lights go on and Egner helps me out of the machine. The noise settles back into the steady and much softer zeeew-kip.

When I leave the room, I’m unsettled to find four researchers behind the control-room glass. I’d thought I was alone, but people I don’t know have been watching my mind at work, perhaps remarking on its effectiveness. A few minutes later, Egner hands me a flimsy sheet, like an X-ray, with a birds-eye view of 20 horizontal layers of my brain. Functional images don’t come the way they’re shown in magazines and journals, with reds and yellows splattered across the brain to show where the most neural activity occurred. The colors are added later by the researchers. In this starker rendering, my brain looks like a walnut in its shell. I ask Egner how I did. He explains to me that individual results can’t be interpreted. “As for your function,” he said, “there’s not much to look at.” If the choice is between that and gross abnormality, I guess I’ll take it.