Cover Story

Ever tried to lie down on a 2-foot-wide plank and stay perfectly still for two hours?

It's a fairly impossible task. And as a result, psychologists and researchers have long been frustrated by how a participant's movements can blur the otherwise remarkable work of the scanners used in functional magnetic resonance imaging (fMRI).

"Even the best subjects move during the course of the scan, be it by breathing, heartbeats or head motion," says Jack Gallant, PhD, a psychology professor at the University of California at Berkeley.

As a result, fMRI data can be blurry, even with good motion-correction software. Researchers are developing new brain-scan technology to overcome this and other limitations. They are refining fMRI by increasing scanner sensitivity and employing stronger magnets. In addition, they are turning to new technologies such as Diffusion Tensor Imaging to map connections between different areas of the brain.

Such advances are giving psychologists an unparalleled look into the brain—with DTI, for instance, researchers can see how areas of the brain work together and how those connections form. With stronger fMRI magnets, researchers can better pinpoint blood flow changes in the brain—giving scientists more information about the neurological underpinnings of, for example, schizophrenia. However, such advances have clear limitations, notes Gallant.

"At some point, the magnets and gradients would become so strong that they might have adverse health effects on the subjects," Gallant says. "So eventually—maybe soon—the easy route of merely increasing magnet/gradient strength will be closed."

Bigger, faster, better

Considered by some to be one of the greatest scientific advances of the last quarter century, fMRI allows researchers to study brain activity by tracking blood flow. The magnets pick up increases in oxygenated blood, and computers use that information to identify likely hot spots of neuronal activity.

But although this technology can pinpoint brain activity within the millimeter range, its temporal resolution is limited: Blood flow takes about two seconds to change course, while a thought can happen in milliseconds. In addition, blood vessels that serve an activated brain area draw oxygenated blood from surrounding vessels, which means that fMRI scans often show larger activation areas than perhaps they should, says Gallant.

To make fMRI show brain activation more clearly, researchers are adding channels to fMRI radio-frequency coils. The coils transmit radio waves into whoever is being scanned and receive signals sent back. The more channels a coil can pick up, the more information the scanner gathers and the clearer the fMRI picture, says Gallant.

Multichannel coils can also help researchers who work with children or other participants who can't help but twitch or wiggle, says Gallant. The coils are somewhat redundant with one another, so some missing data can be recovered, he says.

Over the past decade, the number of fMRI scanner channels has increased from four to 32, notes Ray Lee, PhD, a physicist at Princeton University's Neuroscience Institute. But there is a downside: It can cost hundreds of thousands of dollars to retrofit a scanner for a 32-channel system, Lee says.

The number of coils in each scanner has increased as well, though researchers are beginning to reach the limit of this technology, says Lee.

Russell Poldrack, PhD, a psychology professor at the University of California, Los Angeles, says more coils can reduce distortions in the fMRI image near brain regions where air and tissue meet.

"The area that is most problematic is in the sinuses, which causes problems imaging the bottom part of the frontal lobe," he says. "Similar problems occur in the ear canals and cause problems imaging parts of the temporal lobe."

Also, stronger magnets (7 tesla, for instance, instead of the conventional 1.5 or 3 tesla scanners) are more sensitive to changes that happen outside of the blood vessels and nearer to the neurons themselves. However, there are difficult technical challenges in using higher field magnets for human fMRI, including increased distortion, as well as expense.

"Money, availability and technology are all factors here," Gallant notes.

New scans

Another brain-scan technology that is gaining popularity among researchers is Diffusion Tensor Imaging. DTI, which has been around since 1991, only recently hit the research mainstream, says John Gore, PhD, director of the Center for Imaging Sciences at Vanderbilt University Medical Center.

Scientists use DTI to map connections between different areas of the brain—known as white matter—while fMRI scans show gray matter and general patterns of activation. DTI does this by tracing the flow of water in people's brains, as water tends to run alongside brain cells, which are coated with water-repelling myelin.

Cost-wise, DTI is a winner. Researchers conducting DTI studies use the same scanner as fMRI researchers but direct the software to visualize the movement of water instead of just tracking it as a contrast to blood flow. On a practical level, such analysis takes a little more time than fMRI, and participants must spend an extra 10 to 15 minutes in the scanner, says Gore.

Researchers such as Gore are using DTI to study a variety of topics, including language development and disorders, he says. His team, for instance, is determining the average size of connections between the language regions in the back and front of the brain, and they are linking the integrity of those connections to people's ability to process language. Research on abnormalities in these connections may eventually lead to new treatments for dyslexia or other disorders, he says.

Many psychologists using DTI have become "very creative in studying normal brain organization and neuropsychiatric disorders," says Gore. Aging researchers are making particular use of the technology, he notes. As people age, white matter changes or degenerates, especially among those with dementia or vascular disease. Psychologists are now pinpointing abnormal patterns of change in aging people's white matter, he says.

However, researchers don't yet know exactly what specifically determines the movement of water that DTI measures—and whether DTI is sensitive to myelination, fiber orientation or axon size.

One thing that's true for DTI as well as other advances in brain-scan technology, however, is that researchers need to quickly translate their findings to clinical applications, says Gore. Otherwise, their hard-won advances could wither on the vine, he says.

"Knowing someone has a particular deficiency in the brain doesn't help you at all if you've got no treatment," says Gore.


Dana Wilkie is a writer in Washington, D.C.