Cover Story

"Wendy," 15, lay awake under the bright lights of the operating room. She couldn't see it, but part of her skull was missing, and Minnesota Epilepsy Group physicians had implanted a grid of electrodes on the surface of her brain. Wendy counted to 10 and named pictures on flashcards, allowing the surgeons to map where she processed language—and avoid harming important brain areas while cutting out those responsible for her unremitting epilepsy.

Again and again, the doctors failed to capture a clear picture of Wendy's brain, says Wenbo Zhang, MD, PhD, the team's technical director.

"They couldn't map her language function," he recalls.

Luckily, the surgeons had a fallback plan. Earlier that week, Wendy had spent an hour with psychologist Robert Doss, PsyD, while in the bright white cavity of a magnetic source imaging (MSI) machine. A new technique for measuring brain activity, MSI records the magnetic fields created by electrical activity in the brain.

Unlike most brains, Wendy's brain buried language function deep in its folds instead of planting it on the surface. As a result, the surgeon's electrodes weren't able to pick it up. Luckily, the MSI scan peers deeper into the brain and gave the surgeon a clear picture. So Wendy emerged from the operating room with a little less gray matter, but no fewer words.

"She has been seizure-free and never complained about any cognitive deficits or language deficits," says Zhang.

The Minnesota Epilepsy Group is one of just a handful of medical centers nationwide that are using MSI, fMRI and other brain imaging techniques with actual patients, says Doss. However, surgeons are increasingly asking for functional brain maps of their brain tumor and epilepsy patients—driven, in part, by a new rule that allows insurance companies to reimburse that service.

Pre-surgical mapping, however, is just the beginning, says Doss. In coming years, functional brain imaging may revolutionize neurologic and psychiatric health care, taking some of the guesswork out of diagnosis and helping doctors tailor treatment programs to particular patients. But first, psychologists need to perfect ways to map the brains of particular people, says Doss.

"Neuropsychologists are integral to the successful implementation of these programs," he says.

Better surgeries

People who are undergoing brain surgery for epilepsy have to endure many indignities. In addition to mid-surgery brain mapping, these patients often undergo a procedure known as a WADA test, which identifies which brain hemisphere houses a person's language function. A physician injects an anesthetic into the patient's carotid artery to knock out an entire hemisphere of brain function and then sees if he or she can still talk. Hopefully, it will soon be supplanted, says Peter Bandettini, PhD, chief of National Institute of Mental Health's Section on Functional Imaging Methods

"People are trying to replace these invasive tests with more sophisticated kinds of brain imaging," he says.

One of those people is Doss, who would like to use MSI brain mapping instead.

"The WADA test is a very invasive, uncomfortable procedure, and there's some risk of morbidity," he notes.

A 2006 study by Doss and his colleagues, published in Epilepsia (Vol. 47, No. S4), found that MSI and WADA tests returned identical results 87 percent of the time.

Thanks to Doss's research, many patients who visit the Minnesota Epilepsy Group are now being spared the WADA test.

In addition to making life more pleasant for patients, brain imaging can assist physicians with pre-surgical planning—telling them, for example, how close a tumor lies to a person's language centers, says Keith Thulborn, PhD.

"Sometimes the decision is not to operate if the risk of complications is too great," says Thulborn, a radiology professor at the University of Illinois at Chicago. For eight years, he has worked to harness fMRI's ability to map patients' brain function. If the physicians and patient decide to operate, the brain images are uploaded to a computer in the operating room and help surgeons navigate through patients' brains. A glance at the screen then tells a surgeon how close he or she is to harming a patient's motor or language skills.

Most hospitals already have fMRI scanners in place, says Thulborn, and two years ago, the Centers for Medicare and Medicaid Services created Current Procedural Terminology codes that allow physicians and psychologists to bill insurance companies for pre-surgical brain mapping. However, Thulborn says, most brain-surgery patients aren't yet benefiting from the advances.

"Now that the codes are in place, the rest of the world is playing catch-up," he says.

Faster diagnosis

Pre-surgical brain mapping is the only clinical application for brain-scan technology in mainstream use, says Vince Calhoun, PhD, an electrical engineering professor at the University of New Mexico. Calhoun and others are working to change that by using brain scans to diagnose mental illness.

If scientists can develop a way to fingerprint such disorders as schizophrenia or Alzheimer's disease with brain scans, it would take some of the guesswork out of assigning people to therapy or prescribing medication, says NIMH's Bandettini.

"If you have some sort of brain activation map to complement your behavioral tests, it could make a more accurate diagnosis or help you monitor their improvement better than behavioral tests alone," he says.

Currently, mental health professionals use symptom inventories to determine whether, for instance, a person has schizophrenia or bipolar disorder. While the two disorders have very different courses, they can sometimes look similar—especially when a person's mania symptoms include hearing voices or other psychotic features. Consequently, many people who are bipolar end up starting on anti-psychotic medication when they actually need mood stabilizers, Calhoun says.

A new study by Calhoun, published in Human Brain Mapping (Vol. 29, No. 11), finds that fMRI maps can correctly categorize people who have schizophrenia versus bipolar disorder 94 percent of the time. He achieved this by using data that many researchers previously overlooked: people's brain functions when they are not doing anything in particular, says Bandettini. Rather than asking people to perform tasks in the scanner, Calhoun just recorded their regular brain activity. He then used a new computer algorithm to make sense of the data—one that takes into account correlations in activity across different areas of each person's brain. That process, called functional connectivity mapping, allows the computer to highlight subtle differences in the patients' brains and correctly diagnose them.

"By looking at the brain as a network, we can get a complete picture about how these brain regions are talking to one another," Calhoun says.

Researchers at Stanford have had similar success diagnosing people with Alzheimer's disease. Using a process called independent component analysis—which allows them to visualize the brain's pattern of co-activation—they are developing programs that can distinguish between people who have Alzheimer's disease and healthy controls. This line of research may eventually provide a way to detect and—in conjunction with novel therapies—even prevent Alzheimer's disease before it takes hold in the brain, says Michael Greicius, MD, lead researcher on the project.

"An ideal imaging biomarker would allow us to start treatments early in the disease course before significant cell death has occurred," he says.

What's the holdup?

Most hospitals already have the tools in place to do functional brain imaging with patients, but physicians generally use the machines to take static snapshots of patients' brains. What's keeping functional brain imaging out of the clinical mainstream, says Bandettini, is a shortfall of sophisticated data-processing techniques—particularly ones that can make sense of individual patients' brain activity. Most researchers average brain activity across many different people to get rid of the meaningless data, or "noise." That's a good technique for answering broad research questions, but it's not sensitive enough to diagnose mental illness or locate where in a particular person's brain language function lies.

"One big problem with fMRI is that so much of the research is committed to group studies," Bandettini says. "That kind of research is virtually useless in a clinical setting, where making an individual diagnosis is the key."

Scientists such as Calhoun, Greicius and Doss have found success by gleaning information from data that most researchers would average out, Bandettini notes.

Further advances in interpreting brain imaging and linking them to people's cognitive abilities and inner lives will take close collaboration among psychologists, neurologists and radiologists, adds Doss. The work has the potential to make brain surgery less risky and scary, make Wendy's story commonplace, and perhaps even take the mystery out of mental illness.

"The technology's there, the magnets are big enough, the computers are fast enough, now all we need is more cross-disciplinary collaboration," Bandettini says.


APA offers fMRI primers

In March 2007, APA published two booklets that describe how fMRI research is contributing to psychologists' understanding of the inner workings of the human mind. Since then, APA's Science Directorate has distributed more than 3,000 copies of each to psychologists, high school psychology teachers, graduate students, clinic directors and members of the general public, says Virginia Holt, assistant executive director of APA's Science Directorate. Interest in the booklets is still strong, she notes, as fMRI research continues to make headlines in the areas of addiction, racism and dyslexia.

"We wanted to inform a wide audience about brain imaging advances, so we created two booklets—one for teenagers and one for adults,"she says.

To request hard copies of the booklets, e-mail Science or download the APA Science Directorate pamphlets on fMRI and Psychological Science.