MRI scans are old hat now--people get them for bad backs. But functional MRI is something new: a better way of seeing what the brain is doing.
Medical researchers now have a way of knowing what part of your brain is active, to within a tiny fraction of an inch, when you tap your fingers or listen to someone talk. They have a technique that in principle will allow them to pinpoint where our thoughts and motor impulses originate. Doctors will use the same technique to pinpoint the damage caused by epilepsy or stroke, and perhaps even to improve the treatment of those afflictions. The technique is at once new and not new at all: it is called functional magnetic resonance imaging, and it works much like a standard MRI exam. But instead of revealing the structure of the brain, it reveals the function of the brain’s various parts--parts that can be as small as BBs. There’s never been a way to do this before, says Michael Smith, director of a laboratory at Penn State that has been developing functional MRI, short of opening up people’s heads and implanting electrodes in their brain.
Functional MRI was first tried in England in 1986, but the field has blossomed in the United States over the last few years, in part because MRI scanners are now widely available. In any MRI exam, functional or not, a patient lies motionless inside a narrow tube that is essentially a powerful magnet. The magnetic field forces the magnetic axes of hydrogen nuclei in the patient’s water molecules--70 percent of his body--to line up in parallel. A radio pulse briefly tips the magnetic axes to one side, and when the pulse stops and the axes relax back into line, they emit a weak but detectable radio signal. From the time it takes the nuclei to relax, which depends on how many interacting nuclei there are nearby and thus on the density of the tissue, a computer constructs an image of the brain’s (or the rest of the body’s) internal anatomy. Pathological structures such as tumors show up clearly on such an image.
The goal of a functional MRI exam, however, is not to show structures in the brain but to show the brain doing something. For instance, the researcher might ask the subject to tap a finger. The part of the motor cortex that is responsible for sending the finger-tapping command will do so via a neural impulse. This action requires energy, so more blood flows to that region of the brain. The oxygen in the blood changes the magnetic field in that region in such a way that the radio signal given off by the relaxing hydrogen nuclei becomes more intense--a change the MRI machine detects as it measures the nuclear relaxation time. By subtracting this image from an image of the brain at rest, the researcher acquires a detailed picture of the brain activity in that region. With enough computer power he can even take a rapid series of images, creating a movie of the brain’s activity.
Functional MRI is not the first technique for observing such activity. Positron emission tomography (PET) can also track blood flow to active regions of the brain--indeed, PET scans generate a stronger signal than functional MRI and can therefore detect weaker levels of activity. PET is also better at looking at the entire brain in one fell swoop. But compared with MRI, it has significant drawbacks. For one thing, it requires a radioactive tracer such as an oxygen or fluorine isotope to be injected into the subject’s blood. (The radioactive material emits antimatter particles called positrons, which, on colliding with the brain’s electrons, produce gamma rays that are detected by the scanner.) Thus a person can only be scanned a few times in any one session, and even in any one year, before exposure to radioactivity becomes a health hazard. That makes it hard to do detailed studies of an individual’s brain function. In contrast, there is no known danger associated with repeated MRI scans.
Furthermore, MRI machines cost only about one-fifth as much as PET scanners and are much more widely available. And functional MRI produces sharper pictures: it can resolve regions of the brain as small as a millimeter across, whereas PET’s limit is around two to three millimeters. I see functional MRI as being the dominant brain-mapping technique of the next decade, says neuropsychologist Stephen Rao of the Medical College of Wisconsin. I think it will lead to a major advance in our knowledge of normal human brain functions.
Most of the dozen or so medical centers dabbling in functional MRI are just getting started. But Wisconsin and a few other institutions are engaged in active research. Rao, for instance, is studying how the brain generates motor impulses. He has found that if a person performs a simple repetitive task, such as tapping a finger, only the primary motor cortex is activated. But if the task requires planning--for instance, if Rao assigns numbers to the four fingers and asks the subject to tap out a sequence such as 2-4-3-1--then two secondary motor areas in front of the motor cortex light up as well. And if the person only imagines performing the task, then only the two secondary sites light up--the primary motor cortex remains inactive.
Rao’s ultimate goal is to learn how humans master complex tasks, such as playing the piano. We know that initially learning a motor act involves a lot of conscious effort, he says. But as that same act is repeated over and over again, then it becomes automatic. How do the brain systems change as we learn a new motor skill? The first stages of learning to play the piano are much different than for somebody who is accomplished at playing it, and different brain systems could be involved in that.
Rao’s colleague Jeffrey Binder has used functional MRI to obtain similar results for the auditory cortex. We saw significant differences between activation patterns produced by nonspeech sounds, like tones and white noise, and sounds that were speech, says Binder. The speech sounds are much more complex. They have features that change very rapidly over time, and they seem to activate a much larger area of the auditory cortex than do nonspeech sounds. Although PET studies had also found this distinction between speech and nonspeech, says Binder, functional MRI has shown that the same sound elicits a slightly different pattern of brain activity in every person--an early indication of the new technique’s potential for revealing how individual brains are wired differently.
Functional MRI’s most immediate clinical application will be in the treatment of epilepsy, and several research groups, including Smith’s at Penn State, are working toward that goal. Some epileptic seizures are generalized convulsions of the brain, but in other patients the seizures originate in a distinct region, and in such cases the disease can be treated by surgically removing the focal region. The trick is to locate it.
The current method, called the Wada test, is crude and somewhat risky: the doctor alternately anesthetizes each brain hemisphere while putting the patient through a series of cognitive tests--the idea being to find the damaged region by determining which task the patient can’t do. The Wada test can usually identify the general area that is causing seizures, but it can’t pinpoint the focal region. As a result, the surgeon must remove a lot of perfectly healthy brain tissue, and the patient’s memory is often permanently impaired. Within two or three years, says Smith, functional MRI should allow doctors to do much better: to pinpoint the focal region within a millimeter without subjecting the patient to the risk of anesthesia until it’s time for surgery.
Despite such clinical promise, however, brain-mapping research projects like Rao’s and Binder’s will be the primary role for functional MRI for the next several years. Researchers are still honing and standardizing the technique--at the moment every lab does things differently--and talking about how to interpret its results. But they are downright bubbly about its future. Neuroscientists will have a tremendous tool at their hands that they’ve never had before, and it’s going to be in every medical center within a few years, says Rao. Says Smith: It’s literally taken off, and where it goes, nobody knows.
