If a picture is worth a thousand words, a brain scan might well be worth far more, in terms of publicity, professional renown and even product sales. In particular, functional magnetic resonance imaging, or fMRI, whose output is often digitized into splashy, colorful brain-scan pictures, has captured the hearts of both researchers and the public, both sometimes seemingly without careful consideration. News stories (and the occasional research paper) now routinely announce the discovery of the brain region for X, where X can be pretty much anything from morality to lust to lapses in judgment.
No doubt, fMRI is a revolutionary technique. Though so far it has not been of much use in the clinic, it is one of few tools that allow real-time observation of how the brain changes in healthy people, making possible many studies previously relegated to wishful thinking.
But just because something lets us study the imagination doesn’t make it immune to our own flights of imagination. FMRI is useful but imperfect, its current results more suggestive than explanatory. Cataloging and detailing its flaws are vital, because fMRI is increasingly being recruited, sometimes with great recklessness, as a technological clairvoyant. At least two companies have formed to conduct fMRI-based lie detection; marketing companies use it to attempt to perceive consumer behavior—and some people are even considering a role for it in matchmaking and marriage counseling.
But these images are not real, they are representation; MRI machines spit out numbers, and the subsequent pictures are graphical estimations created after those data have been parsed, analyzed and cleaned up (the process has been the topic of a major recent controversy in neuroscience). The numbers themselves are also a proxy, since they measure blood flow as a sign of brain activity instead of measuring the activity itself. The blood response takes up to 15 seconds, far slower than many of the thought processes being studied. And for a research participant, being in an MRI machine is unpleasant; it’s small enough that it can induce claustrophobia, and moving your head even a bit can throw off the results. How genuine results from such an alien environment can be is debatable.
As Russell A. Poldrack points out in a new Cerebrum article, there are other limitations, too. These types of imaging studies usually use small numbers of test subjects, and results often have not been verified by other research groups. They measure group averages, leaving their usefulness for individuals in doubt.
Saying there is a brain region for something is inevitably an oversimplification, since much of the brain is always active and every brain process involves disparate regions. Some brain regions activate to shut down activity in other regions, making interpretation of the data trickier.
Yes, fMRI scans have potential, he says, especially for lending some much-needed quantitative rigor to our understanding and diagnosis of psychiatric disorders. But we’re not there yet, he adds, and understanding precisely why is vital to keeping expectations reasonable and ensuring that we are savvy enough to assess the research and marketing claims we encounter.
Comments