The visual images of brain activity captured by functional MRI machines are undeniably fascinating. But two new studies point out troubling aspects of this technology.

Functional magnetic resonance imaging has, in a very real sense, opened up a window to the brain. By measuring blood oxygen levels, scientists can now determine which parts of the brain are active or inactive as subjects participate in various activities. This has provided a much clearer picture of how we process information and make decisions.

In recent years, researchers working with fMRI machines have examined everything from how the brain responds to violent movie images to the way jazz musicians shut off their self-critical censors during improvisation.

But now, there is talk of using these machines as part of interrogation of terrorism suspects, according to Jonathan Marks, head of the Bioetchics and Medical Humanities Program at Penn State. Marks reports that some in the intelligence community believe fMRIs could serve as high-tech lie detectors -- a notion he considers extremely problematic.

Theoretically, an fMRI reading could tell intelligence experts that a suspect is responding to a particular image or word -- say, the photo of a known terrorist. But, Marks noted, the machine can’t explain the reason why this brain stimulation is occurring. “I spent years living in London, listening to reports of IRA bombings,” he said. “My brain would light up if you mentioned the word semtex (a plastic explosive).”

“MRI machines are very useful diagnostic tools, but using them to claim that certain things are going on inside people’s minds is a major jump,” Marks added. “One of the real concerns I have is that you can see people begin to say, ‘The fMRI picked him out as a terrorist, so let us give him a going over in the interrogation room.’ ” 

That gives new meaning to the phrase “tortured logic.”

Meanwhile, psychologists David McCabe of Colorado State University and Alan Castel of UCLA raise a less dramatic but nevertheless troubling question: Do those wow-inducing pictures showing various areas of the brain lit up give studies an underserved amount of extra validity?

In an experiment, they had students read fictional articles summarizing cognitive neuroscience research. Some articles were accompanied by a brain image, others by a bar graph, while still others had no imagery at all. Participants were “asked to rate the soundness of the scientific reasoning in the article,” and they gave top marks to the studies that were illustrated with the brain images.

The researchers conclude that the brain images are persuasive because “they provide a physical basis for abstract cognitive processes.” While they find this somewhat disturbing, they also see an up side to the fascination with this imagery: It is making cognitive neuroscience research more accessible and giving it greater credibility with the general public.