Tag Archives: neuroprediction

convicted brain

When a press release carrying the title, “Can brain scans predict future criminal behavior?” came across the Mom Psych feed recently, I knew it needed a little extra commentary. If you clicked through to the article just now, you may have had the same reaction I did. Or maybe not. Were you curious? Hopeful? Skeptical? Doubtful? Apprehensive?

Personally, I will admit to a distinct inward cringe, and not just because an image of Tom Cruise sprang unbidden to my couch—er—I mean mind. Naturally I thought of his character in Minority Report, accused of murder long before any such intention had entered his head. But I’m not paranoid enough (yet) to think we’ll ever use fMRI technology to justify “Precrime” units. Still, I couldn’t shake a sinking feeling that establishing a culture of rehabilitation in the criminal justice system may be an impossible mission.

This is not a criticism of the research. We do learn quite a bit about the human brain and behavior from fMRI studies and these investigators are very clearly thinking in terms of therapeutic interventions for likely re-offenders. But of course, that doesn’t mean politicians, attorneys, judges or juries will apply the technology in the intended spirit. As we know, history is replete with examples of science gone wrong: discoveries that initially seem to portend good for mankind all too frequently end up being used to inflict murder and mayhem, and all sorts of other injustices along the way.

My point (in case it needs to be stated outright) is that a reasonable degree of caution is warranted when applying studies like this one to real life. Not only because they may lead down a perilous road from a policy perspective, but also because “neuroprediction” in itself can be tricky business, as Russell Poldrack points out. Director of the Imaging Research Center at the University of Texas at Austin, Poldrack calls this particular study an “impressive” one, although he notes that the predictive accuracy wasn’t validated on out-of-sample data. After performing the missing cross-validation on his own, however, Poldrack found a “slight benefit” to out-of-sample prediction of future rearrest, and concluded:

The take-away message from this analysis is that fMRI can indeed provide information relevant to whether an individual will be rearrested for a crime.  However, this added predictability is exceedingly small, and we don’t know whether there are other (unmeasured) demographic or behavioral measures that might provide similar predictive power.  In addition, these analyses highlight the importance of using out-of-sample prediction analyses whenever one makes a claim about the predictive ability of neuroimaging data for any outcome.  We are currently preparing a manuscript that will address the issue of “neuroprediction” in greater detail.

Well, that’s one very real problem. But statistical analysis aside, there’s another important question we need to ask ourselves: Why do we want fMRI scans to predict recidivism? If we’re profiling so that we can better target therapies for rehabilitation—okay, fine. On the other hand, if we hope to use fMRI scans to make decisions about sentencing, probation, parole, involuntary commitment or juvenile detention, we would be overstepping bounds on many fronts. Even if we could be sure there was nothing misleading about a defendant’s individual scan results seen apart from the average of many test subjects (a big “if”), we would essentially be dismissing the brain’s ability to change. But the brain is malleable—researchers have shown there are ways to help people alter activity levels throughout the brain, including in the anterior cingulate cortex (ACC), which happens to be the area of focus in this study.

Make no mistake, there is still much to learn about the neural mechanisms underlying human behavior, despite the fact that fMRI studies add to the store of knowledge on a daily basis. But it is clear that neuroscience has moved into the legal system to stay, spawning a relatively new area of study christened neurolaw. “Neurolaw studies not only the descriptive and predictive issues of how neuroscience is and will be used in the legal system but also the normative issues of how neuroscience should and should not be used within the legal system,” explain researchers Annabelle (Mimi) Belcher and Walter Sinnott-Armstrong. A neuroscientist and neuroethicist respectively, Belcher and Sinnott-Armstrong’s 2010 paper on the topic points out some of the prevailing thinking about how neuroscience may build on psychology and social science research in the courtroom.

“We are not, of course, endorsing any of the many uses of neuroscience mentioned here,” they caution. “The point is only that neuroscience can, in principle, be used in a great many ways within the legal system; some of these uses of neuroscience are coming quickly, and we had better get ready. We need to get ready both to prevent misuses and accompanying mistakes and also to encourage the best uses of any legitimate applications of neuroscience in law.”

. . . . .

CBC RADIO Interviews Annabelle (Mimi) Belcher (Along with Michio Kaku and Steve Laken): The End of Privacy

%d bloggers like this: