A paper from Jack Gallant's group at Berkeley posted on-line before publication in Nature is getting lots of attention, due in no small measure to the eye-catching headline of the accompanying news piece: "Mind-reading with a brain scan." The paper itself is somewhat more modestly titled, "Identifying natural images from human brain activity", and this represents the data quite a bit better than the slightly sensational news item. But if ever there was a paper that was newsworthy to neuroethics, this is it.
The authors showed 1,750 images to two subjects (who happen to be the first and second authors on the paper) while imaging blood flow in their brains using fMRI. The resultant responses were catalogued and then the subjects were shown 120 novel images; using the information gleaned from the previous data run, the investigators were able to predict the blood flow response as measured by fMRI in response to these images with reasonable accuracy (72% for one subject and a whopping 92% accuracy for the other).
So is this mind-reading? The answer is decidedly no, and as Gallant points out,
The next step is to interpret what a person is seeing without having to select from a set of known images. “That is in principle a much harder problem,” says Gallant. You’d need a very good model of the brain, a better measure of brain activity than fMRI, and a better understanding of how the brain processes things like shapes and colours seen in complex everyday images, he says. “And we don’t really have any of those three things at this time.”
In this sense, the study is more properly characterized as an attempt to unravel the neural code, albeit using the indirect measure of the BOLD signal detected by the fMRI machine. Nonetheless, the findings are sure to raise all kinds of alarms about bona fide mind-reading. If this study spurs serious discussion of what we will do when we really develop mind-reading technology, it will have made important contributions both to deciphering the neural code and triggering neuroethical debate. Not bad for a day's work.
Comments