ACLU Files for Info on New Brain-Scan Tech

An anonymous reader writes "According to their website, the ACLU has filed a FOIA request seeking information on the new Functional Magnetic Resonance Imaging service being made available to the government for use on suspected terrorists which can produce 'live, real-time images of people's brains as they answer questions, view images, listen to sounds, and respond to other stimuli. [...] These brain-scanning technologies are far from ready for forensic uses and if deployed will inevitably be misused and misunderstood," said Barry Steinhardt, Director of the ACLU's Technology and Liberty Project. "This technology must not be deployed until it is proven effective -- and we are a long way away from that point, according to scientists in the field,"'"

Unless we magically isolated the "lying" part

[Too many errors, bai]

of the brain, using fMRI to detect lies is a load of dung. Way slower to react than lie detectors, and a horrible image resolution. I'm not saying it's entirely impossible, I just severely doubt the possibility of determining guilt by brain lobe activation levels.

Comments

[venicebeach]

First, a correction. The article says:

The most likely technology to be used for anti-terrorism purposes is Functional Magnetic Resonance Imaging (fMRI), which can produce live, real-time images of people's brains as they answer questions, view images, listen to sounds, and respond to other stimuli.

fMRI does not produce live, real-time images of brain activity. At best, this is misleading. First, temporal resoultion of fMRI is very poor when compared with the speed of firing of neurons. A typical fMRI experiment takes a picture of the whole brain every 1.5-4 seconds. Furthermore, the blood oxygenation changes measured with fMRI are slow and cause an effective temporal blurring of the data (blood peaks about 6 seconds after brain activity). To determine which changes relate to changes in psychological function, much offline processing is necessary. Yes, it is possible and has been acheived in some cases to have semi-real-time online analysis, but this is certainly not the norm. What you typically end up with at the end of an fMRI experiment is a static map showing the extent to which signal at each voxel correlates with your task of interest.

Now as for the issue at hand, it is certainly premature to use fMRI as a reliable lie detector or something like it. However, the article does not really specify how it is being used. If data is being collected to advance the reliabilty of this tool as a lie detector then it could be effective sooner rather than later.

Re:Faulty systems can still work some of the time.

[CosmeticLobotamy]

if someone is questioned about a large number of things, and he gets nervous when answering certain questions, that might be a good place to start investigating.

Or they know that that question is the one you think they did. I had to be polygraphed for a job ("Of course it's voluntary. We're just not hiring you because we liked the other guy's hair better."). In the pre-interview, they ask if you've ever been questioned by police, so I said yes. Which is true. When I was a kid, I was in the wrong place at the wrong time. Didn't do a damn thing, and the cops knew it, but this guy didn't ask them. He asked me about it 45 times in the machine, and obviously that question was important, and it made me nervous every time.

They don't actually tell you the results of those things, but for some reason, I went from being a lock with, "It's just a formality. Call when they're done, we'll get you set up," to not answering my calls for a week until they called to tell me they offered the job to someone else.

Obviously I can't be sure that's why. Maybe my fly was open. But the polygraph's the only reason I can think of.

What I particularly loved was at the end, the guy looks upset and says, "Were you controlling your breathing?" Yes! You strapped a frigging cable around my torso and told me to keep still! Stupid frigging *grumble* *grumble*...

Why is this a "Civil Liberties" issue?

[mi]

These brain-scanning technologies are far from ready for forensic uses and if deployed will inevitably be misused and misunderstood.

The results, if any, will be presented in courts, with experts from defense and prosecution debating their merits in front of juries. This happens to fingerprints, DNA, speed radars, and all other technologies used in crime-fighting.

In short, I feel, my ACLU donation is being misused...

Re:This is a joke right?

[venicebeach]

Well, you don't necessarily need to know much except how the overall pattern differs between lying and telling the truth. There have been several studies to do it recently that have had some success. For example:

Kozel FA, Johnson KA, Mu Q, Grenesko EL, Laken SJ, George MS. (2005) Biol Psychiatry. Oct 15;58(8):605-13.
Detecting deception using functional magnetic resonance imaging.

BACKGROUND: The ability to accurately detect deception is presently very limited. Detecting deception might be more accurately achieved by measuring the brain correlates of lying in an individual. In addition, a method to investigate the neurocircuitry of deception might provide a unique opportunity to test the neurocircuitry of persons in whom deception is a prominent component (i.e., conduct disorder, antisocial personality disorder, etc.). METHODS: In this study, we used functional magnetic resonance imaging (fMRI) to show that specific regions were reproducibly activated when subjects deceived. Subjects participated in a mock crime stealing either a ring or a watch. While undergoing an fMRI, the subjects denied taking either object, thus telling the truth with some responses, and lying with others. A Model-Building Group (MBG, n = 30) was used to develop the analysis methods, and the methods were subsequently applied to an independent Model-Testing Group (MTG, n = 31). RESULTS: We were able to correctly differentiate truthful from deceptive responses, correctly identifying the object stolen, for 93% of the subjects in the MBG and 90% of the subjects in the MTG. CONCLUSIONS: This is the first study to use fMRI to detect deception at the individual level. Further work is required to determine how well this technology will work in different settings and populations.

Re:The ACLU - some people's rights but not others

[Ungrounded Lightning]

There are plenty of other organizations willing to defend your 2nd amendment rights. The ACLU is a private association, it can defend rights however it sees fit.

And they're even welcome to print their poster on the bill of rights that leaves off the second (and a couple of others) entirely. (Even if it is as revolting as flag burning, it IS free speech.)

But IMHO they crossed a line when they provided a lawyer for the shooting victim of a crook to sue for damages the person from whose locked safe the gun had been stolen.

Re:Why is this a "Civil Liberties" issue?

[TFoo]

It is important to note that the ACLU has only file a FOIA request at this point: they haven't filed expensive lawsuits or spent a ton of money yet -- so don't jump to complain just yet.

The government will prevail, unfortunately

[pongo000]

These brain-scanning technologies are far from ready for forensic uses

So are polygraph tests, yet these are routinely used in a "forensic" capacity.

Since when has the unsuitability of polygraphs for forensic use [psychologymatters.org] ever stopped the government from using such technology to their own purposes?

Bravo to the ACLU for taking this on. Unfortunately, their actions will be minimalized over the government's assertion that this technology will catch more terrorists. And before you know it, you'll be submitting to brain scans during your next employment interview, or police interrogation.

Re:The ACLU - some people's rights but not others

[Saanvik]

You know, I've seen this claim a few times (ACLU lawsuit relating to a stolen gun), but I've never seen any sources to back it up.

Do you have any?

Re:First post(?)

[Dun Malg]

A better idea is if the Alphabet Agencies (CIA/DoD/NSA/DoJ/etc) uses FMRI's for security screenings, in the same way that polygraph's are used. That way science can build up a body of knowledge at the Federal Gov'ts expense and the results can be backed up with polygraphs.

Polygraphs can't back up shit. They're a pile of crap. There are no physiological reactions that can be specifically atributed to deception. That's why they're not permitted as evidence in any court. Why do you think it is that the two possible results of a polygraph are "shows signs of deception" or "inconclusive"? Polygraph results are highly subjective interpretations of ill-defined measurements. Baseline questions are asked that supposedly set the thresholds for "truth" and "deception", but the machines largely rely on the subject's subconscious fear that the machine is catching them in the lie. There isn't a red light or buzzer on the machine that goes off every time the subject lies. What you have is just one man's opinion of what a lot of jumpy marks on graph paper mean in relation to your guilt or innocence-- influenced, of course, by his guess, based upon what he has heard about you, and deductions he draws from how you appear and act.

Re:First post(?)

[BWJones]

Furthermore, if you know what you are doing, you can influence the polygraph any way you want it (trust me, I teach neurophysiology to medical students). There are other methods of lie detection that are harder to spoof such as the P300 method (cortical evoked potential at 300ms delay in normal persons signifying recognition) being investigated, but even this method has it's problems in that you cannot discriminate why someone may elicit a P300. I would also suspect that interpretation of fMRIs can also be confused by someone who "knows" how to lie. The trick is to avoid delivering "tells" that are physiologic manifestations of deception and build yourself a reality behind the lie. I've said it before, but the truth is that there is no foundation in physiology that mandates that one has to reveal anything when stating something that is not in fact, the truth. A good liar will be able to deceive the device and more importantly, the interpreter of the device because they are able to LIVE the lie.