The Limits of fMRI

August 19, 2008 12:20 PM

Picturing our thoughts."We're looking for too much in brain scans." [Via]
posted by homunculus (16 comments total)

Yeah I've never understood the fascination with fmri other than 'it looks sciency'. Its pretty clear that the significant events in the brain happen on the order of 50-200ms, outside of the time resolution of most fmri. Furthermore, who really cares where 'processing' takes place, what we want to know is HOW it takes place.
posted by norabarnacl3 at 12:33 PM on August 19, 2008 [1 favorite has favorites]

Very interesting, thanks!
posted by Defenestrator at 12:43 PM on August 19, 2008

AKA "The Technological Hype Cycle".
posted by Bora Horza Gobuchul at 12:48 PM on August 19, 2008

(disclaimer: i'm not really a cognitive neuroscientist, but i work with a few)

Logothetis's article on the limits of fMRI comes at a time when people are really making some big strides, especially regarding the use of pattern classification algorithms for clustering brain states. Norman and Haxby et al. have a really good paper, Beyond Mind Reading (PDF), summarizing some recent developments. Especially interesting is how they honed in on the neural representations of line orientation in the visual cortex. In another study, Identifying natural images from human brain activity, Kay et al. showed it was possible to figure out what pictures of natural scenes subjects were looking at just by examining their brain activity. I don't think we're so far away from being able to reconstruct static visual images right from fMRI data. fMRI is no silver bullet, but man can it do some really neat stuff, and Logothetis might be just a bit pessimistic.
posted by avianism at 12:49 PM on August 19, 2008 [1 favorite has favorites]

Still starin' at there goats... yep. Starin'.

Not much happening with these goats yet. Nope. Not much.

Also, fmri is one of the have-a-hammer-and-everything-looks-like-a-nail kind of tools. There aren't any otherr tools out there so people do fmri studies which are still better than the handwaving that goes on. I suppose it's a little like the divide between theoretical physicists and experimentalists in particle physics. Most of those cloud chamber diagrams don't tell us much but at least they can validate theories once in a while.
posted by GuyZero at 12:51 PM on August 19, 2008 [1 favorite has favorites]

My brain activity spiked when I read the phrase "hide the sausage." Didn't need a fMRI to tell me that.
posted by strangeleftydoublethink at 1:05 PM on August 19, 2008

I've been petitioning my boss for nigh on a year now to allow me to conduct a study to find the fMRI center of the brain. You know, the part of your brain that lights up when you're placed inside an MRI machine.
posted by logicpunk at 1:34 PM on August 19, 2008 [1 favorite has favorites]

More recently, a paper from the lab of Mriganka Sur at MIT reported that the vast majority of the signal detected by fMRI machines was actually a byproduct of astrocytes, which are common cells in the brain that provide neurons with oxygen and energy. When the activity of astrocytes was blocked, but the activity of neurons remained unchanged, 80 to 90 percent of the fMRI signal disappeared. This suggests that fMRI machines are vulnerable to any disorder or drug that leads to changes in astrocyte function, since such changes will dramatically skew the imaging data. "Astrocytes remain so poorly understood," says Sur, "but researchers should keep in mind that astrocytes are largely what fMRI is measuring, and not neural activity directly."

I sense a market for Astrocyte inhibitors to be sold to people who suspect they may be subjected to coercive fMRI scans for law enforcement/intelligence interrogation purposes.
posted by scalefree at 1:43 PM on August 19, 2008

There are a lot of skeptics about the claims made in the name of neuroscience, whether by the fMRI experimenters or people who just get weak in the knees when brain scans are pulled out as evidence for this or that claim, often about gender/sexuality issues. For a change, here's one about Literature and neuroscience.
posted by kozad at 2:15 PM on August 19, 2008

Science article summarizing some of the recent critiques of fMRI studies, including a sidebar where they described how the same "science article" was more widely believed and trusted if it was accompanied by a 3-D brain scan graphic.
posted by bumpkin at 2:46 PM on August 19, 2008

fMRI images = instant grant money, due to science-y appeal!

On a more serious note, I've watched grad students massage their BESA algorithms until a desired result (or any result, really) was produced, so I think a bit of skepticism is healthy.
posted by availablelight at 3:54 PM on August 19, 2008

[...and yes, BESA--and the research I was familiar with--was EEG/MEG, but I'm guessing there is an equal amount of wiggle room and hand-waving in fMRI data processing]
posted by availablelight at 4:05 PM on August 19, 2008

Oh, this is classy:

Last year, the New York Times published an op-ed that used fMRI to investigate the brains of swing voters as they stared at photos and videos of presidential candidates. For instance, the scientists found that pictures of Mitt Romney led to activity in the amygdala, while pictures of Hillary Clinton activated the anterior cingulate. (Interestingly, the only two candidates who inspired "little activity in areas of the brain associated with thought or feeling" were Barack Obama and John McCain.)

For the record, the amygdala processes the memory of emotional reactions, while the anterior cingulate cortex oversees some autonomic functions, like regulating blood pressure and heart rate, as well as reward anticipation, decision-making, empathy and emotion.

Conclusion? Mormons make people remember traumatic events, and Hillary Clinton can have a direct impact on the human circulatory system. Science: 1, Politics: 0.
posted by Marisa Stole the Precious Thing at 4:24 PM on August 19, 2008

fmri is one of the have-a-hammer-and-everything-looks-like-a-nail kind of tool

...but it sure beats trying to push nails into wood with your bare hands. Data about which regions of the brain are receiving the most oxygenated blood is data which we just didn't have before. fMRI won't locate your "adventure center," but used judiciously in conjunction with other evidence it can tell us some things which we wouldn't otherwise know.
posted by justsomebodythatyouusedtoknow at 8:20 PM on August 19, 2008

but it sure beats trying to push nails into wood with your bare hands

Sure, absolutely. I'm not trying to be dismissive towards fmri researchers, but they really do have a very limited set of tools with which they can study the brain directly in real-time. There are a number of avenues of research for which something other than fmri would be better but those techniques just don't exist, so it ends up that the tool drives the research. But it's not like neuroscientists are blind to this, they're just stuck.

(of course, I'm hardly an expert in neuroscience and I'm sure someone is trying some other better research technique)
posted by GuyZero at 9:27 AM on August 20, 2008

Is it possible for BBC Horizon to reach the 40minute mark without showing a meaningless brain scan?
posted by dydecker at 10:27 AM on August 20, 2008