Reimagined and Revamped. Fighting the spread of nonsense often feels like a Sisyphean task. However, the joy is in making the information available, not the hope of conversion.

fMRI doesn't do what you think it does.

I'm writing this because I keep reading, quite often now, that something is true since an fMRI study shows neural activity during the application of <insert pseudoscience here>. But the interpretations of these studies, regardless of the conclusions in the actual study itself, are often wrong. So I thought I'd do a little overview of fMRI and why some of these conclusions can not be made based on the results of an fMRI scan.

First, lets start with the basics:


Your red blood cells have hemoglobin molecules (hundreds of them) which have a cool feature. They can hold on to oxygen molecules when they are available, and release them when needed. Your red blood cells can take on oxygen in an oxygen rich area, you know, like on the surface of the inside of your lungs, and get pumped around and when they get into the capilary system, the oxygen leaves when the red blood cell gets near tissue that needs it.

Oxyhemoglobin is the hemoglobin when it is rich in oxygen.
Deoxyhemoglobin is the hemoglobin when it oxygen free.

It turns out the oxyhemoglobin and deoxyhemoglobin have different magnetic properties. Oxyhemoglobin is diamagnetic, meaning when a magnetic field is applied to it, it will generate an opposite magnetic field. Bismuth shares this property. Deoxyhemoglobin is paramagnetic, meaning application of a magnetic field will make the molecule magnetic, kind of like a piece of iron. Needless to say, the characteristics are very weak, but they happen to be measurable.

Well, we have known for over 100 years, that neurons in your brain use less oxygen when they are not being used. Makes sense, doesn't it? Therefore when areas of your brain are active, more oxyhemoglobin is converting to deoxyhemoglobin than in other areas of the brain.

There are some very good resources on the internet that describe how an MRI scanner works, so I wont really go into it here too much. To sum up: The big magnet in the machine is turned on an aligns hydrogen atoms in your body in a certain direction. This action gives the atoms some energy. When the field is turns off, the hydrogen atoms go back how they want, they relax, and give off some energy. The released energy is detected by the machine, and it uses these measurements to make an image. Most of the hydrogen atoms in your body are in water molecules, which is why such differences can be seen between tissues that have more or less water in them, like bones, organs and blood vessels.

fMRI uses an MRI scanner to detect the differences in magnetic signature of oxyhemoglobin and deoxyhemoglobin over and above the normal energy release of hydrogen atoms. But here is the most important point of fMRI: It is a statistical analysis. It is not a direct measurement of neural activity. In order to image an area where there is activity, a thought has to be made, and an image taken a couple of seconds afterward (because the oxygen conversion is not instantaneous). This process must be repeated a couple of times to be able to filter out false positives.

All an fMRI tells you is where there is more activity compared with other nearby regions. It does not tell you why a region is being stimulated. That is always up to the investigator.

Which is why fMRI studies paired with known pseudoscience almost always gets into trouble.

fMRI and the Woos
It doesn't take long to find fMRI being used to justify nonsense. Here is a short list.

fMRI and Reflexology (pressing areas on your feet helps areas of your body)
fMRI and Acupuncture (and again and again and again....)
fMRI and Rolfing (what a name huh? Its essentially a woo form of massage)
fMRI and Healing Touch (or any of the non touching healing woo, like Therapeutic Touch or prayer, heck here's the whole kit and kaboodle)

and so on. While I am quite tempted to rip apart each and every one of those studies, I'd get distracted from my main point. Suffice to say, I simply can't understand why studies that concern these areas of obvious nonsense can't be constructed and carried out in a way that would make them respectable.

The main issue with these is that the fMRI is studying a neurological response of the patient, not the effect of the procedure. If the patient thinks that procedure is going to affect his liver, then neurons that are associated with the liver are likely to shine. If you try out two different procedures (say reiki and acupuncture) that are both supposed to treat the same malady, it is highly likely that the same area of the brain will light up.

The only way to get any information about the neurological response to a treatment with an fMRI is to compare it to a sham treatment, where the patient doesn't know if he is getting sham or not.

fMRI doesn't do what woos think it does.

Here is proof that the brain simply lights up in response to what the patient thinks is true, not what is actually true.

To summarize that study, 30 people, half of whom thought (pretended?) they were electrosensitive, were placed in an fMRI machine. They were told they would touch a cell phone and a thermode which would be set to various temperatures. They asked the participants to rate their level of discomfort from 1 to 5.

Not surprisingly both groups had similar fMRI responses to the thermode. Similar areas of the brain lit up, and similar levels of discomfort were reported. However, not surprisingly, when it came to the cell phone, those that were electrosensitive had areas of their brain light up that were not similar to those that were not sensitive to electricity.

Also not surprising was that fact that the cell phone was fake. It had no electricity at all.

The point is that fMRI doesn't do what more of the woo-mongers think it does. It doesn't prove that acupuncture is real, or that meridians or qi or and of these magical forces are real At best it may show that the patient thinks the magic energy is real, and nothing more.

Sidenote on this study: Reading the comments at the article, its clear the woo-meisters are complaining that the study was doomed to fail since it was done with an MRI machine (once again, unsurprisingly). If that were really an issue, why weren't the patients in constant acute pain? I'm sure Emily Rosa could work out a test for electrosensitivity.

Sidenote on woo and fMRI. Even if the brain lit up in response to these pretend therapies, it still doesn't mean that they provide any benefit. The list of things they pretend to cure is endless, an fMRI doesn't show anything with regard to their efficacy, only the patients perception.


On 12/4/08, 1:42 PM , Anonymous said...

I've done the FMRI - devastated to have spent money (a lot) and it didn't work. Very concerned about others in more extreme situations trusting this technology to "save" them. I really believed in it. To be fair, I think the intentions were good of the folks doing the testing. But there is something off about it.

On 9/6/10, 3:11 PM , Sildenafil said...

what thought provoking piece, I like your analysis and the way you expressed your opinions and thoughts about this particular topic, I share some of your thoughts, thanks for informative post!

On 6/13/11, 5:08 PM , free online games said...

i am so happy can be here and read this