I learn new things every time I volunteer for a research study. Today, I learned that neuroimaging researchers have no idea how the fMRI works.

Oh, they're aware that there's a very large magnet in it, and that the people who paid for the very large magnet said NO METAL IN THE MRI ROOM in equally large block capitals. Beyond that they don't really have any idea. They know they're measuring brain activity, and I think some of them know what marker they're using (blood flow), but they don't appear to know by what arcane sorcery big magnet + blood = animated GIF of the brain.

[Hemoglobin is magnetically susceptible. Only very, very faintly, hence the mad scientist-grade 4T magnet in the MRI scanner. Hemoglobin that's carrying oxygen has a slightly different response than hemoglobin that's already dropped it off, and brain regions that are active get an influx of the oxygenated stuff. Awesomely powerful, but also awesomely picky, and if you don't do the stats right you can accidentally get results out of dead fish.]

They absolutely definitely don't know how anything else in the room works, especially if it has magnets in. They almost didn't let me into the scanner, as their little hand-scanning magnetic wand kept going off at the back of my head. If I have a plate in my skull, it's news to me. Maybe aliens?

Further experimentation showed that it wasn't going off on me, but at that spot in the room. Also a number of other spots in the room with nothing in them. I pointed out that we were under a bunch of fluorescent lights, which have electromagnetic ballasts. Two highly-educated neuroscientists gave me blank looks.

I also asked how, if there were no metal or magnetic objects allowed in the MRI chamber, they got the earphone/speaker things to work. This had apparently never occurred to them. A brief explanation of electromagnetic speaker drivers got the same blank look as the ballast comment.

I hope their baseline scan of my brain looked interesting. I mainly spent those eight minutes trying to work out how they got the sound piped in. I still don't know what they actually use, but I came up with hydraulics -- you could transmit the compression and rarification of the waveform in pretty much the same way they used to make delay line computer memory, and use it to move a speaker membrane at the end. Any liquid would do. Turing infamously suggested gin.

The second thing I learned is that you can get the attention of every neuroscientist in the room by telling one of them that you have synesthesia, and the MRI noises have a color. There are actually two sources of noise in the tube; the low tone thumping is the motor moving the scanner components back and forth, but the higher buzz is the adjustment magnet ramping up and down. The field requires a lot of rapid adjustments to get a picture, so essentially the thing is a gargantuan electromagnetic buzzer that gives off a sort of a nasty poky diode-y waveform. Goes right through the earplugs. Synthetic noises with lots of hard corners like that come out very bright and flat, and as it happened, the fMRI was a high enough tone that it was pretty much a very rapidly-dashed bright white line.

They did show me pictures of my own brain at the end, which was nice. The last time I did one of these, they wouldn't let me see the scan. The researcher even claimed my brain looked normal, although he may have just been polite.