Post #1704495

2026-04-15 12:51 UTC

Accidentally went down a rabbithole trying to work out whether cornflakes contained enough iron to be problematic in an MRI scanner. It seems iron can experience magnetic forces 250 times its weight around the scanner, meaning things will lift off the ground if they&#39;re &gt;0.4% iron. https://mriquestions.com/uploads/3/4/5/7/34572113/panych_et_al-2018-journal_of_magnetic_resonance_imaging.pdf The most heavily fortified cereal I can find is Grape Nuts, with 28mg per 100g, which is to say, they&#39;re 0.028% iron. https://www.tesco.com/groceries/en-GB/products/285353353 That&#39;s more iron than you need to be eating, but still they&#39;d need 14x that much to lift. But they&#39;d slide along a surface if the coefficient of friction were 0.07 or below, which is very low but not unrealistic for the right materials. https://www.engineeringtoolbox.com/friction-coefficients-d_778.html So: if you use a really powerful scanner and some overly fortified cereal, on a particularly slippery plate, you can probably get it to move around a bit, but not dangerously so. Might be a fun scicomm demo if for some reason anyone trusted with an MRI scanner will let you try it. Be careful of the dust in the bottom of the packet, though — some of that&#39;s probably tiny particles of pure iron so now your MRI scanner just has stubble forever

Replies (8)

@andrewt@mathstodon.xyz 2026-04-15 15:09

Obviously I should add: this is idly looking into it with Google. Magnetism is really complicated and I'm no expert. If I've overlooked a complexity it'll probably reduce the effect, although as scanners get more powerful it'll likely increase. Normally you do these kinds of sums and it says "you are nine orders of magnitude away from anything happening" and you can assume the complexities don't matter. This little one-and-a-half orders of magnitude difference is *not much* — I'm confident in saying cornflakes in an MRI scanner aren't a hazard, but I've really no idea whether or how much they'd slide about

Open ##1716864
@BillBraine@newsie.social 2026-04-15 12:54

@andrewt https://www.youtube.com/watch?v=hSofzQURQDk

Open ##1716865
@tal@mastodon.social 2026-04-15 13:18

@andrewt what T rating is that for? Most I've ever been in is 3T, but I think 5Ts are becoming common, maybe even some 7Ts somewhere? Also, um... there's hardly any gadolinium fortified cereals out there?

Open ##1716866
@lethe@chaos.social 2026-04-15 13:38

@andrewt @ljrk Wish I would've known that back when I was working in MRI research for a short period. We definitely had some strong scanners too^^ But I can warmly recommend throwing a tennis ball with a 5 cent coin in it into the scanner ^-^

Open ##1716868
@Linza@kamu.social 2026-04-15 15:13

@andrewt MRIs are kind of terrifying

Open ##1716872
@memnus@swingset.social 2026-04-15 15:23

@andrewt you nerdsniped me on the question of "they can't be fortifying foods with ferromagnetic compounds, can they?" I assumed that bioavailable iron would be more similar to hemoglobin than iron shavings, but apparently I'm wrong: https://pubs.rsc.org/en/content/articlehtml/2020/fo/c9fo02370d And now, as a Grape-Nuts Enjoyer, I am mildly concerned.

Open ##1716873
@kranzi@mastodon.green 2026-04-15 15:28

@andrewt fortunately it doesn't work this way: iron atoms are NOT ferromagnetic, this happens only in the bigger christalline structure. Otherwise MRI wouldn't be possible for humans, because hemoglobin is ~3% iron and your blood would pull at you very strongly...

Open ##1716875
@cass@dice.camp 2026-04-15 15:28

@andrewt Just to be clear, there's no risk to eating grapenuts before getting an MRI? I'm probably not the only person who was wondering 😅

Open ##1716881