(Nuclear) magnetic resonance imaging is used for non-invasively looking at a person's giblets and as such is really useful for doctors. Unlike x-ray imaging it is best for looking at squishy bits, x-rays are best used on bones.
The core of the technique is to put your subject in a VERY BIG magnet. The magnets used for medical imaging are typically 1.5Tesla or 3.0Tesla, this is hundreds of times the strength of a little bar magnet and approaching 100,000 times the strength of the earth's magnetic field. A big nmr spectrometer will have a field of up to about 20T.
I'm not going to try explaining magnetic resonance imaging or nuclear magnetic resonance, the linked wikipedia articles look pretty good. It's worth noting that you can make a good guess that someone is explaining nmr to a neophyte by the distinctive waving of their arms. nmr scientists seem inordinately fond of acronyms (DANTE, ADEQUATE, INADEQUATE, WEFT, NOESY, GRASE...) and I have the sneaking suspicion that they spend more time inventing their acronym than designing their experiment.
The idea with the mock magnet is that patients find the fairly confined space of the MRI machine a bit claustrophobic, but you can get them used to the idea using the mock magnet. Also if you're doing experiments in the machine you can check your bits and pieces are going to fit in offline.
I enjoyed my time with the high field magnet, they're clad in a shiny cryostat and they're fond of ferromagnetic materials (like steel), this means they tend to stand alone in the middle of the lab. You can make the owner of a high field magnet very nervous by casually waving a screwdriver around in the presence of his "precious". Even at a distance of 3metres a reasonable size magnet exerts a very noticeable pull. Once you've got the screwdriver stuck to the magnet it's rather difficult to remove. Have a look on YouTube for a whole pile of videos on introducing metal objects to MRI machines.
Magnetic resonance imaging is a development of chemical shift nuclear magnetic resonance (nmr) spectroscopy, which has been keeping chemists happy for years. People are inordinately scared of the word "nuclear", so for presentational reasons the "nuclear" is dropped from the name of the imaging technique.
An exciting aspect of nuclear magnetic resonance (and other areas involving big magnets) is the "magnet quench" phenomena, this is the problem that has caused so much trouble at CERN recently. High field magnets work by putting a large current (tens of amps) into a superconducting coil (or solenoid). The superconductors used in magnets have to be cooled by a combination of liquid helium (really cold) and liquid nitrogen (not quite so cold, but cheaper). Whilst superconducting, everything is fine - the current can flow for ever producing a magnetic field. However, if by accident or design, the superconductor stops superconducting then it heats up it's liquid gas jacket, which vapourises potentially very rapidly. Here's another video of a controlled magnet quench, featuring a pretty, shiny magnet. The first minute is best, the venting helium sounds like a strange musical instrument.
I'm idly wondering whether a magnetic resonance imager embodies the most Nobel Prizes of any device you're likely to meet in everyday life. Heike Kamerlingh Onnes won the 1913 physics prize "for his investigations on the properties of matter at low temperatures which led, inter alia, to the production of liquid helium". Bardeen, Cooper and Schrieffer got the 1972 prize for their theoretical work on superconductors. Lauterbur and Mansfield received the 2003 prize for medicine for the development magnetic resonance imaging. There's a whole bunch of other nuclear magnetic resonance prizes, arguably you could include Marconi for his development of "wireless telegraphy".
1. Hopkinson, I., R.A.L. Jones, P.J. McDonald, B. Newling, A. Lecat, and S. Livings, Water ingress into starch and sucrose:starch systems. Polymer, 2001. 42(11): p. 4947-4956.
2. Hopkinson, I., R.A.L. Jones, S. Black, D.M. Lane, and P.J. McDonald, Fickian and Case II diffusion of water into amylose: a stray field NMR study. Carbohydrate Polymers, 1997. 34(1-2): p. 39-47.