Positron Emission Tomography (PET) is not in fact limited to brain metabolism studies, but is a powerful and general imaging technique. PET works because of the fact that when a positron and an electron mutually anihillate, two gamma rays are emitted with a specific energy (about 511KeV each) and in precisely opposite directions. By detecting these two gammas it is possible to localize that anihillation event, which can be safely assumed to be in the near vinicinity of a positron emitter.
Any element which decays by the positron (also known as beta plus) pathway can be used as a PET tracer; commonly used are fluorine-18 and oxygen-15, which are chosen for their short half-lives, ease of production (via proton bombarbment at 20-40MeV in a reasonably inexpensive cyclotron) and chemical utility.
These elements are then chemically incorporated into biologically useful molecules; in the case of F18, usually fluorodeoxyglucose, a useful label for sugar metabolisis imaging, and sometimes (as in my lab) fluorestradiol, used for breast cancer studies. O15 is usually used as water to do blood flow/volume studies.
PET tracer production is one of the few places outside of a reactor facility one is likely to encounter Curie-quantity radioisotopes. A Curie, like a Farad, is an extremely large unit; microcuries or picocuries are more commonly encountered.