Positron Emission Tomography (PET)


What is a PET scan?

     PET is one of the newest, most advanced methods for studying organs in the body such as the brain. Like an X-ray, or MRI scan PET is painless and relatively noninvasive (i.e., no surgery or opening of the body is required). Unlike X-rays and traditional MRI, PET does not produce a picture of the "structure" or anatomy of the brain, but rather it gives an image of  brain "function" or physiology. In other words, it can be used to image what the brain is doing.


How does it work? Steps in the PET process:

    In PET a chemical substance the investigator would like to follow through the brain is "labeled" with radioactive atoms. The labeled substance (called a radiotracer) is introduced into the body, usually by injection, where it migrates to the structures of interest. PET radiotracers emit positrons. Positrons travel a short distance in the brain before colliding with an electron. The positron and the electron annihilate each other and in the process emit two photons at 180 degrees to each other. The photons are detected as "pairs in coincidence" by a series of specialized detectors arranged in a ring around the subject. A computer is then used to assemble an image from all of the pairs of coincidental photons detected by the scanner. Areas of the brain which emitted the most photons will produce a more intense PET signal and areas which emitted few photons will produce a weak signal.

    Labeling is a process whereby a "tag" is attached to a substance that the investigator wants to follow. With PET labeling is accomplished by using a (very expensive) device called a cyclotron. The radioactive atoms created by the cyclotron decay by "emitting" particles called  "positrons" which can be monitored with a special array of detectors called a PET camera.  The substances that can be labeled for PET scanning are limited only by the imagination of the investigators and the physical half-life of the positron emitting label (half-life is the amount of time it takes for half of the label to decay -- the faster the decay, the less time between injection and scanning). The real power of PET is that atoms which naturally comprise the organic molecules utilized in the body can be labeled (turned into positron emitters). These atoms include oxygen, carbon and nitrogen. Since these atoms occur naturally in organic compounds, replacing the naturally occurring atoms with labeled versions leaves a compound that will behave in a manner identical to its unlabeled sibling, but which will be traceable in the body.

    In most cognitive studies using PET the labeled compound is oxygen (O15) and it is injected into the body in the form of radioactive water. The basic idea is that areas of the brain that are working relatively harder tend to get increased blood flow relative to areas that are not working as hard. This results in more labeled oxygen migrating to these areas  -- that is, these harder working regions have a higher oxygen concentration and hence a more intense PET signal than areas which are not working as hard.

    To obtain a PET image two conditions must be run. One where the subject engages in the cognitive process of interest (experimental or "task state") and one that involves all of the same processes as the task state except for the one of interest ("control state"). The PET image from the control state is then subtracted from the task state image. Here are two PET images from an experiment where subjects heard or read words (the control state task was viewing a fixation cross). As can be seen areas normally thought to subserve vision in the occipital lobe "light up" for read words while areas thought to be involved in audition and spoken language light up for spoken words.