The most commonly used positron emitting radionuclides are carbon, nitrogen, oxygen, and fluorine. These radionuclides, which emit positrons, are normal components of human tissues either individually or when coupled with some other compound. PET, therefore, can provide an in-vivo study of naturally existing compounds in the human body.

A positron is a positively charged electron that is emitted from the nucleus of a radionuclide. Once emitted this positron (or anti-electron) travels several millimeters until it meets a free electron from the surrounding atoms, at which time an annihilation event takes place. The masses of the electron and positron are converted to electromagnetic radiation. Due to conservation of energy and momentum, two "annihilation" photons appear (two gamma rays). The total energy of these two photons will equal the rest mass of the original electron and positron (511 keV) and they will be emitted in a 180 degree opposite direction to one another.

A ring of detectors surrounds the patient and when two 511 keV gamma rays are simultaneously recorded by opposing detectors, an annihilation event is known to have taken place on or about a line connecting the centers of the two detectors. PET, therefore, uses the principle of annihilation coincidence detection.

To be recorded as an annihilation coincidence event, the gamma rays detected by the opposing detectors must occur within a very short interval of time called the coincidence window, which is usually ten to twenty nanoseconds. The PET scanner continuously records the coincidence events, and this data must be manipulated to generate images. In the past, filter back projection was used, but a newer technique called iterative reconstruction is favored because it eliminates some of the artifacts generated with filtered back projection. The reconstructed data can be displayed in a three dimensional rotating volume as well as standard tomographic slices in the transaxial, coronal, and sagittal planes.