Detected Events

Ideally the only prompt events registered by the detectors are those which arise from "real" positron annihilation. However, a number of other unwanted events that satisfied the coincidence criteria are also registered. The detection of unwanted events causes noise and degradation in spatial resolution. Therefore, their correction is essential to improve the quantification. In general, five types of event can be detected by PET scanner, and four of them are illustrated in Fig. 2.4.

A true coincidence refers to an event that two photons are emitted back-to-back from a single positron-electron annihilation, and are detected simultaneously by opposing detectors within the region of coincidence detection and within the coincidence timing window of the system.

Figure 2.4: Types of coincidence event recorded by a full-ring PET system. The white circle indicates the site of positron annihilation, and the solid line represents the gamma ray, (A) true coincidence, (B) scattered coincidence, (C) random (or accidental) coincidence, and (D) multiple coincidence. The mispo-sitioned line of response is indicated by the dashed line.

Figure 2.4: Types of coincidence event recorded by a full-ring PET system. The white circle indicates the site of positron annihilation, and the solid line represents the gamma ray, (A) true coincidence, (B) scattered coincidence, (C) random (or accidental) coincidence, and (D) multiple coincidence. The mispo-sitioned line of response is indicated by the dashed line.

Scattered coincidence occurs when one or both of the emitted photons undergo a Compton scatter interaction in tissue. Compton scattering causes a loss in energy of the photon and change in direction of the photon. Since the direction is changed, the origin where the photons were emitted cannot be located correctly and, as a result, the event is mispositioned, leading to decreased contrast and deteriorated quantification.

A random (or accidental) coincidence occurs when two unrelated photons, which have not originated from the same site of positron annihilation, strike opposing detectors within the coincidence timing window. Since the random events are produced by photons emitted from unrelated decays, they are spatially uncorrelated with the activity distribution. The random coincidences are a source of noise, the rate of which is approximately proportional to the square of the activity in the field of view (FOV). The performance of PET scanner for high count rate studies is degraded and therefore, correction for randoms is necessary.

Multiple events are similar to random events, except that three photons originated from two positron annihilations are detected within the coincidence timing window. Because of the ambiguity in positioning the event, these coincidences are normally discarded.

A single event for which only one photon is emitted is also possible due to some physical factors. The single events are usually rejected by the coincidence detection circuit since detection of only one event within the timing window violate the condition of coincidence. Yet in practice, about 1-10% of single events are converted into paired coincidence events.

Was this article helpful?

0 0

Post a comment