The NEC dependence of different scintillators for positron emission tomography

In positron emission tomography (PET), the concept of noise equivalent count rate (NEC) is a measure of image quality. It has been shown that the local signal-to-noise ratio in the images reflects the global signal-to-noise ratio, which, in turn, can be related to the NEC. Factors that affect the NEC include the scanner geometry and scintillator material. The peak NEC has long been considered an indicator of PET scanner performance, whereas the sensitivity, represented by the slope of the NEC curve at the origin is equally important. This initial slope is proportional to the true coincidence sensitivity times the factor (1-SF), where SF is the scatter fraction. The peak NEC value is a strong function of scanner geometry and scintillator material, scatter fraction, system dead time and random coincidence fraction. The scatter fraction depends mainly on the low level discriminator setting of the system. Access to time-of-flight information may have a strong impact on the signal-to-noise properties and hence on the NEC performance. However, this work will focus specifically on the effect of the scintillator properties on the NEC, in particular the stopping power, light output, decay time and the interaction with the electronics. Three scintillators have been compared, the standard BGO, the fast high density LSO and a new fast, high light yield scintillator LaBr₃. The scintillators have been compared for the fixed scanner geometry of the ECAT EXACT.