Image quality assessment of LaBr3 based 3D PET scanners

The main thrust for this work is the investigation and design of a PET scanner based on new lanthanum bromide scintillators. We use Monte Carlo simulations to generate data for a 3D PET scanner based on LaBr₃ detectors, and to assess the count-rate capability and the reconstructed image quality of phantoms with hot and cold spheres using contrast and signal-to- noise ratios. Previously we have shown that LaBr₃ has very high light output, excellent energy resolution, and fast timing properties which can lead to the design a time-of-flight whole-body PET camera. The data presented here illustrate the performance of LaBr, without the additional benefit of TOF information, although our intention is to develop a LaBr₃ scanner with TOF measurement capability. The only drawback of LaBr₃ is the lower photo-fraction that can potentially lead to a poorer spatial resolution than a high Z scintillator such as GSO or LSO. However, in 3D PET imaging where energy resolution is very important for reducing scattered coincidences in the reconstructed image, the image quality of a LaBr₃ scanner can potentially surpass other scanners. Our results show, that there is a gain in NEC (factors of 3 and 1.5 over GSO and LSO) arising from the reduced scatter and randoms fractions in a LaBr₃ scanner. The reconstructed image resolution is slightly worse than a high Z scintillator, but at increased count-rate, reduced pulse pileup leads to an image resolution similar to that of LSO. Image quality simulations predict reduced contrast for small hot spheres over an LSO scanner, but improved noise characteristics especially in a large 35×70cm phantom.