Imaging study of a clinical PET scanner design using an optimal crystal thickness and scanner axial FOV

The aim of this study is to understand the trade-off between crystal thickness and scanner axial FOV in the use of high light output crystals for clinical PET imaging. Clinical scanner design has evolved towards using thick crystals and long scanner axial FOV, as well as TOF imaging. While Monte Carlo studies demonstrate that longer axial FOV and thick crystals will lead to higher scanner sensitivity, cost has prohibited the building of commercial scanners with >; 25-cm axial FOV. We performed a series of system simulations while varying the crystal thickness and scanner axial FOV. Our results show that there is an optimal crystal thickness which maximizes the rate of change in relative sensitivity as a function of crystal thickness (about 10 mm for LSO, 20-25 mm for LaBr3) and independent of the scanner axial FOV. Simulations of a line source cylinder show that a scanner design using 20-mm thick LaBr3crystals and 38-cm axial FOV can provide similar NEC to that achieved in an LSO scanner with 20-mm thick crystals and an 18-cm axial FOV, but with twice the crystal volume. While there is a gain in the NEC, indicating a reduction in total scan time there are additional benefits in the use of shorter crystals, such as improved timing resolution leading to improved image quality with TOF imaging. Also, reduced parallax error due to shorter crystals allows the design of a smaller ring diameter, which improves system sensitivity and also provides the ability to develop a clinical PET insert for use in a PET/MR system. Results from simulation studies of lesion phantoms will be presented for scanners using an optimized distribution of thin crystals with their intrinsic performance characteristics.