A 0.38 mm LYSO and GAGG checkerboard pattern detector with inter-crystal scatter identification capability for high-resolution mouse brain PET imaging

Ultra-high spatial resolution PET imaging, such as mouse brain imaging, is significantly influenced by detector’s intrinsic resolution (crystal size), position decoding accuracy, and the capability to identify inter-crystal scatter.In this work, a 0.38 mm LYSO and GAGG checkerboard pattern detector is proposed. The detector consists of a $28 \times 28$ array of $0.38 \times 0.38 \times 4 \mathrm{~mm} 3$ crystals, where each crystal is surrounded by crystals of different type. A 2 mm thick glass light guide was placed between the scintillator array and a 64-channel Hamamatsu MPPC. A PETsys TOFPET2 ASIC records both integrated charge (QDC) and time over threshold (ToT) for each MPPC channel to discriminate the LYSO, GAGG and inter-crystal scatter (ICS) events (ICS-QT). A Look-Up-Table based method was used to identify more ICS events (LYSO-Out and GAGG-Out).Flood images were acquired using a Na-22 point source. Majority of the 0.38 mm crystals were clearly resolved. The average peak-to-valley ratio (PVR) is 8.6 for GAGG and 6.8 for LYSO. A total of 37.6% of the measured events were identified as ICS events (ICS-QT + LYSO-Out + GAGG-Out).GATE simulation was used to investigate the fraction of the ICS events. The fraction of ICS events, which correspond to the events that interact with multiple crystals, is 68.0%. A simulated imaging study was also performed. By using a box-shaped scanner comprising four LYSO and GAGG checkerboard pattern detectors, 0.4 mm diameter hot rods can be resolved with an average PVR of 1.8.