Simulation study of annihilation gamma‑photon transport in concentric 8‑ring brain‑PET with monolithic thin LYSO slab detectors

To increase spatial resolution and sensitivity of brain-PET, we are developing an advanced brain-PET scanner. In this project, we use Monte Carlo simulations to investigate the annihilation photons transport in multi-concentric 8-ring detector system with 7mm wide airgaps between rings, and 3mm-thick monolithic LYSO slabs optically coupled to ultrafast SiPM. A spherical radioactive water phantom (170mm diameter) was adopted as a brain model. We determined that transport of annihilation photons through a multi-concentric-ring detector system can be described by mono-exponential functions and can be divided into three distinct processes: (i) intra-ring true single and pseudo-single via photo-absorption events and intra-ring Compton interactions– dominant; (ii) inter-ring forward Compton scatter; (iii) inter-ring backward Compton scatter – weakest. Inclusion of Compton events increases sensitivity by >2.5X. Application of thin (3mm) slabs LYSO significantly enhances the detector spatial resolution, vs. thick slabs ΔDOI (0.3mm) and ΔXY resolution (0.5mm) for 3mm vs. a single 24mm slab (1.3mm ΔDOI, 1.0mm ΔXY). The time difference between Compton scatter and the following photo absorption event in inter-ring scatter can be measured with sufficiently fast detectors and optimized airgaps sizes can be used to further improve spatial resolution. Introduction of multi-concentric ring geometry with thin monolithic LYSO slabs incorporating first-order Compton scattering events into coincidence processing not only enables the recovery of otherwise rejected true coincidences but also provides critical benchmarks for optimizing detector module design, adjusting ring-ring airgaps, thereby laying a methodological foundation for the development of next-generation, ultrahigh-resolution dedicated brain PET imaging systems.