Development and Evaluation of a Proof-Of-Concept Prototype Time-Of-Flight PET System Based on High Quantum Efficiency Multi-anode PMTs

Objectives We evaluated a proof-of-concept prototype time-of-flight (TOF) PET system (average coincident resolving time = 340 ps) based on a high quantum efficiency (QE) multi-anode PMT with low transit time spread and an FPGA-based time-to-digital converter (TDC). We performed various phantom studies to demonstrate the superb timing performance of this system.

Methods A single-ring PET system of 40 detectors, each containing 15 × 15 L_0.95GSO 3 × 3 × 20 mm³ scintillation crystals coupled to a Hamamatsu H10966A-100 PMT (8 × 8 anodes, 33% QE at 420 nm, eight dynode stages) employs FPGA-based 40 channel dual-phase tapped-delay-line TDCs equipped with a Xilinx Virtex 6^® device (1). We estimated spatial resolution using a ²²Na point source (10 μCi). We also acquired the images of NEMA IEC Body and 3D Hoffman Brain phantoms. The images were reconstructed using a TOF LM-OSEM algorithm. In addition, we applied a simultaneous activity and attenuation estimation algorithm (maximum-likelihood activity and attenuation correction factor) to the phantom data. To further demonstrate the capability of TOF, we generated body phantom images using only partial (50% and 75 %) ring geometries.

Results The transverse (axial) spatial resolutions were 2.15 mm (2.41 mm) at 1 cm off axis. Phantom images reconstructed with TOF information showed higher contrast compared with non-TOF reconstruction. The body phantom image with TOF produced uniform background even without any attenuation correction. Simultaneous estimation algorithms generated an activity image comparable to the image whose attenuation was corrected using CT-based attenuation map. In partial ring geometries, TOF images yielded fewer artifacts and higher contrasts than non-TOF images.

Conclusions We evaluated a prototype TOF PET system based on multi-anode high QE PMTs. The system generated high-quality phantom images demonstrating its fine timing resolution. We will further improve the performance of the developed PET system and hope to present quantitative results in the future.