Abstract
We evaluated a timestamp-based real-time coincidence system that is being developed for our high sensitivity brain-dedicated PET scanner. The coincidence system consisted of sorters, delay, coincidence processors, and multiplexers and was implemented in a Virtex-6 field-programmable gate array (FPGA). The coincidence processor used less than 10% of resources that is available in the FPGA chip. To validate the functionality of the coincidence system, single events generated from GATE Monte Carlo simulation toolkit was processed using the coincidence system as well as a Matlab code. Then, error rate was calculated by comparing the number of resultant coincidence pairs. For the simulation, a uniform phantom with activity of 3.7-111 MBq was scanned using a small-ring PET scanner with the axial field-of-view of 21 cm. The error rate was smaller than 0.1% and 5% in the number of estimated prompt and random coincidences, respectively. Then, the coincidence system was evaluated by implementing it on a prototype time-of-flight PET scanner. Compared to offline coincidence processing, count rate performance was improved more than 30%. Another advantage of the developed coincidence system is that it can be easily applied to different PET scanners without modifying its internal structure.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call