Analysis of time resolution in monolithic crystal PET detectors using different digital time extraction algorithms

New digital PET detectors are based in an early digitalization of the detector signals. In these systems, digital signal processing (with FPGAs, DSPs or both) is applied to extract the information from the detected events. Algorithms adapted to the available hardware resources should be proposed to detect gamma-ray energy, impact position and event arrival time. In this paper, eight different digital algorithms for time extraction are evaluated. Each of them is a combination of different signal processing methods, which are low pass filter interpolation, charge calculation, bipolar signal digital constant fraction algorithm and digital constant fraction algorithm. It is important to analyze how electronic noise, ambient noise, clock jitter and signal shaping affect to the time resolution obtained using the different proposed algorithms. The developed test allows us to find the best electronics and to choose the appropriated algorithm to be implemented. Results show that both the analog electronics shaping time and the noise have an important influence in timing resolution.