A Time-Walk and Timing-Shift Correction Method for Dual-Ended Readout TOF-DOI PET Detectors

Positron emission tomography (PET) detectors suffer from time-walk when the leading edge discriminator is employed for timing pick-off as well as a timing-shift when thick crystals are utilized due to the depth-of-interaction (DOI) effect. In this study, a combined time-walk and timing-shift correction method was proposed for dual-ended readout PET detectors. To evaluate the proposed method, a pair of dual-ended readout PET detectors was constructed. Each detector was based on two Hamamatsu S14161-3050-08 silicon photomultiplier (SiPM) arrays coupled to both ends of an $8\times 8$ arrays of $3.1\times 3.1\times 20~{\mathrm { mm}}^{3}$ lutetium-yttrium oxyorthosilicate crystals with a 3.2-mm pitch. By employing the relationship between the energies and detected time differences of events, the time-walk and timing-shift were effectively corrected. The coincidence time resolution of the two detectors improved from $260.7~\pm ~1.0$ ps to $229.4~\pm ~1.0$ ps when a 400–650 keV energy window was used to select events. These results demonstrate the effectiveness of the proposed time-walk and timing-shift correction method.

I. Introduction

Time-of-flight positron emission tomography (TOF-PET) can enhance the signal-to-noise ratio (SNR) of the reconstructed images by estimating the interaction position of annihilation photons along the line of response (LOR) [1], [2]. Compared to non-TOF-PET, the TOF gain factor for the SNR of the reconstructed image roughly follows the formula , where L is the size of the object (such as the human body or brain), c is the speed of light, and is the coincidence time resolution (CTR) of the positron emission tomography (PET) system. This is equivalent to an effective gain increase of for the system sensitivity [3], [4].