ToF Performance Evaluation of PET Modules With Digital Silicon Photomultiplier Technology During MR Operation

Abstract

In 2012, we presented the Hyperion-II <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D</sup> preclinical PET insert which uses Philips Digital Photon Counting's digital SiPMs and is designed to be operated in a 3-T MRI. In this work we use the same platform equipped with scintillators having dimensions closer to a clinical application. This allows an investigation of the time of flight (ToF) performance of the platform and its behavior during simultaneous MR operation. We employ LYSO crystal arrays of 4×4 ×10 mm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> coupled to 4 ×4 PDPC DPC 3200-22 sensors (DPC) resulting in a one-to-one coupling of crystals to read-out channels. Six sensor stacks are mounted onto a singles processing unit in a 2 ×3 arrangement. Two modules are mounted horizontally facing each other on a gantry with a crystal-to-crystal spacing of 217.6 mm (gantry position). A second arrangement places the modules at the maximum distance of approximately 410 mm inside the MR bore (maximum distance position) which brings each module close to the gradient system. The DPCs are cooled down to approximately 5-10° C under operation. We disable 20% of the worst cells and use an overvoltage of V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ov</sub> = 2.0 V and 2.5 V. To obtain the best time stamps, we use the trigger scheme 1 (first photon trigger), a narrow energy window of 511 ±50 keV and a minimum required light fraction of the main pixel of more than 65% to reject intercrystal scatter. By using a <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">22</sup> Na point source in the isocenter of the modules, the coincidence resolution time (CRT) of the two modules is evaluated inside the MRI system without MR activity and while using highly demanding gradient sequences. Inside the B <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> field without any MR activity at an overvoltage of V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ov</sub> = 2.0 V, the energy resolution is 11.45% (FWHM) and the CRT is 250 ps (FWHM). At an overvoltage of V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ov</sub> = 2.5 V, the energy resolution is 11.15% (FWHM) and the CRT is 240 ps (FWHM). During a heavy z-gradient sequence (EPI factor: 49, gradient strength: 30 mT/m, slew rate: 192.3 mT/m/ms, TE/TR: 12/25 ms and switching duty cycle: 67%) at the gantry position and an overvoltage of V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ov</sub> = 2.0 V, the energy resolution is degraded relatively by 4.1% and the CRT by 25%. Using the same sequence but at the maximum distance position and an overvoltage of V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">ov</sub> = 2.5 V, we measure a degradation of the energy resolution of 9.2% and a 52% degradation of the CRT. The Hyperion-II <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">D</sup> platform proofs to deliver good timing performance and energy resolution inside the MRI system even under highly demanding gradient sequences.