Prism-PET 2x2 Super Module with Interleaved Multiplexing (iMux) Design

Introduction: Positron Emission Tomography (PET) is a nuclear medical imaging technique that entails the detection of pairs of 511 keV gamma-ray photons emitted during the annihilation process. Previously, we developed an innovative single-ended readout depth-encoding PET detector module with a segmented prismatoid light guide array, called the Prism-PET. In this study, we innovatively created the Prism-PET super module by assembling four newly designed tapered Prism-PET detector modules into a 2x2 array (Figure a), which can further enhance the high-resolution performance of Prism-PET detector modules by alleviating artifacts at edges and corners for the single detector block while optimizing the light-sharing effect to maximize signal-to-background ratio. The 2x2 Prism-PET super module is also implemented in the second prototype of the Prism-PET brain scanner (Prism-PET II). We have also incorporated our proposed interleaved multiplexing (iMux) scheme (Figure b) into the 2x2 super module to mitigate the large number of readout channels, consequently reducing scanner complexity, power consumption, heat output, and overall cost.

Methods: The 2x2 super module consisted of a 32x32 array of tapered 1.5x1.5x20 mm³ lutetium yttrium oxyorthosilicate (LYSO) crystals. The tapered and fully polished LYSO array was coupled in a 4-to-1 configuration to a 16x16 array of 3x3 mm² SiPM pixels on one side and to a prismatoid light guide array on the opposite side. Barium sulfate (BaSO₄) was inserted between crystals and light guides to ensure optical isolation. The proposed interleaved multiplexing scheme can be seen in Figure c where signals from every other SiPM pixels across rows and columns are shorted together (i.e., same ASIC channel). However, the nearest neighboring SiPM pixels that overlap with the same prismatoid mirrors are connected to distinct ASIC channels to facilitate for accurate signal demultiplexing across the entire detector block. Standard flood data acquisition was performed on the 2x2 super modules by uniformly exposing them with a 3-MBq Na-22 point source placed 15 cm away above the center of the super modules. Coincident events’ position encoding in a light-sharing PET module with iMux readout was performed using the weighted compensation Raise-to-the-Power (WC-RTP) method. The performance of the 2x2 Prism-PET super modules were experimentally characterized in terms of flood histogram, energy resolution, depth-of-interaction (DOI) resolution, as well as timing resolution.

Results: The flood histogram of Na-22 uniform irradiation for the 2x2 Prism-PET super module is shown in Figure d. Excellent crystal identification was achieved across the entire array of 1024 LYSO crystals, demonstrating that the decoding error of the super module is negligible. Figure e illustrates the energy resolution for the entire array of 1024 crystals and Figure f reveals a mean energy resolution of 7.6% for the entire super module. Figure g shows the energy-based and optical photon TOF (oTOF) based DOI histograms of 2x2 super module with DOI resolution of 2.2 mm and 5.1 mm FWHM across all crystals, respectively. Figure h shows the TOF histograms of the 2x2 super module with iMux readout. The list-mode data were processed to filter inter-crystal scatter (ICS) events in the photopeak. The best average coincidence timing resolution (CTR) achieved 311 ps FWHM across all crystals following additional corrections involving DOI information, the nearest-neighbor light-sharing timestamps (LSTS), and line of response (LOR)-based fine-tuning.

Conclusions: We have developed and experimentally characterized our proposed 2x2 Prism-PET super module combining using the iMux scheme, which improves on the already cost-effective and high-resolution Prism-PET detector module and provides 16-to-1 crystal-to-readout multiplexing with high energy, DOI and timing performance.

• Download figure
• Open in new tab
• Download powerpoint