Characterization of a high resolution and high sensitivity pre-clinical PET scanner with 3D event reconstruction

COMPET is a preclinical PET scanner aiming towards a high sensitivity, a high resolution and MRI compatibility by implementing a novel detector geometry. In this approach, long scintillating LYSO crystals are used to absorb the gamma-rays. To determine the point of interaction (P01) between gamma-ray and crystal, the light exiting the crystals on one of the long sides is collected with wavelength shifters (WLS) perpendicularly arranged to the crystals. This concept has two main advantages: (1) The parallax error is reduced to a minimum and is equal for the whole field of view (FOV). (2) The P01 and its energy deposit is known in all three dimension with a high resolution, allowing for the reconstruction of Compton scattered gamma-rays. Point (1) leads to a uniform point source resolution (PSR) distribution over the whole FOV, and also allows to place the detector close to the object being imaged. Both points (1) and (2) lead to an increased sensitivity and allow for both high resolution and sensitivity at the same time, while keeping a low number of readout channels. In total, COMPET incorporates 1080 readout channels (600 crystals, 480 WLS). It has an axial FOV of 80 mm and adjustable bore opening between 30 mm and 80 mm. It consists of four modules with five layers each. Simulations show a PSR of below I mm in the transaxial plane and a sensitivity of up to 16\% in the center of the FOV. The readout is based on time over threshold signals, sampled with an FPGA, which allows for the measurement of high event rates at the order of mega-counts per seconds. Its compact design and compatibility to high magnetic fields will allow to use it as an insert for an already existing MRI scanner. A first semi-layer with 12 WLS and 10 LYSO crystal was built and connected to the COMPET readout system. Coincidence data between this module and a tagger crystal using a small Ge-68 and a 60 MBq F-18 source was taken. (C) 2011 Elsevier B.V. All rights reserved.