Abstract
2143
Objectives We have developed a submillimeter resolution small animal PET scanner using the virtual pinhole PET technology. The Microinsert-II system is designed to fit inside the Siemens Inveon PET/CT scanner’s bore and operate in conjunction with the Inveon system. In addition, it is designed to be a MRI compatible device that allows integration with the 4.7T Bruker small animal scanner.
Methods The Microinsert-II system consists of 48 SiPM based detectors which are arranged in a ring geometry with 12 transaxial and 4 axial detectors. Each detector module consists of a 20 x 20 LYSO crystal of dimension 0.8 x 0.8 x 3 mm3 coupled to a 4 x 4 SiPM array via a custom light guide. The system has an inner diameter of 6.5 cm and an axial depth of 6.6 cm. It is read out using a combination of custom-made and Siemens QuicksilverTM electronics that enables both stand-alone and simultaneous operation inside the Inveon PET/CT system.
Results We have optimized the light guide design and detector assembly techniques such that a 20 x 20 crystal array can be decoded using a 4 x 4 SiPM array using resistive readout and Anger logic. 48 such detectors have been assembled and tested. The average energy resolution of a given detector is 12% and the average timing resolution is 710 ps. Monte Carlo simulations show that for a Derenzo pattern phantom, spheres as small as 0.75 mm diameter can be resolved.
Conclusions We have completed the assembly of the Microinsert-II system which is expected to provide sub-millimeter spatial resolution as demonstrated by Monte Carlo simulations. It can operate in a high-resolution stand-alone mode, or in conjunction with Inveon small animal PET scanner where it can provide higher spatial resolution than a stand-alone Inveon system. Performance measurements of the stand-alone system using phantom and animal experiments will be presented.
Research Support This work is supported in part by the National Cancer Institute (R01-CA136554, R33-CA110011, and P30-CA91842) and the Washington University Center for High Performance Computing (funded by NIH grant NCRR 1S10RR022984-01A1).