Analysis of scattered radiation cross-talk in a high-resolution gamma ray tomography detector with GATE Monte-Carlo simulation

Limited energy resolution in scintillation type gamma ray detectors leads to systematic errors in photon counting because the pulse height discrimination stages cannot accurately discriminate interactions with full respectively partial deposition of isotopic emission energy. The resulting error is a systematic positive count rate offset originating from erroneously counted scattered photons. The origin of scattering may be the detector itself (scintillation crystals and other construction material) as well as components of the setup, including the object of investigation. In this article results of a simulation study are presented which was carried out to assess the role of different design parameters for the count rate accuracy of a high resolution gamma ray detector used for transmission tomography. Thereby the simulation software Geant4 Application for Emission Tomography (GATE) was used. As a target parameter we evaluated the radiation cross-talk, which is the amount of erroneously counted interactions from photons which have undergone Compton scattering in neighbouring crystals. For the given detector design it was found that cross-talk obtained from the simulated data is in good agreement with experimentally determined cross-talk. It could further be shown by virtual detector design changes that radiation cross-talk can be reduced only to a degree that would still require additional software correction measures, such as scattering correction algorithms, if quantitative accuracy it demanded.