Description
Conventional positron emission tomography (PET) scintillator detectors (arrays of parallel scintillator rods with a photodetector readout at the back of the array) offer limited depth-of-interaction (DOI) resolution which reduces reconstruction accuracy and thus PET image quality. Modifications such as dual-readout arrays of parallel rods have improved DOI resolution but unfortunately increases costs significantly. Lack of spatial resolution and other factors (limited time of flight (TOF) information, reduced energy resolution) all contribute to the degradation of scintillation signal. Several experiments have been conducted to determine if the optical photon transfer/ extraction process can be improved upon and consequentially lead to an improved DOI method for PET scintillator systems while simultaneously limiting manufacturing costs. CsI:Tl and LYSO scintillation FWHM data were compared with several sources in a standard gamma ray spectroscopy experiment with results showing the FWHM of the LYSO crystal (Na-22) source was 23.3 keV FWHM. Additional data were taken using a translate method of scanning a PMT across the axis of a 3x3x15 mm3 LYSO crystal that was irradiated with two separate Co-57 and Cs-137 sources to localize scintillation events along the axis of said crystal. The data show that there is a 6% increase in the number of total counts under the photo peak using the Co-57 @ 122 keV source v. a Cs-137 @ 662 keV source v. distance of 11.5 mm. Bessel and Cubic Spline interpolations were then compared with the Bessel interpolation showing a more accurate fit to the raw data. An experiment was then devised by adhering a Lucite rod to the center of an LYSO crystal to quantify the number of optical photons collected at the PMT via several lengths. This experiment was conducted to determine signal loss with respect to potential 3D TOF PET systems that could utilize readout data from a second axis providing depth of interaction (DOI). Cs-137 (662 keV) and Mn-54 (835 keV) sources were used in this experiment due to the similar energies produced in the positron annihilation process. Graphs of FWHM v. Lucite length produced a highly linear fit (R2 = .99) for the Cs-137 source and a linear fit (R2 = .76) for the Mn-54 source. Following the prior experiment, an arrangement of five LYSO crystals were placed side by side, then an identical arrangement was place on top of and perpendicular to the “lower” arrangement. Each vertical surface was then coated with a reflective white coating to prevent light sharing with its adjacent crystal. The two crystal layers were then coupled to each other via optical grease to maximize photon transmission between layers, but not adjacent crystals. The intent is to determine DOI and localization of scintillation events in three dimensions to a spatial accuracy to that of the physical size of the LYSO crystal. Preliminary results show that a DOI can be achieved to a degree of (± 2 mm) in the x, y, and z directions. These results were limited by physical constraints in the lab (PMT size et al.) and the manufacturability of smaller and smaller crystals. The results of this experiment do show a correlation to crystal size and spatial resolution.
