Room: Karl Dean Ballroom A1
Purpose: Light diffusion and absorption within scintillators cause depth-dependence in their gain and blur for x-ray detection. These variations constitute noise that degrades the detective quantum efficiency (DQE) of indirect flat panel detectors (I-FPDs) at all spatial frequencies. We have developed experimental methods to directly measure these effects in scintillators, for the first time. The knowledge is used to explore FPD designs that maximize scintillator thickness for dose-efficiency, while minimizing loss in DQE(f) due to scintillator depth-effects.
Methods: Powder Gdâ‚‚Oâ‚‚S:Tb (GOS) and columnar CsI:Tl scintillators (170-1200 Âµm thickness) were coupled to an ultra-high-sensitivity optical camera, and a micro-slit synchrotron beam (20 Âµm, 32 keV) was used to localize x-ray interactions to fixed depths within them using a â€œside-onâ€? irradiation geometry. Each sampleâ€™s depth-dependent average gain g(z) and spatial resolution MTF(z,f) was determined directly from depth-localized interaction images. A modular I-FPD allowing both â€œfront-irradiationâ€? (FI) and â€œback-irradiationâ€? (BI) of an interchangeable scintillator was used to evaluate the impact of irradiation geometry on imaging performance. This device was also used to estimate the DQE(f) of dual-screen detectors comprising FI â€œtopâ€? and BI â€œbottomâ€? scintillators coupled to a bidirectional sensor array. MTF(z,f) and g(z) measurements were used to determine the optimal dual-screen configurations for higher-energy imaging , e.g. RQA9 beam used in CBCT. The results were compared to the imaging performance of commercial I-FPDs.
Results: GOS scintillators had more depth-dependence in g(z) and MTF(z,f) than thicker CsI:Tl scintillators (e.g. 290 Âµm GOS vs. 1200 Âµm CsI:Tl). BI provided superior spatial resolution and x-ray sensitivity to FI, especially in scintillators with stronger depth-dependence in g(z) and MTF(z,f). Dual-screen DQEs exceeded those of current I-FPDs at low frequencies (DQE(0)=0.8, 45% improvement) and were comparable at higher frequencies.
Conclusion: Direct knowledge of g(z) and MTF(z,f) can inform potential I-FPD improvements through BI and dual-screen design concepts.