Purpose: Single x-ray imaging (SXI) is an emerging experimental approach which measures the response of scintillators to single x-ray interactions using an ultra-high sensitivity camera. We propose a theoretical framework for computing frequency-dependent pulse height spectra (PHS(u)) using SXI data. In addition to allowing the computation of traditional detector performance metrics such as DQE(u), PHS(u) provides new insight for optimizing indirect FPDs as well as novel detector designs which use scintillators. Monte Carlo (MC) simulation of the Hybrid-AMFPI, a novel multi-layer detector consisting of a scintillator coupled to an a-Se direct detector, is used to validate and demonstrate the merits of our approach.
Methods: MC was used to simulate the Hybrid-AMFPI as a stack of gadolinium oxysulfide (GOS) and a-Se, using 10,000 input x-rays with an energy of 60 keV. SXI images were summed in 1D and Fourier transformed to construct frequency-dependent pulses, i.e. â€œFourier burstsâ€?. Fourier burst ensembles were then used to construct PHS(u). Moments of PHS(u) were used to calculate a frequency-dependent Swank factor, As(u), which was used to compute DQE(u). Simulations of both front irradiated (FI) and back irradiated (BI) geometries were carried out, with various GOS thicknesses.
Results: DQE(u) of the Hybrid-AMFPI was significantly higher than that of a-Se alone at low frequencies, and asymptotically approached that of a-Se at high frequencies. Hybrid-AMFPI performance was optimal using a BI geometry. MC results achieved convergence at approximately 2000 detected x-rays, which provides higher computational efficiency than other MC methods for DQE.
Conclusion: Our framework provides a method for using SXI data to compute PHS(u), which leads to DQE(u). Additionally, PHS(u) yields information about the distribution of single x-ray responses which may be used to gain insight into key features of detector performance, and to guide optimization of x-ray detectors that utilize scintillators.
Monte Carlo, Imaging Theory, Flat-panel Imagers