PURPOSE: To find comprehensive equations for the frequency-dependent MTF and DQE of photon counting detectors including the effect that the combination of crosstalk with an energy threshold is changing the pixel sensitivity profile and to compare the results with measurements. METHODS: The framework of probability-generating functions (PGF) is used to find a simple method to derive the MTF and the DQE directly from a Monte-Carlo model of the detection process. RESULTS: In combination with realistic model parameters for the detector, the method is used to predict the MTF and the DQE for different pixel sizes and thresholds. Particularly for small pixels, the modification of the sensitivity profile due to crosstalk substantially affects the frequency dependence of both quantities. CONCLUSION: The phenomenon of the pixel sensitivity profile, i.e. the fact that the choice of the threshold is affecting the detector sharpness, may play a substantial role in exploiting the full potential of photon counting detectors. The model compares well with measurements: with only two model parameters, the model can predict the MTF(f) and the DQE(f) for a wide range of thresholds.
PURPOSE: To find comprehensive equations for the frequency-dependent MTF and DQE of photon counting detectors including the effect that the combination of crosstalk with an energy threshold is changing the pixel sensitivity profile and to compare the results with measurements. METHODS: The framework of probability-generating functions (PGF) is used to find a simple method to derive the MTF and the DQE directly from a Monte-Carlo model of the detection process. RESULTS: In combination with realistic model parameters for the detector, the method is used to predict the MTF and the DQE for different pixel sizes and thresholds. Particularly for small pixels, the modification of the sensitivity profile due to crosstalk substantially affects the frequency dependence of both quantities. CONCLUSION: The phenomenon of the pixel sensitivity profile, i.e. the fact that the choice of the threshold is affecting the detector sharpness, may play a substantial role in exploiting the full potential of photon counting detectors. The model compares well with measurements: with only two model parameters, the model can predict the MTF(f) and the DQE(f) for a wide range of thresholds.
Authors: Joakim da Silva; Fredrik Grönberg; Björn Cederström; Mats Persson; Martin Sjölin; Zlatan Alagic; Robert Bujila; Mats Danielsson Journal: J Med Imaging (Bellingham) Date: 2019-10-15