Monte Carlo studies of x-ray energy absorption and quantum noise in megavoltage transmission radiography.

The subject contrast of bony anatomy in megavoltage medical radiographs is very low, making detection of bony landmarks difficult if additional noise sources are introduced into the images. One source of noise, which is inherent to the x-ray detection process, is x-ray energy absorption noise. X-ray energy absorption noise results from variations in the amount of energy deposited in the imaging detector per interacting x ray. These variations increase the noise content of the image. In this study, EGS4 Monte Carlo simulations of x-ray interactions in metal plate phosphor screen detectors have been performed to determine the distribution of energy absorption events within the phosphor screen. From these "absorbed energy distributions (AEDs)", the x-ray energy absorption noise and the quantum absorption efficiency of the detector are determined. These calculations are performed for a range of detector thicknesses (0.1-4 mm) and x-ray energies (0.1-10 MeV). A number of conclusions can be drawn from these investigations. (i) The x-ray absorption noise reduces the detective quantum efficiency (DQE) of metal plate/phosphor screen detectors by as much as 50% at energies used in megavoltage imaging (1-10 MeV). (ii) It is important to include secondary particle (electron) transport in estimating the quantum absorption efficiency of these detectors. For instance, the quantum efficiency of a typical portal detector is approximately 2%, even though 4%-5% of the incident photons are attenuated. (iii) The metal "conversion" plate commonly used in megavoltage imaging enhances the DQE of the phosphor screen by increasing the quantum absorption efficiency and reducing the magnitude of the x-ray absorption noise.