Dual-energy computed tomography (DECT) is widely used to identify explosives or otherwise characterize substances relevant to transportation security screening. However, DECT data taken with energy-integrating x-ray detectors are susceptible to the effects of beam hardening, which can introduce cupping and streaking artifacts into reconstructed images, thereby complicating image analysis. While photon-counting detectors can circumvent this issue by retaining the full spectrum of attenuation information for a probed material, the effects associated with charge sharing among pixels and pulse pile-up can introduce other errors if left uncorrected. Techniques were devised for using basis material decomposition (BMD) as a calibration to correct for the non-linear response of photon-counting detectors. The attenuation spectra from copper, aluminum, and polyethylene phantoms were used as basis functions that could reproduce the attenuations of other measured materials. Material properties relevant to detection, such as linear attenuation coefficient (LAC), electron density (ρe), and effective atomic number (Ze), can then be accurately calculated from energy-resolved CT data. In addition, the calibrated data could produce reconstructed images that were relatively free of the beam hardening artifacts associated with traditional DECT.
|