Practical aspects of energy weighting in X-ray imaging

The technique of Energy Weighting takes advantage of the upcoming energy resolving properties of X-ray pixel detectors. It can improve the image quality (in terms of signal-to-noise ratio compared to integrating detectors) up to a factor of 1.5 as shown in [J. Giersch et al. (2003)]. The optimal weighting function depends on materials and thicknesses, but a good approximation is 1/E³. In the presence of scattered radiation the usability of lower photon energies is limited. For this case we derived an adapted weighting function. Furthermore, we studied the influence of detector properties on the image quality improvement. Our results show that the energy bin size (i.e. number of energy channels) of the imaging system has a comparatively small impact on the benefit. Additionally, with optimised bin border positions, it is possible to achieve about 90% of maximum improvement with only three bins. To investigate the practical feasibility of Energy Weighting with present detectors, we used the Medipix2 in our X-ray setup. The Medipix2 is a photon counting hybrid pixel detector, which is in principle capable of energy sensitive X-ray measurements. By moving the thresholds together with the tube voltage we are able to take energy resolved images. Application of Energy Weighting results in SNR improvement in good agreement with theory. With Monte-Carlo simulations on computed tomography (CT) we are able to suppress spectral artifacts using Energy Weighting: an increased image quality without cupping artifacts can be achieved by applying the method after CT reconstruction for each energy bin.