Beam Hardening Correction in X-Ray Computed Tomography: A Comparison of Two Iterative Model-Based Reconstruction Methods

While the expenses for computational power decrease, iterative reconstruction methods for x-ray computed tomography (CT) that allow the use of improved model assumptions, become more attractive. In applications for non-destructive testing one of the most common degradations of image quality with standard x-ray CT reconstruction methods is beam hardening. Techniques for beam hardening correction using filtered back projection usually account for a single material only. In the case of multiple materials further improvement can be achieved using iterative methods. We developed a framework for reconstruction based on the principles of maximum-likelihood estimation that accounts for beam hardening by adapting the algorithms of L. Brabant (2012) and B. De Man (2001). These algorithms are tested in simulation studies as well as in measurements with multiple materials. We compare the results and show that iterative model-based reconstruction methods turn out to be a promising tool in order to reduce beam hardening.