Mathematics

In computed tomography, the quality of a tomographic reconstruction can be degraded by the effects of a large, distributed radiation source. Large, distributed sources can arise if the focal spot itself is large, and from off-focal radiation. A large, distributed radiation source results in general blurring, as well as halo-like artifacts at boundaries of reconstructed images. We present an iterative, statistical, sinogram restoration data preprocessing method developed by Rivière et al. in the paper "Penalized-Likelihood Sinogram Restoration for Computed Tomography" that can be used to correct for the effects of a large, distributed source. This is done by estimating ideal transmitted intensities from the measurement data and knowledge of the structure of the distributed source by minimizing a quadratic penalized weighted least squares objective function prior to reconstruction. We demonstrate the effectiveness of this method by comparing the result of a reconstruction of simulated 3D cone beam neutron transmission data from a large, distributed neutron source to a reconstruction of the ideal transmitted intensities computed by applying the sinogram restoration method to the same data.