Scatter correction algorithm for cone-beam computed tomography

Abstract

Circular cone-beam computed tomography (CBCT) provides three-dimensional anatomy of an object from the projection data acquired in a single-rotation scan. Due to the large illumination area, a high level of scatter limits the image quality of CBCT by reducing the image contrast, increasing the image noise, and introducing inaccurate CT numbers in the reconstructed images. In this work, a scatter correction method using beam-block strips was presented. A single scan was performed on a cylindrical uniform phantom, and scatter was estimated from each projection data. Using beam-block strips that were positioned in a longitudinal direction to the rotation axis, detected signals in the shadow regions were estimated through interpolation. In addition, a separate estimation of signals was made through interpolation in the illuminated regions; in order to get rid of the effect of unwanted partial transmission through the beam-blocking material. The estimated scatter was extracted from the two estimates. A backprojection filtration (BPF) algorithm was applied to image reconstruction. The results showed that about 30% of the error due to scatter has been corrected by the proposed algorithm.