Metal artifact reduction in cone-beam X-ray computed tomography using statistical iterative reconstruction

In general, image reconstruction from metal-embedded data causes streak artifacts that reduce the quality of the reconstructed image. In this paper, the attempt has been conducted to solve the problem of metal artifacts in cone-beam X-ray CT. The proposed method is applied directly to CT measurement data. First, the metal objects in the reconstructed image are detected and then reprojected to obtain the raw data using cone-beam reconstruction. The missing projections caused by the metal objects are replaced with their surrounding unaffected area through interpolation. Finally, the corrected raw data are reconstructed with the convex algorithm, which is the iterative algorithm for maximizing the likelihood function. The reconstructed images of metal artifact data using statistical reconstruction tends to be superior to conventional filtered backprojection (FBP) reconstruction.