What RTK has is mostly based on iterative reconstruction but I guess you don’t have the projection images and the scanner geometry? I have never worked on any post-processing for such artifacts, since it’s motion and sampling issues in the tomography, working on it during reconstruction is more natural.