A modified 4D ROOSTER method using the Chambolle-Pock algorithm

Proc. of the Third International Conference on Image Formation in X-Ray Computed Tomography

Abstract: The 4D RecOnstructiOn using Spatial and TEmpo- ral Regularization method is a recent 4D cone beam computed tomography algorithm. 4D ROOSTER has not been rigorously proved to converge. This paper aims to reformulate it using the Chambolle & Pock primal-dual optimization scheme. The convergence of this reformulated 4D ROOSTER is therefore guaranteed.