Image reconstruction based on adaptive weighting enhanced total variation for CT with a displaced detector array

In order to reduce manufacturing costs or radiation doses,one of the practical needs in computer tomography (CT) is to obtain a larger field of view (FOV) with a limited-size detector,which can be realized by placing the detector with a laterally offset. However,conventional reconstruction algorithms cannot accurately reconstruct images from these projection data with detector offset. To solve this problem,this paper proposes a reconstruction model based on adaptive weighting enhanced total variation minimization (WAwrTV) and its Chambolle-Pock (CP) solving algorithm. The model constructs an adaptive weighting enhanced gradient norm as a regularization term and includes a bias-weighted fidelity term. In experiments,Projections with detector offset were simulated using both the ThoraxRecon phantom and real CT images. Reconstruction images are qualitatively and quantitatively analyzed. Results demonstrate that the proposed algorithm effectively reconstructs projection data with detector offset,improves reconstruction accuracy,and exhibits good noise resistance.