Denoising for Low-Dose CT Image by Discriminative Weighted Nuclear Norm Minimization

Low-dose computed tomography (LDCT) is an effective approach to reduce radiation exposure to patients. Lots of mottle noise and streak artifacts, however, are introduced to the reconstructed image. Current weighted nuclear norm minimization (WNNM) denoising method cannot remove the streak artifacts in LDCT image completely, even if many time-consuming iterations are adopted. In this paper, an effective image denoising algorithm, which is based on discriminative weighted nuclear norm minimization (DWNNM), is proposed to improve LDCT image. In the D-WNNM method, the local entropy of the image is exploited to discriminate streak artifacts from tissue structure, and to tune WNNM weight coefficients adaptively. Additionally, a preprocessed image is used to improve the accuracy of block matching, and the total-variation (TV) algorithm is applied to further reduce the residual artifacts in the recovered image. We evaluate the D-WNNM method on the simulated pelvis phantom, the actual thoracic phantom, and the clinical thoracic data, and compared it to several other competitive methods. Experimental results show that the proposed approach has better performance in both artifacts suppression and structure preservation. Particularly, the number of iterations required in the proposed algorithm is substantially reduced (only twice), when compared with that required in the WNNM method (at least eight iterations).