Sparse-view x-ray computed tomography reconstruction via mumford-shah total variation regularization

The regularization plays an important role in the sparse-view x-ray computer tomography (CT) reconstruction. Based on the piecewise constant assumption, total variation (TV) regularization has been widely discussed for the sparse-view CT reconstruction. However, TV minimization often leads to some loss of the image edge information during reducing the image noise and artifacts. To overcome the drawback of TV regularization, this paper proposes to introduce a novel Mumford-Shah total variation (MSTV) regularization by integrating TV minimization and Mumford-Shah segmentation. Subsequently, a penalized weighted least-squares (PWLS) scheme with MSTV is presented for the sparse-view CT reconstruction. To evaluate the performance of our PWLS-MSTV algorithm, both qualitative and quantitative analyses are executed via phantom experiments. Experimental results show that the proposed PWLS-MSTV algorithm can attain notable gains in terms of accuracy and resolution properties over the TV regularization based algorithm.