Image Reconstruction for Few-view Computed Tomography Based on ℓ 0 Sparse Regularization

Abstract

Accurate CT images are expected to be obtained from low-dose/limited projection data. In this work, an ℓ0 sparse regularization based optimization model was investigated for few-view CT. With the aim to effectively solve the optimization model, original optimization problem was transformed following the framework of iterative reconstruction based on alternating direction (ADM) method. Experiments are demonstrated to validate the efficiency of the proposed algorithm.