报告主题： Edge Based Joint Multi-energy CT Image Reconstruction
报告人：陈韵梅 教授 （美国Florida大学）
报告摘要：We propose a new joint image reconstruction method by recovering edge directly from observed data, then reconstructing images with their edge information. For joint edge reconstruction our model minimizes the fidelity term together with the matrix Frobeniun norm or spectral norm or nuclear norm across the contrasts and L1 norm over the pixels. Derivation of data fidelity for the Jacobian of the multi-energy images also detailed. The new minimization problem yields an optimal O(1/k2) convergence rate, where k is the iteration number, and the per-iteration cost is low thanks to the close-form matrix-valued shrinkage. The numerical results indicate that the proposed method improves reconstruction efficient and accuracy compared to the state-of-the-arts.