References & Citations
Mathematics > Numerical Analysis
Title: Joint image edge reconstruction and its application in multi-contrast MRI
(Submitted on 6 Dec 2017 (v1), last revised 8 Dec 2017 (this version, v2))
Abstract: We propose a new joint image reconstruction method by recovering edge directly from observed data. More specifically, we reformulate joint image reconstruction with vectorial total-variation regularization as an $l_1$ minimization problem of the Jacobian of the underlying multi-modality or multi-contrast images. Derivation of data fidelity for Jacobian and transformation of noise distribution are also detailed. The new minimization problem yields an optimal $O(1/k^2)$ convergence rate, where $k$ is the iteration number, and the per-iteration cost is low thanks to the close-form matrix-valued shrinkage. We conducted numerical tests on a number multi-contrast magnetic resonance image (MRI) datasets, which show that the proposed method significantly improves reconstruction efficiency and accuracy compared to the state-of-the-arts.
Submission history
From: Xiaojing Ye [view email][v1] Wed, 6 Dec 2017 01:43:19 GMT (3634kb,D)
[v2] Fri, 8 Dec 2017 14:07:49 GMT (3475kb,D)
Link back to: arXiv, form interface, contact.