Fast Algorithms for Image Reconstruction with Application to Partially Parallel MR Imaging

Abstract

This paper presents two fast algorithms for total variation-based image reconstruction in partially parallel magnetic resonance imaging (PPI) where the inversion matrix is large and ill-conditioned. These algorithms utilize variable splitting techniques to decouple the original problem into more easily solved subproblems. The first method reduces the image reconstruction problem to an unconstrained minimization problem, which is solved by an alternating proximal minimization algorithm. One phase of the algorithm solves a total variation (TV) denoising problem, and second phase solves an ill-conditioned linear system. Linear and sublinear convergence results are given, and an implementation based on a primal-dual hybrid gradient (PDHG) scheme for the TV problem and a Barzilai-Borwein scheme for the linear inversion is proposed. The second algorithm exploits the special structure of the PPI reconstruction problem by decomposing it into one subproblem involving Fourier transforms and another subproblem that can be treated by the PDHG scheme. Numerical results and comparisons with recently developed methods indicate the efficiency of the proposed algorithms.

An Accelerated Linearized Alternating Direction Method of Multipliers

“…There are also several works concerning the tuning of the stepsize η t in L-ADMM, including [50,51,11].…”

A New Augmented Lagrangian Approach for $L^1$-mean Curvature Image Denoising

“…Thus, the minimum exists and is unique [45]. More importantly, the AL techniques discussed in Chapter 4 can be applied to the model (see, e.g., [36,45,53,126,170,175]). The so-called split Bregman iteration-based solvers have also received some attention [80,130].…”

Efficient, Convergent SENSE MRI Reconstruction for Nonperiodic Boundary Conditions via Tridiagonal Solvers

“…Variable splitting (VS) [3] [4] [5] is a versatile optimization approach for these cost functions. VS decouples a costly nonlinear optimization into simpler problems via the augmented Lagrangian (AL) framework.…”