Unfolded Low-rank + Sparse Reconstruction for MRI

Abstract

We apply the methodology of deep unfolding on the problem of reconstruction of DCE-MRI data. The problem is formulated as a convex optimization problem, solvable via the primal–dual splitting algorithm. The unfolding allows for optimal hyperparameter selection for the model. We examine two approaches – with the parameters shared across the layers/iterations, and an adaptive version where the parameters can differ. The results demonstrate that the more complex model can better adapt to the data.