Efficient Blind Compressed Sensing Using Sparsifying Transforms with Convergence Guarantees and Application to MRI

TL;DR

This paper introduces a fast, convergent blind compressed sensing framework that jointly reconstructs images and learns sparsifying transforms, demonstrated on MRI data with highly undersampled measurements.

Contribution

It proposes a novel, globally convergent algorithm for blind compressed sensing that simultaneously recovers images and learns sparsifying transforms from undersampled data.

Findings

Faster reconstruction compared to synthesis dictionary methods

Proven global convergence of the algorithms

Effective MRI reconstruction from highly undersampled measurements

Abstract

Natural signals and images are well-known to be approximately sparse in transform domains such as Wavelets and DCT. This property has been heavily exploited in various applications in image processing and medical imaging. Compressed sensing exploits the sparsity of images or image patches in a transform domain or synthesis dictionary to reconstruct images from undersampled measurements. In this work, we focus on blind compressed sensing, where the underlying sparsifying transform is a priori unknown, and propose a framework to simultaneously reconstruct the underlying image as well as the sparsifying transform from highly undersampled measurements. The proposed block coordinate descent type algorithms involve highly efficient optimal updates. Importantly, we prove that although the proposed blind compressed sensing formulations are highly nonconvex, our algorithms are globally convergent (i.e., they converge from any initialization) to the set of critical points of the objectives defining the formulations. These critical points are guaranteed to be at least partial global and partial local minimizers. The exact point(s) of convergence may depend on initialization. We illustrate the usefulness of the proposed framework for magnetic resonance image reconstruction from highly undersampled k-space measurements. As compared to previous methods involving the synthesis dictionary model, our approach is much faster, while also providing promising reconstruction quality.