Meeting Banner
Abstract #3805

Cardiovascular MRI Reconstruction with Data-Driven Sparsifying Transform

Qiu Wang1, Jun Liu1, Nirmal Janardhanan1, Mariappan S. Nadar1

1Imaging and Computer Vision, Siemens Corporation, Corporate Technology, Princeton, NJ, United States

Dynamic cardiovascular MRI facilitates the assessment of the structure and function of the cardiovascular system. To fit the data acquisition time, the data must be highly undersampled. Compressed sensing or sparsity based MR reconstruction takes advantage of the fact that the image is compressible in some transform domain, and enables reconstruction based on under-sampled k-space data thereby reducing the acquisition time. The design of such transform is a key to the success of the reconstruction. In this paper, we propose to use tight frame learning for computing data-driven transforms. Empirical results demonstrate improvement over the transform associated with the redundant Haar Wavelets.

Abstract PDF

Keywords

acceleration achieved acquisition adaptive anatomical applying appropriate approximates artifacts assessment basic bottom breaking canonical captured cardiac cardiovascular clinical coefficient coil coils columns comparing complex compressed comprises computer computing concatenating consider construct construction corporate cycles dashed described design designed disclaimer discrete domain driven dynamic effective empirical enables enforcing every example facilitates fact filter filters form formulation frame frames fully function generate generates generating going ground healthy highly illustrates includes initiate inside iteratively larger leads learn learning little marked matrix measure medicine minimization minimizations missing motion must noise norm operator paper parallel peak performance pixels plot preprint problem profile prolonged propose reconstructed reconstructing reconstruction reducing redundant regularization remainder represents rightmost sampled scanner sensing sensitivity shall simulate solid solves solving space sparsely sparsity spatial square structure subject submitted success system take takes technology temporal term terms texture thereby tight together training transform transforms truth unknowns validated vector vectorized version vertically vision volunteer wavelet wavelets