Fast Toeplitz Based Iterative SENSE Reconstruction
Olafsson V, Lee S, Fessler J, Noll D
The University of Michigan
Reconstruction of images from sensitivity encoded imaging (SENSE) is computationally demanding when non-Cartesian k-space sampling is used in the data acquisition. Most of the reconstruction time is accounted for by the multiple matrix-vector multiplication required in conjugate gradient (CG) iterations. We present a new fast iterative reconstruction method for SENSE providing off-resonance correction with high accuracy. Simulations show that the proposed method is significantly faster than NUFFT based method with substantial correction for field map.