CS-SENSE or Denoised SENSE: The Influence of Irregular Sampling in L1 Regularized SENSE Reconstruction

Keywords

accelerate acceleration accounting achievable adapted addition almost altogether amplification answer application applies becomes best better bottom brain called candidate channel clinical coil combination compressed conditioned conditioning considered convergence converges correlation counteract curves decreases density determined disk encoding error especially established estimation expected exploits faster formulated frequency full fully function generalizing gradient great hamburg head homogeneous improve improvement included incoherent increasing influence inspired inverse investigate irregular isotropic iterations iterative king knowledge leads least limiting literature maps math matrix measured moderate needed noise normalized noticeable operator optimal optimization overdetermined parallel pattern patterns periphery peter posed prior problem promising proposed quality question reconstructed reconstruction reduced reduces reduction regular regularization regularized regularly resolution respect retrospectively sampled sampling scanner scheme selects sense sensing sensitivities sensitivity sets severe since solution solve soviet space sparsity speed squares support term theory transform unchanged uniform usually variable wavelet words