L1-Regularized GRAPPA Kernel Estimate

Keywords

absolute accepted accuracy adaptively amplification amplitude applied apply applying array artifact assuming automatically bandwidth benefit bottle channel chosen cine clinical coefficients coil compact compactness computation computed computer condition core corporate corporation dashed degrade ding domain effectively either error established estimation evaluated even fidelity frame frames fully future gain generated hints identical implementation implemented improve improved incorporation independent indicates instead international introducing iterative kernel kernels larger latter linear longer magnitude matrix memory metric minimization model much nature needs noise norm noted novel often optimal optimization optimize parallel personal phantom pixel portion positive potentially practice preserves prevent problem proceedings promote propose rapid real receiver reconstruct reconstructed reconstruction rectangular reduce regression regularization regularized relation relatively remaining replace robustness sampled scanner select selectively sensing series shape shorter significantly simple software solution solved space sparsity square statistical still strength strengths studied studies support suppressed suppressing thick traditional transform typically uniformly utilizing warranted water wavelet wide written