Fast Approximators for Least-Norm Reconstructions of Undersampled Non-Cartesian MRI Data

Keywords

according accurate acquisition added adjoint almost approximate approximation arguments asserts authors averaging beyond biomedical calculation challenges channel characterized clinic clinical closely commonly compensation compressive computation computational computationally conjugate consider considered constant constrained conversely core correction crosstalk definiteness density descent describe design despite determined determining direct directly discrete dual efficient employing enforce engineering equality equation equivalent estimation exact example fast fidelity fortunately full function generating gradient greater ideally identity implementation impractical incorporated instead intensive iteration iterations iterative kernel least loss machine matches mathematical matrix maximize mayo mink mitigate model necessarily negativity nonnegative norm note novel numerical optimization optimized origin phantom physiology popular positive possessing problem profiles projected projector proposed prunes radiology rank readily real realize reconstruction reconstructions recovering reduce referred related repeat required requires requiring resolution response sampling scaled scenarios semantically sensing several shot since solution solving spectral spiral squares standalone strategies strategy submitted suppose system term theory true uniform unit zero