A Novel Compressed Sensing Approach to Accelerated Quantitative MRI Using Model-Driven Adaptive Sparsifying Transforms

Keywords

absence accelerated acceleration accuracy accurately achievable acknowledge acknowledgments acquisition adaptive adjust aliasing allow although analytical anatomy applications apply approaches approximation artifacts available behavior block brain clinical coil collected combined complicated compressed conjugate constant control dependence dependencies described design dimension driven enhances equation error errors estimation example exploited fact financial fitting fold free fully functions generality glean good healthy help improves increments individual inexpensive initially interleaves intuitively inversion iterations iterative limiting linear locker look loss mapping maps matrix measured minutes model models near nearly noise normalized novel numerical operator outline parametric perfect phantom physics piecewise practice prescribed prior problem propagate properties propose proposed pulse quantitative random realistic receivers reconstructed reconstruction recovery regularization regularized rely representation residual respectively retrospectively root sampling scanner sense sensing series settings several simulated since solution space sparsity spin square step superior suppose system theoretical theory transform transforms typically uniformly update updates user utilize variable vector white yield zero