Maxwell parallel imaging

Purpose To develop a general framework for parallel imaging (PI) with the use of Maxwell regularization estimation sensitivity maps (SMs) and constrained optimization parameter‐free image reconstruction. Theory Methods Certain characteristics both SMs images are routinely used to regularize otherwise ill‐posed optimization‐based joint reconstruction from highly accelerated PI data. In this paper, we rely on fundamental property SMs—they solutions equations—we construct subspace all possible SM distributions supported in given field‐of‐view, promote that belong subspace. addition, propose scheme reconstruction, as second step, once an accurate is available. The resulting method, dubbed (MPI), works 2D 3D, Cartesian radial trajectories, minimal calibration signals. Results effectiveness MPI illustrated various undersampling schemes, including radial, variable‐density Poisson‐disc, Cartesian, compared against state‐of‐the‐art methods. Finally, include some numerical experiments demonstrate memory footprint reduction constructed basis help tensor decomposition, thus allowing full 3D reconstructions. Conclusion provides physics‐inspired method efficient arbitrary scans.