MRI-NUFFT: An open source Python package to make non-Cartesian MR Imaging easier

Motivation: Non-Cartesian imaging remains complicated to use for MRI due to the high computational cost of the Non-Uniform Fourier Transform for image reconstruction. Goal(s): To provide a uniform interface for reconstructing magnetic resonance images from non-Cartesian k-space data and a collection of non-Cartesian sampling trajectories Approach: We propose an open-source Python package (https://github.com/mind-inria/mri-nufft/) providing a standard interface to existing NUFFT libraries, with extended models for multi-coil imaging and static-field (B0) inhomogeneities correction. Results: MRI-NUFFT can generate sampling trajectories, compliant with hardware constraints, as well as simple forward/adjoint operations and density compensation for use in advanced image reconstruction scenarios. Impact: With MRI-NUFFT, non-Cartesian MRI trajectories and reconstruction algorithms become accessible, efficient, and affordable to everyone for research and education purposes.