Some embodiments of the present disclosure disclose systems and methods for robust magnetic resonance image reconstruction that can model for all or nearly all components in the magnetic resonance imaging system, that possess compressibility features to speed up reconstructions, and that can be optimized such that the reconstruction can be performed within a short period of time.

A system and method for performing accelerated k-space shift correction calibration scans for non-Cartesian trajectories is provided. The method can include applying an MRI sequence, performing a calibration scan based on the MRI sequence using the non-Cartesian trajectory to acquire k-space shift data, wherein one or more partitions are skipped during the calibration scan, interpolating the skipped one or more partitions using the k-space shift data from adjacent partitions, and calibrating the MRI system using the k-space shift data and the interpolated k-space shift data. In some embodiments, an acceleration factor Acc can be defined and the calibration scan acquires k-space shift data for only one partition in every Acc partitions.

A method for producing a streak-suppressed MR image of a subject includes (i) generating an interference correlation matrix from M coil images, (ii) producing eigenvectors of the interference correlation matrix, and (iii) determining, from the subspace-eigenvectors, a projection matrix of the interference null space. The subspace-eigenvectors include a plurality of subspace-eigenvectors that span an interference subspace and a plurality of null-space-eigenvectors that span an interference null space. The method also includes generating, from N coil images derived from a respective one of N MR signals, N streak-suppressed multi-coil images by either (i) preprocessing the N coil images with the projection matrix and applying an image-reconstruction technique to each of the resultant N preprocessed coil images, or (ii) applying an image-reconstruction technique to each of the N coil images to obtain N reconstructed coil images and post-processing the resultant N reconstructed coil images with the projection matrix.

A plurality of reception coils are used to acquire magnetic resonance signals using parallel imaging and a k-space acquisition scheme, in which alternatingly the central region and one of the peripheral k-space portions are imaged in acquisition steps of a pair, such that after a partition number of such pairs, the whole k-space to be acquired has been imaged and a sliding reconstruction window can be applied to reconstruct an additional magnetic resonance image after each acquisition of such a pair. A time series of magnetic resonance images forming the magnetic resonance data set is then reconstructed from the magnetic resonance signals and sensitivity information regarding the plurality of reception coils by using the sliding reconstruction window and a reconstruction technique for undersampled magnetic resonance data. The k-space trajectories for each acquisition step are chosen to allow controlled aliasing in all three spatial dimensions including the readout direction.

In a method for recording measurement data, frequency-dependent parameters characterizing a gradient unit are loaded, a k-space trajectory planned for a MR measurement and having at least one frequency component is loaded, MR measurement data is acquired based on the planned k-space trajectory and reconstructing image data from the MR measurement data, wherein the planned k-space trajectory is corrected based on the at least one frequency component of the planned k-space trajectory and the frequency-dependent parameters, and an electronic signal representing the reconstructed image data is provided as an output of the MR system. The reconstructed image data may be stored and/or displayed. Advantageously, the correction can be employed flexibly for k-space trajectories with different frequency components.

In a method for acquiring measurement data using a magnetic resonance (MR) system having a gradient unit, frequency-dependent parameters characterizing the gradient unit of the MR system are accessed (e.g. loaded from a memory), a k-space trajectory of a RESOLVE (Readout Segmentation Of Long Variable Echo trains) sequence planned for a MR measurement is accessed, MR measurement data is acquired based on the planned k-space trajectory and reconstructing image data from the MR measurement data, and an electronic signal is provided that represents the reconstructed image data as an output of the MR system. The k-space trajectory may have a frequency component in at least one direction. The planned k-space trajectory may be corrected based on at least one frequency component of the planned k-space trajectory and the frequency-dependent parameters.

An excitation region setting method according to an embodiment includes: receiving a designation of a first region from a user, the first region being designated in a distortion-corrected image that is a magnetic resonance image in which an effect of a distortion of a magnetic field has been corrected; calculating an actual excitation region where a subject is to be excited, based on the designated first region and the effect of the distortion of the magnetic field; and correcting imaging conditions including at least one of an orientation of a slice plane that defines the actual excitation region, or a frequency of a high-frequency magnetic field applied to the subject, in such a manner that the calculated actual excitation region becomes closer to an ideal excitation region represented as the first region.

Techniques for correcting gradient non-linearity bias in mean diffusivity measurements by MRI systems are shown and include minimal number of spatial correction terms to achieve sufficient error control using three orthogonal diffusion weighted imaging (DWI) gradients. The correction is based on rotation of system gradient nonlinearity tensor into a DWI gradient frame where spatial bias of b-matrix is described by its Euclidian norm. The techniques obviate time consuming multi-direction acquisition and noise-sensitive mathematical diagonalization of a full diffusion tensor for medium of arbitrary anisotropy.

To avoid discontinuities between echoes from becoming large level differences in a k-space and to reduce artifacts generated in a reconstructed image due to the discontinuities in the k-space, an MRI apparatus of the present invention uses phase characteristics of multiple echoes to be collected after a single RF excitation to control an arrangement order in the k-space where the multiple echoes are arranged when a pulse sequence of the fast spin echo method that collects the multiple echoes using a spin flip after a single RF excitation is executed. The arrangement is controlled so that echoes with small phase errors between the echoes at least near the center of the k-space are adjacent to each other.

A method for recording a magnetic resonance image data set includes providing a magnetic resonance sequence. The magnetic resonance sequence includes at least one radio-frequency pulse and a slice-selection gradient pulse applied during or before the radio-frequency pulse, which is configured as non-constant. The method includes providing at least one correction term for compensating a magnetic field change of the slice-selection gradient pulse. The magnetic field change is ascertained via a transfer characteristic of the gradient system of the magnetic resonance system. The method also includes recording at least one magnetic resonance image data set with the magnetic resonance sequence using the correction term.