A computer-implemented method of correcting phase and reducing noise in magnetic resonance (MR) phase images is provided. The method includes executing a neural network model for analyzing MR images, wherein the neural network model is trained with a pair of pristine images and corrupted images, wherein the corrupted images include corrupted phase information, the pristine images are the corrupted images with the corrupted phase information reduced, and target output images of the neural network model are the pristine images. The method further includes receiving MR images including corrupted phase information, and analyzing the received MR images using the neural network model. The method also includes deriving pristine phase images of the received MR images based on the analysis, wherein the derived pristine phase images include reduced corrupted phase information, compared to the received MR images, and outputting MR images based on the derived pristine phase images.

First k-space data are received for a MRI examination of a subject in a first field of view (FOV), and second k-space data are received for a second field of view that is adjacent to or overlaps the first field of view. Alternatively, second k-space data, comprising phase and/or slice oversampling k-space data of the first field of view are retrieved from a non-transitory data storage medium. The first k-space data are reconstructed by using at least a portion of the second k-space data as phase and/or slice oversampling to generate a first extended image of a first extended field of view that encompasses the first field of view and extends into the second field of view. The first extended image is cropped to the first field of view to generate an image of the first field of view for the first MRI examination.

A computer-implemented method is provided for improving image quality with shortened acquisition time. The method comprises: determining an accelerated image acquisition scheme for imaging a subject using a medical imaging apparatus; acquiring a medical image of the subject according to the accelerated image acquisition scheme using the medical imaging apparatus; applying a deep network model to the medical image to improve the quality of the medical image; and outputting an improved quality image of the subject, for analysis by a physician.

A problem of time elongation in image reconstruction due to necessity of many iterations is solved, when an image is reconstructed by performing iterative approximation on measurement data acquired by an MRI apparatus without full-sampling. Density is estimated in the course of the iteration, and density is corrected using thus estimated density. Density estimation is performed not only on initial values but also on every iteration process, thereby enabling convergence with a smaller number of iterations.

The present disclosure relates to a method and a magnetic resonance imaging device for two-dimensional (2D) magnetic resonance (MR) imaging of a subject. The disclosure further relates to a corresponding computer program and a corresponding computer-readable storage medium. In one exemplary method, a k-space dataset of the subject is acquired using a simultaneous multi-slice technique. Therein, a blipped phase-encoding gradient is applied in a pseudo-random manner to achieve an incoherent undersampling at least in a k-space direction perpendicular to a slice select direction. A compressed sensing reconstruction is then performed based on the acquired k-space dataset to generate an MR image of the subject.

An information processing apparatus according to an embodiment of the present disclosure includes a processing circuitry. The processing circuitry obtains a first g factor generated by using first magnetic resonance data acquired through a first parallel imaging process performed by using a plurality of reception coils and a second g factor generated by using second magnetic resonance data related to a second parallel imaging process performed by using the plurality of reception coils. The second parallel imaging process is different from the first parallel imaging process. The processing circuitry adjusts the first g factor so as to reduce a difference between the first g factor and the second g factor.

A system for image reconstruction in magnetic resonance imaging (MRI) is provided. The system may obtain undersampled k-space data associated with an object, wherein the undersampled K-space data may be generated based on magnetic resonance (MR) signals collected by an MR scanner that scans the object. The system may construct an ordinary differential equation (ODE) that formulates a reconstruction of an MR image based on the undersampled k-space data. The system may further generate the MR image of the object by solving the ODE based on the undersampled k-space data using an ODE solver.

Method for eliminating aliasing artifacts in a magnetic resonance image, comprising the steps of obtaining a first and a second starting image (100a,100b) obtained by a determined acquisition sequence and using, respectively a phase encoding for columns, and a phase encoding for rows. Both the first and the second starting image (100a,100b) are organized in according to a matrix structure (m.Math.n) comprising a plurality of portions (101a,101b) arranged according to m rows and n columns, each of which is associated to a respective numerical value corresponding to the light intensity of the portion. The method provides a translation step for translating at least one between the first and the second starting image (100a,100b) with respect to a respective reference system, in such a way to minimize the differences among the numerical values of the homologous portions of the first and of the second starting image due to the fact that the first and the second starting image are obtained by a different encoding phase.

A computer-implemented method of correcting phase and reducing noise in magnetic resonance (MR) phase images is provided. The method includes executing a neural network model for analyzing MR images, wherein the neural network model is trained with a pair of pristine images and corrupted images, wherein the corrupted images include corrupted phase information, the pristine images are the corrupted images with the corrupted phase information reduced, and target output images of the neural network model are the pristine images. The method further includes receiving MR images including corrupted phase information, and analyzing the received MR images using the neural network model. The method also includes deriving pristine phase images of the received MR images based on the analysis, wherein the derived pristine phase images include reduced corrupted phase information, compared to the received MR images, and outputting MR images based on the derived pristine phase images.

A fully sampled calibration data set, which may be Cartesian k-space data, is used to obtain targeted and optimal interpolation kernels for non-regularly sampled data. The calibration data are self-calibration data obtained from a time-averaged image, or re-sampled data. ACS data are resampled for calibration of region-specific kernels. Subsequently, an explicit noise-based regularized solution can be utilized to estimate region-specific kernels for reconstruction.