The subject matter discussed herein relates to a fast magnetic resonance imaging (MRI) method to suppress fine-line artifact in Fast-Spin-Echo (FSE) images reconstructed with a deep-learning network. The network is trained using fully sampled NEX=2 (Number of Excitations equals to 2) data. In each case, the two excitations are combined to generate fully sampled ground-truth images with no fine-line artifact, which are used for comparison with the network generated image in the loss function. However, only one of the excitations is retrospectively undersampled and inputted into the network during training. In this way, the network learns to remove both undersampling and fine-line artifacts. At inferencing, only NEX=1 undersampled data are acquired and reconstructed.

Techniques are disclosed for ascertaining a point spread function (PSF) for reconstructing image data from scan data recorded by means of a magnetic resonance system. The techniques include a comparison of values determined for a planned k-space trajectory for parameters characterizing the k-space trajectory with baseline values of the parameters characterizing the k-space trajectory deposited in a database for the magnetic resonance system, in each case together with an associated point spread function PSF to ascertain baseline values of the deposited baseline values that are as similar as possible to the values determined for the planned k-space trajectory for the parameters characterizing the k-space trajectory and, on the basis of this deposited PSF, a PSF to be used for a reconstruction of final image data is ascertained.

In one aspect the disclosed technology relates to embodiments of a method (e.g., for automatic cine DENSE strain analysis) which includes acquiring magnetic resonance data associated with a physiological activity in an area of interest of a subject where the acquired magnetic resonance data includes one or more phase-encoded data sets. The method also includes determining, from at least the one or more phase-encoded data sets, a data set corresponding to the physiological activity in the area of interest where the reconstruction comprises performing phase unwrapping of the phase-encoded data set using region growing along multiple pathways based on phase predictions.

The invention relates to a method of Dixon-type MR imaging. It is an object of the invention to provide an MR imaging technique using bipolar readout magnetic field gradients with an improved estimation of the main field inhomogeneity to eliminate residual artifacts. In accordance with the invention, a method of MR imaging of an object placed in a main magnetic field within an examination volume of a MR device is proposed, wherein the method comprises the steps of: —subjecting the object (10) to an imaging sequence to generate at least two sets of echo signals at two or more different echo times using bipolar pairs of readout magnetic field gradients, one set of echo signals being generated at a first echo time (TE1) and the other set of echo signals being generated at a second echo time (TE2), —acquiring the echo signals from the object (10), —reconstructing a first image from the echo signals attributed to the first echo time (TE1) and a second image from the echo signals attributed to the second echo time (TE2), —computing modified first and second images by compensating for a fat shift in the reconstructed first and second images respectively, —estimating phase errors in the acquired echo signals on the basis of the first and second images and the modified first and second images using a signal model including the resonance spectra of fat and water and the spatial variation of the main magnetic field, and —reconstructing a water image and/or a fat image by separating the signal contributions of fat and water to the acquired echo signals using the estimated phase errors. Moreover, the invention relates to a MR device (1) and to a computer program to be run on a MR device (1).

In a method for correcting influences on magnetic resonance imaging of an examination object caused by fluctuations in a basic magnetic field, an MR data set is generated for two or more measurement periods, and a regression analysis is performed. Each of the MR data sets may contain at least one two-dimensional individual data set. The regression analysis may determine at least one phase correction value for a measurement period to be corrected. Two or more different individual data sets may be taken into account in the analysis. An MR image may generated based on the MR data sets and the at least one phase correction value.

Disclosed are a magnetic-resonance imaging method, apparatus and system, and a storage medium. The method includes acquiring an initial model of magnetic-resonance imaging and establishing an initial imaging model according to an iterative algorithm used for solving the initial model, where the iterative algorithm includes at least one of an undetermined parameter, an undetermined solving operator or an undetermined structural relationship; training the initial imaging model on the basis of sample data to generate a magnetic-resonance imaging model, where training of the initial imaging model is used for learning the at least one of the undetermined parameter, the undetermined solving operator or the undetermined structural relationship in the iterative algorithm; and acquiring under-sampled K-space data to be processed, inputting the under-sampled K-space data into the magnetic-resonance imaging model, and generating a magnetic-resonance image.

A method for determining an undersampling error for a magnetic resonance fingerprinting (MRF) pulse sequence includes retrieving a plurality of sets of spatial response functions. Each set of spatial response functions is associated with a tissue type in a reference image and is based on a tissue mask of the reference image for each tissue type. A signal evolution for each tissue type may be generated based on, for example, the MRF pulse sequence, An undersampled image may be generated for each tissue type using the set of spatial response functions and the signal evolutions for the tissue type. At least one quantitative parameter may be determined by comparing an undersampled image series created from the undersampled images to an MRF dictionary. An undersampling error for the MRF pulse sequence may be generated by comparing a quantitative map (or maps) for the quantitative parameter (or parameters) and the reference image.

A system and method for correcting phase shift in echo images are provided. The method may include one or more of the following operations. A plurality of echo images may be obtained. Homogeneous pixels in the plurality of echo images may be identified. A vector corresponding to each of at least some of the identified homogeneous pixels may be determined. A vector of a homogenous pixel includes a phase element and an amplitude element. A first complex linear model of phase shift may be determined based at least in part on the determined vectors. Phase shift of at least one of the plurality of echo images may be corrected based on the first complex linear model.

In a method for operating an MRI device, image data is acquired using a spin echo sequence with an additional readout per pulse train for acquiring correction data. By comparing subsequent correction data of later pulse trains to reference data acquired during a first pulse train of the sequence a difference indicating a parameter shift is determined. A corresponding compensation is then automatically determined in dependence on the difference and is applied to a set of predetermined parameters for at least a respective next pulse train and/or to the image data acquired in at least a respective next pulse train of the sequence.

The present invention is directed to enabling high-accuracy Nyquist ghost correction without using a reference image. After at least one of a plurality of images for use in diagnosis is used to perform low-order phase correction without causing aliasing of an image, a 2D phase map including remaining high-order phase errors and phase errors in a phase encode direction is calculated. The low-order phase correction is performed on a pair of pieces of data for image obtained by inverting a readout gradient magnetic field as image data for use in 2D phase map calculation, and positive-polarity/negative-polarity errors of the readout gradient magnetic field are calculated with odd lines and even lines of the pair of pieces of data for image rearranged. In the case of DWI imaging, an image with b-value=0 can be used for 2D phase map calculation.