A method may include obtaining a plurality of imaging signals collected by applying a wave encoding gradient to a region of interest (ROI) of a subject. The method may also include obtaining a plurality of auxiliary signals associated with the ROI. The method may also include obtaining a point spread function corresponding to the wave encoding gradient. The method may also include determining, based on the plurality of auxiliary signals, temporal information relating to at least one temporal dimension of the ROI. The method may also include determining, based on the plurality of auxiliary signals, the plurality of imaging signals, and the point spread function, spatial information relating to at least one spatial dimension of the ROI. The method may also include generating at least one target image of the ROI based on the temporal information and the spatial information.

A method of medical imaging including receiving k-space data that is divided into multiple k-space data groups, selecting one of the multiple k-space data groups as a reference k-space data group, and calculating spatial transform data for each of the multiple k-space data groups by inputting the multiple k-space data groups and the reference k-space data group into a transformation estimation module. The spatial transformation estimation module is configured for outputting spatial transform data descriptive of a spatial transform between a reference k-space data group and multiple k-space data groups in response to receiving the reference k-space data group and the multiple k-space data groups as input. The method further comprises reconstructing a corrected magnetic resonance image according to the magnetic resonance imaging protocol using the multiple k-space data groups and the spatial transform data for each of the multiple k-space data groups.

A method for producing an image of an object with a MRI system includes providing a Shear In Readout Encoding Imaging (SIREN) gradient pulse in a phase gradient signal waveform. The phase gradient signal waveform is applied to a phase gradient coil of the MRI system. The application of the SIREN gradient pulse provides a SIREN k-space of the object which has SIREN k-space lines with a shear angle. A MR image space data from the SIREN k-space is then obtained by applying a reconstruction technique. Finally, the image of the object is generated by transforming SIREN MR image space data into regular image space data using a decoding algorithm based on the shear angle.

Systems and methods for generative adversarial networks (GANs) to remove artifacts from undersampled magnetic resonance (MR) images are described. The process of training the GAN can include providing undersampled 3D MR images to the generator model, providing the generated example and a real example to the discriminator model, applying adversarial loss, L2 loss, and structural similarity index measure loss to the generator model based on a classification output by the discriminator model, and repeating until the generator model has been trained to remove the artifacts from the undersampled 3D MR images. At runtime, the trained generator model of the GAN can be generate artifact-free images or parameter maps from undersampled MRI data of a patient.

Methods, computing devices, and magnetic resonance imaging systems that improve image quality in turbo spiral echo (TSE) imaging are disclosed. With this technology, a TSE pulse sequence is generated that includes a series of radio frequency (RF) refocusing pulses to produce a corresponding series of nuclear magnetic resonance (NMR) spin echo signals. A gradient waveform including a plurality of segments is generated. The plurality of segments collectively comprise a spiral ring retraced in-out trajectory. During an interval adjacent to each of the series of RF refocusing pulses, a first gradient pulse is generated according to the gradient waveform. The first gradient pulses encode the NMR spin echo signals. An image is then constructed from digitized samples of the NMR spin echo signals obtained based at least in part on the encoding.

Training a neural network to correct motion-induced artifacts in magnetic resonance images includes acquiring motion-free magnetic resonance image (MRI) data of a target object and applying a spatial transformation matrix to the motion-free MRI data. Multiple frames of MRI data are produced having respective motion states. A Non-uniform Fast Fourier Transform (NUFFT) can be applied to generate respective k-space data sets corresponding to each of the multiple frames of MRI; the respective k-space data sets can be combined to produce a motion-corrupted k-space data set and an adjoint NUFFT can be applied to the motion-corrupted k-space data set. Updated frames of motion-corrupted MRI data can be formed. Using the updated frames of motion corrupted MRI data, a neural network can be trained that generates output frames of motion free MRI data; and the neural network can be saved.

According to one embodiment, a medical information processing apparatus has processing circuitry. The processing circuitry acquires medical data on a subject, acquires numerical data obtained by digitizing an acquisition condition of the medical data, and applies a machine learning model to input data including the numerical data and the medical data, thereby generating output data based on the medical data.

Calibration data is generated for completing undersampled measurement data acquired via a magnetic resonance system. This includes recording N measurement data sets using an acquisition scheme, and undersampling the k-space with an acceleration factor R, with N being greater than or equal to R, and the N measurement data sets together scanning the k-space completely. Phase images are generated from the N recorded measurement data sets, at least one homogeneity value of the created phase images is determined, and a complete calibration data set is generated based upon the recorded measurement data sets, taking into account the at least one homogeneity value. Thus, it is possible to determine which measurement data sets are subject to undesired phase errors, the measurement data sets used for the creation of the calibration data sets can be selected optimally, and input of the detected phase errors into the calibration data sets can be avoided.

Super-resolution processing is performed on an MRI image by using an NMR signal as a point spread function (PSF) . Image processing of increasing a resolution is performed on a reconstructed image by using the point spread function. The point spread function is a signal obtained by, after a phantom disposed in an imaging space is irradiated with a high-frequency magnetic field, acquiring a nuclear magnetic resonance signal from the phantom without applying frequency encoding and phase encoding, and performing Fourier transform on the acquired nuclear magnetic resonance signal.

In a method for avoiding artifacts during acquisition of MR data, a first measurement data set (MDS) of a target region of the examination object and at least one second MDS of the target region are acquired, and a combined MDS is created based on the acquired data sets. The first MDS does not sample a first region of k-space to be sampled according to Nyquist and corresponding to a first partial factor, and a second MDS does not sample a second region of k-space to be sampled according to Nyquist and corresponding to a second partial factor. The first and second regions of the k-space are different from each other. Advantageously, a k-space region acquired in none of the acquisitions made can be minimized by the inventive variation in the respective sampling pattern of the acquired MDS, so artifacts are reduced/avoided in MR images reconstructed from the MDS.