A system and method for modifying RF pulse generated by an MRI system are provided. The method may include: obtaining an excitation variable-rate selective excitation (VERSE) factor and a refocusing VERSE factor; determining a first slice-selection gradient waveform based on an excitation factor and a reference excitation slice-selection gradient waveform; determining a second slice-selection gradient waveform based on a refocusing factor and a reference refocusing slice-selection gradient waveformslice-selection gradient waveformslice-selection gradient waveform; determining an excitation pulse based on the first slice-selection gradient waveform; determining a refocusing pulse based on the second slice-selection gradient waveform, wherein a ratio of the decimal part of the excitation factor to the decimal part of the refocusing factor is equal to a ratio of the amplitude of the first reference waveform to the amplitude of the reference refocusing slice-selection gradient waveform.

In a method and apparatus for generating corrected magnetic resonance measurement data in an examination region of an object undergoing examination, a magnetic resonance sequence is applied to a subject in order to acquire magnetic resonance measurement data from an examination region within a time period. A first resonant frequency of nuclear spins in the examination region is determined at a first time point within the time period. A second resonant frequency of nuclear spins in the examination region is determined at a second time point within the time period. Magnetic resonance measurement data that are acquired at a further time point within the time period are corrected, based on the first resonant frequency and the second resonant frequency.

A method for MRI imaging of a subject includes spatially encoding spins in a slice of the subject in orthogonal first and second directions. The encoding includes applying a chirped radiofrequency (RF) pulse concurrently with application of a magnetic field gradient pulse along the first direction. After applying of the RF pulse, a second chirped RF pulse is applied concurrently with application of a second magnetic field gradient pulse, with polarity opposite that of the first gradient pulse. An encoding magnetic field gradient, constant from applying the first RF pulse until the end of applying the second RF pulse, is concurrently applied along the second direction. Following the encoding, a spin signal is measured concurrently with application of a constant readout magnetic field gradient.

In a method for correcting chemical shift artifacts, CSA, a convolutional neural network (CNN) may be provided, which is trained using acquisition data acquired in phase and in opposed phase by a DIXON MR method that may include acquisition data that contains CSA in mutually opposite directions and acquisition data that contains CSA only in one direction. The CNN may be trained to transform acquisition data obtained by the fast DIXON MR method so the acquisition data acquired by the fast DIXON MR method exhibits CSA that arise only in the same direction. The method may further include acquiring fast DIXON MR acquisition data using respective control instructions and applying the trained CNN to the acquisition data to minimize or entirely remove the CSA and to calculate corrected acquisition data. The CSA may arise in the magnetic resonance DIXON method when using fast DIXON MR to capture the echoes.

For determination of motion artifact in MR imaging, motion of the patient in three dimensions is used with a measurement k-space line order based on one or more actual imaging sequences to generate training data. The MR scan of the ground truth three-dimensional (3D) representation subjected to 3D motion is simulated using the realistic line order. The difference between the resulting reconstructed 3D representation and the ground truth 3D representation is used in machine-based deep learning to train a network to predict motion artifact or level given an input 3D representation from a scan of a patient. The architecture of the network may be defined to deal with anisotropic data from the MR scan.

The present invention pertains to magnetic resonance imaging, notably of separate body parts with open space between them. A bridge member containing MR responsive material is provided in the open space to establish a correspondence between the body parts. The MR responsive material generates magnetic resonance signals in response the RF excitation, so that between the separate body parts via the bridge member magnetic resonance signal are obtained from positions between which there is at most a limited spatial variation of the main magnetic field, so that phase ambiguities between the signals from these positions are avoided. Thus chemical shift separation, notably water-fat separation though a region-of-interest containing several (both) body parts may rely on a smoothness condition imposed on the spatial distribution of the main magnetic field. This avoids artefacts, such as water-fat swaps when separating water and fat contributions in the reconstructed magnetic resonance image.

Disclosed herein is a medical system (100, 300) comprising a memory (110) storing machine executable instructions (120). The medical system further comprises a computational system (104). Execution of the machine executable instructions causes the computational system to: receive (200) initial pulse sequence commands (122), wherein the initial pulse sequence commands are configured for controlling a magnetic resonance imaging system (302) to acquire k-space data (332) following a non-Cartesian k-space sampling pattern (604, 604), wherein the initial pulse sequence commands are configured for controlling the magnetic resonance imaging system to sample the non-Cartesian k-space sampling pattern by repeatedly sampling a Cartesian k-space sampling pattern (126) that is rotated for each acquisition, wherein the non-Cartesian k-space sampling pattern has an effective water-fat shift direction (606, 606); receive (202) a chosen water-fat shift direction (124); and construct (204) modified pulse sequence commands by rotating the non-Cartesian k-space sampling pattern such that the effective water-fat shift direction is aligned with the water-fat shift direction.

Capturing MR image data of an examination object using an MR apparatus, including: performing a balanced steady-state free precession sequence with phase progress of 180 degrees per repetition time using the MR apparatus; in the balanced steady-state free precession sequence, providing a white-marker gradient in order at least partially to balance a dephasing caused by a magnetic-field-changing object in the examination object; capturing image data of the examination object using the MR apparatus at an echo time; and adjusting a phase development between phase magnetization of a first and second materials, which form an interface in the examination object, in the balanced steady-state free precession sequence using the MR apparatus, wherein due to the adjusting of the phase development before an effect of the white-marker gradient, a co-phasal alignment of a magnetization of the first material and of the second material at the interface is effected at the echo time.

Disclosed herein is a medical system (100, 300, 500) comprising a memory (110) storing machine executable instructions (120) and a convolutional neural network (122). The convolutional neural network is configured for receiving an initial Dixon magnetic resonance image (124, 126) as input. The convolutional neural network is configured for identifying one or more water-fat swap regions (128) in the initial Dixon magnetic resonance image. The medical system further comprises a processor (104) for controlling the medical system. Execution of the machine executable instructions causes the processor to: receive (200) the initial Dixon magnetic resonance image; and receive (204) the one or more water-fat swap regions from the convolutional neural network in response to inputting the initial Dixon magnetic resonance image into the convolutional neural network.

Improved magnetic resonance imaging systems, methods and software are described including a low field strength main magnet, a gradient coil assembly, an RF coil system, and a control system configured for the acquisition and processing of magnetic resonance imaging data from a patient while utilizing a sparse sampling imaging technique.