US-12620151-B2 - Systems and methods for magnetic resonance imaging

US12620151B2US 12620151 B2US12620151 B2US 12620151B2US-12620151-B2

Abstract

The present disclosure provides a system and method for magnetic resonance imaging. The method may include obtaining a first set of imaging data, the first set of imaging data being sampled in multiple shots, each shot of the multiple shots corresponding to a plurality of echo times, the first set of imaging data including partially sampled data in a first k space; obtaining a second set of imaging data, the second set of imaging data including fully sampled data in a central region of a second k space; determining fitting data in the first k space based on the first set of imaging data and the second set of imaging data; and/or generating a target image based on the fitting data in the first k space and the first set of imaging data in the first k space.

Inventors

Boyu JIANG
Jianmin Yuan
Wending TANG

Assignees

SHANGHAI UNITED IMAGING HEALTHCARE CO., LTD.

Dates

Publication Date: 20260505
Application Date: 20220412
Priority Date: 20210518

Claims (19)

1 . A method for magnetic resonance imaging, which is implemented on a computing device including at least one processor and at least one storage device, comprising: obtaining a first set of imaging data, the first set of imaging data being sampled in multiple shots, each shot of the multiple shots corresponding to a plurality of echo times, the first set of imaging data including partially sampled data in a first k space, wherein the first set of imaging data is continuously distributed along an echo time direction, and phase-encoding gradients corresponding to data points of the first set of imaging data filled at different echo times are different; obtaining a second set of imaging data, wherein the second set of imaging data includes fully sampled data in a central region of a second k space, and the second set of imaging data is continuously distributed along the echo time direction; determining fitting data in the first k space based on the first set of imaging data and the second set of imaging data; and generating a plurality of echo images based on the fitting data in the first k space and the first set of imaging data in the first k space, wherein each echo image of the plurality of echo images is reconstructed by imaging data acquired at a same echo time in different shots.
2 . The method of claim 1 , wherein the first k space is a first k y -t space, and/or the second k space is a second k y -t space, k y representing a phase-encoding direction, t representing the echo time direction.
3 . The method of claim 1 , wherein the first set of imaging data is generated using multi-shot echo planar imaging, and the first set of imaging data is generated by: obtaining a first sub-set of imaging data in each shot of the multiple shots; and filling the multiple first sub-sets of imaging data in the first k space to obtain the first set of imaging data.
4 . The method of claim 3 , wherein the first sub-set of imaging data is filled in a cyclically changing trajectory in the first k space.
5 . The method of claim 1 , wherein the first set of imaging data is generated by sampling at the different echo times and filling in different portions of the first k space; the second set of imaging data is generated by sampling at the different echo times and filling in the central region of the second k space, and the second set of imaging data only includes imaging data in the central region of the second k space.
6 . The method of claim 1 , wherein the second set of imaging data is acquired with no phase-encoding gradient.
7 . The method of claim 1 , wherein the determining the fitting data in the first k space based on the first set of imaging data and the second set of imaging data comprises: determining weight values associated with missing data in the first k space, based on at least one portion of the second set of imaging data; and determining the fitting data corresponding to the missing data in the first k space based on the weight values and the first set of imaging data.
8 . The method of claim 7 , wherein the weight values are characterized by determining a correspondence between the missing data in the first k space and the first set of imaging data.
9 . The method of claim 7 , wherein the determining the weight values associated with the missing data in the first k space comprises: determining the weight values based on a machine learning model.
10 . The method of claim 1 , wherein the generating the plurality of echo images comprises: generating at least two echo images based on the fitting data in the first k space and the first set of imaging data in the first k space; and generating the plurality of echo images based on the at least two echo images.
11 . The method of claim 1 , wherein the second set of imaging data is acquired with slice selection gradient, readout gradient, and no phase-encoding gradient.
12 . The method of claim 1 , wherein a count of shots for the second set of imaging data is larger than a count of shots for the first set of imaging data.
13 . The method of claim 1 , wherein a count of shots for the second set of imaging data is determined based on a coverage of imaging data in the first k space that is collected in each shot of the multiple shots, or the count of shots for the second set of imaging data is dynamically adjusted, wherein for different scanning parts or objects, the count of shots for the second set of imaging data is different.
14 . The method of claim 1 , further comprising: obtaining at least two temporally sequential echo images, wherein each of the echo images of the at least two temporally sequential echo images is generated by reconstructing imaging data of a mapped k space corresponding to a second same echo time through multiple excitation planar echo data acquisitions; determining a fat-water transition region and a phase distribution of the each of the echo images of the at least two temporally sequential echo images; and determining a quantitative distribution of water and fat based on the fat-water transition region and the phase distribution of the each of the echo images of the at least two temporally sequential echo images to obtain a fat quantitative map.
15 . The method of claim 14 , wherein the determining the fat-water transition region and the phase distribution of the each of the echo images of the at least two temporally sequential echo images includes: for all adjacent pixels in the each of the echo images of the at least two temporally sequential echo images, determining whether the adjacent pixels are in the fat-water transition region and marking the fat-water transition region; dividing the each of the echo images of the at least two temporally sequential echo images into multiple sub-regions based on the marked fat-water transition region, and determining a fat and water distribution in each sub-region to determine the fat-water transition region and the phase distribution of the each of the echo images of the at least two temporally sequential echo images.
16 . A method for magnetic resonance imaging, which is implemented on a computing device including at least one processor and at least one storage device, comprising: generating a first set of imaging data using multi-shot echo planar imaging (EPI), the first set of imaging data including a plurality of first sub-sets of imaging data, each first sub-set of imaging data corresponding to an EPI shot of a plurality of EPI shots, wherein the first set of imaging data is continuously distributed along an echo time direction, and phase-encoding gradients corresponding to data points of the first set of imaging data filled at different echo times are different; filling the first set of imaging data in a first k y -t space by filling the each first sub-set of imaging data in a cyclically changing trajectory in the first k y -t space, to obtain partially sampled data in the first k y -t space, k y representing a phase-encoding direction, t representing the echo time direction; generating a second set of imaging data using multi-shot EPI with no phase-encoding gradient, the second set of imaging data is continuously distributed along the echo time direction; filling the second set of imaging data in a second k y -t space to obtain fully sampled data in a central region of the second k y -t space; and generating a plurality of echo images corresponding to the first k y -t space based on the first set of imaging data, wherein each echo image of the plurality of echo images is reconstructed by imaging data acquired at a same echo time in different shots.
17 . The method of claim 16 , wherein each EPI shot of the plurality of EPI shots uses cyclically changing phase-encoding gradient blips; the phase-encoding gradient blips in each EPI shot are preceded by application of a pre-winding gradient; the pre-winding gradients for the plurality of EPI shots have different gradient moments; and the gradient moments of the pre-winding gradients for the plurality of EPI shots are randomly generated.
18 . The method of claim 16 , wherein the generating the plurality of echo images comprising: determining fitting data in the first k y -t space based on the first set of imaging data and the second set of imaging data; and generating the plurality of echo images based on the fitting data in the first k y -t space and the first set of imaging data in the first k y -t space.
19 . A system for magnetic resonance imaging, comprising: at least one storage device storing a set of instructions; and at least one processor in communication with the storage device, wherein when executing the set of instructions, the at least one processor is configured to cause the system to perform operations including: obtaining a first set of imaging data, the first set of imaging data being sampled in multiple shots, each shot of the multiple shots corresponding to a plurality of echo times, the first set of imaging data including partially sampled data in a first k space, wherein the first set of imaging data is continuously distributed along an echo time direction, and phase-encoding gradients corresponding to data points of the first set of imaging data filled at different echo times are different; obtaining a second set of imaging data, wherein the second set of imaging data includes fully sampled data in a central region of a second k space, and the second set of imaging data is continuously distributed along the echo time direction; determining fitting data in the first k space based on the first set of imaging data and the second set of imaging data; and generating a plurality of echo images based on the fitting data in the first k space and the first set of imaging data in the first k space, wherein each echo image of the plurality of echo images is reconstructed by imaging data acquired at a same echo time in different shots.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS This application claims priority of Chinese Patent Application No. 202110539454.6, filed on May 18, 2021, and the contents of which are hereby incorporated by reference. TECHNICAL FIELD This disclosure generally relates to systems and methods for imaging, and more particularly, relates to systems and methods for magnetic resonance imaging. BACKGROUND Magnetic resonance imaging (MRI) is widely used for generating images of the interior of a patient for medical diagnosis and/or treatment purposes. During an MRI process, a plurality of acquired radiofrequency (RF) signals may be filled into k space. The data in k space may be transformed to reconstruct an MRI image. MRI images generally have a good soft tissue contrast. However, the required image acquisition time for MRI is relatively long, and motion artifacts can be easily introduced into MRI images of motion sensitive organs or tissues (e.g., the heart, a lung, an abdomen, etc.). Thus, it is desired to provide systems and methods for improving imaging speed and imaging quality of magnetic resonance imaging. SUMMARY In one aspect of the present disclosure, a method for magnetic resonance imaging is provided. The method may include one or more operations. The one or more operations may be implemented on a computing device having one or more processors and one or more storage devices. The one or more operations may include: obtaining a first set of imaging data, the first set of imaging data being sampled in multiple shots, each shot of the multiple shots corresponding to a plurality of echo times, the first set of imaging data including partially sampled data in a first k space; obtaining a second set of imaging data, the second set of imaging data including fully sampled data in a central region of a second k space; determining fitting data in the first k space based on the first set of imaging data and the second set of imaging data; and generating a target image based on the fitting data in the first k space and the first set of imaging data in the first k space. In some embodiments, the first k space may be a first ky-t space, and/or the second k space may be a second ky-t space, ky representing a phase-encoding direction, t representing an echo time direction. In some embodiments, the first set of imaging data may be generated using multi-shot echo planar imaging. In some embodiments, the first set of imaging data may be generated by: obtaining a first sub-set of imaging data in each shot of the multiple shots; and filling the multiple first sub-sets of imaging data in the first k space to obtain the first set of imaging data. In some embodiments, the first sub-set of imaging data may be filled in a cyclically changing trajectory in the first k space. In some embodiments, the first set of imaging data may be generated by sampling at different echo times and filling in different portions of the first k space. In some embodiments, the second set of imaging data may be generated by sampling at different echo times and filling in the central region of the second k space. In some embodiments, the second set of imaging data may be acquired with no phase-encoding gradient. In some embodiments, the determining fitting data in the first k space based on the first set of imaging data and the second set of imaging data may include: determining weight values associated with missing data in the first k space, based on at least one portion of the second set of imaging data; and determining the fitting data corresponding to the missing data in the first k space based on the weight values and the first set of imaging data. In some embodiments, the weight values may be characterized by determining a correspondence between the missing data in the first k space and the first set of imaging data. In some embodiments, the determining weight values associated with missing data in the first k space may include: determining the weight values based on the at least one portion of the second set of imaging data using at least one of interpolation, a low rank method, or a machine learning model. In some embodiments, the generating the target image may include: generating at least two echo images based on the fitting data in the first k space and the first set of imaging data in the first k space; and generating the target image based on the at least two echo images. In another aspect of the present disclosure, a method for magnetic resonance imaging is provided. The method may include one or more operations, including: generating a first set of imaging data using multi-shot echo planar imaging (EPI), the first set of imaging data including a plurality of first sub-sets of imaging data, each first sub-set of imaging data corresponding to an EPI shot of a plurality of EPI shots; and filling the first set of imaging data in a first ky-t space by filling the each first sub-set of imaging data in a cyclically changing trajectory in the first ky-t