Magnetic Resonance Imaging Method and Magnetic Resonance Imaging System

Abstract

The present disclosure relates to a magnetic resonance imaging method and a magnetic resonance imaging system. The magnetic resonance imaging method includes the following steps: fully sampling and occupying the central region of a first k-space along a Cartesian trajectory, and undersampling and occupying the peripheral region of the first k-space along a non-Cartesian trajectory; obtaining sensitivity distribution information of a receiving coil; merging Cartesian data of the central region into a third k-space according to multiple channels based on the sensitivity distribution map; applying parallel imaging and compressed sensing to the undersampled non-Cartesian trajectory based on the sensitivity distribution map to reconstruct an image, and obtaining a second k-space through transformation. When synthesizing the second k-space and the third k-space, replacing the corresponding region of the third k-space with the central region of the second k-space to obtain a k-space suitable for image reconstruction. It realizes the reconstruction of the k-space of a hybrid MR trajectory including an undersampled part by using sensitivity distribution information and compressed sensing to improve the k-space and obtain an image suitable for medical diagnosis.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging, and particularly to a method for collecting and recording magnetic resonance signal data based on a hybrid trajectory to perform magnetic resonance image reconstruction by combining parallel imaging and compressed sensing reconstruction. Background Art

[0002] Magnetic Resonance Imaging (MRI) is a medical imaging technique that irradiates an object with radio frequency pulse signals using an antenna under certain magnetic field conditions and forms an image based on the received modulated radio frequency signals from the object. The internal structure, substance composition, physiological processes, etc. of the object can be studied using magnetic resonance imaging technology. A radio frequency pulse with a Larmor frequency causes the spin nucleons, such as hydrogen nuclei (i.e., H+), in the irradiated object to precess with a deflection angle, generating a magnetic resonance radio frequency signal after excitation, which is received by a receiving coil and processed by a computer to form an image.

[0003] Existing magnetic resonance scanning protocols based on pulse sequences with ultra-short echo time (TE) have the advantage of extremely low noise and can greatly improve the comfort of patients during magnetic resonance scanning. Therefore, the above magnetic resonance scanning protocols are naturally widely used in pediatric magnetic resonance imaging, dental imaging, lung imaging, and magnetic resonance scanning and examination of conventional organs. However, in the above magnetic resonance scanning protocol, for example, the k-space of the PETRA sequence has the magnetic resonance signal data collected along a non-Cartesian trajectory, such as a 3D radial trajectory, occupying the outer peripheral region of its k-space through the control of the gradient magnetic field, and occupying and recording along Cartesian points in the central region of the k-space. To obtain a high-resolution image based on the above magnetic resonance scanning protocol, the Nyquist law needs to be satisfied, so a large number of radial trajectories are collected, thus consuming a large amount of time. Therefore, there is a need to provide a magnetic resonance scanning technology that can meet the image quality required for diagnosis and reduce the acquisition duration. Summary of the Invention

[0004] In view of this, on the one hand, the present disclosure provides a magnetic resonance imaging method that is at least based on compressive sensing reconstruction and, for the case of multiple receive coils, also combines parallel imaging to address issues such as magnetic resonance image reconstruction of k - space involving hybrid trajectory acquisition and recording of undersampled portions and obtain images suitable for medical diagnosis, where the undersampled portion includes the k - space of non - Cartesian trajectories. The method includes the following steps: The magnetic resonance signal data collected by at least one receive coil fully samples and occupies the central region of the first k - space along Cartesian trajectories under the control of a gradient magnetic field, and the collected magnetic resonance signal data undersamples and occupies the peripheral region of the first k - space along non - Cartesian trajectories under the control of a gradient magnetic field; Reconstruct the first image data based on the data occupying the non - Cartesian trajectories in the peripheral region of the first k - space to construct a second k - space, where the reconstruction of the first image data at least includes suppressing artifacts caused by the undersampling in the first image data during image reconstruction using the sparse representation of the first image data in the transform domain, and where, in the second k - space obtained by transforming the reconstructed first image data, the data occupies Cartesian trajectories; Synthesize the data occupying Cartesian trajectories in the central region of the first k - space with the second k - space to generate a k - space suitable for magnetic resonance imaging.

[0005] Optionally, in the magnetic resonance imaging method, the synthesizing the data occupying Cartesian trajectories in the central region of the first k - space with the second k - space to generate a k - space suitable for magnetic resonance imaging includes: Replacing the data occupying the corresponding central region of the Cartesian trajectory in the second k - space with the data occupying the Cartesian trajectory in the central region of the first k - space extracted to generate a k - space suitable for magnetic resonance imaging.

[0006] Optionally, the magnetic resonance imaging method further includes: Obtaining sensitivity distribution information reflecting multiple receive coils.

[0007] Optionally, in the magnetic resonance imaging method, the reconstructing the first image data based on the data occupying the non - Cartesian trajectories in the peripheral region of the first k - space to construct a second k - space includes: The reconstruction of the first image data further includes performing parallel imaging with the aid of the sensitivity distribution information.

[0008] Optionally, in the magnetic resonance imaging method, reconstructing first image data based on data occupying the non-Cartesian trajectory in the peripheral region of the first k-space to construct a second k-space includes: performing parallel imaging on data of a plurality of the first k-space peripheries occupying the non-Cartesian trajectory to obtain first image data; using the first image data to sparsely represent in a transform domain and suppressing artifacts caused by the undersampling in the first image during image reconstruction to obtain reconstructed first image data; and applying a Fourier transform based on the reconstructed first image data to obtain the second k-space.

[0009] Optionally, in the magnetic resonance imaging method, synthesizing data occupying the Cartesian trajectory in the central region of the first k-space and the second k-space to generate a k-space suitable for magnetic resonance imaging includes: applying sensitivity distribution information to data occupying the Cartesian trajectory in the central region of the first k-space corresponding to a plurality of the receive coils to construct a third k-space.

[0010] Optionally, in the magnetic resonance imaging method, applying sensitivity distribution information to data occupying the Cartesian trajectory in the central region of the first k-space corresponding to a plurality of the receive coils to construct a third k-space includes: transforming data occupying the Cartesian trajectory in the central region of the first k-space corresponding to a plurality of the receive coils to the conjugate product of the image domain and the sensitivity distribution information, and merging according to the measurement channels corresponding to the receive coils to obtain second image data to construct the third k-space.

[0011] Optionally, in the magnetic resonance imaging method, applying sensitivity distribution information to data occupying the Cartesian trajectory in the central region of the first k-space to construct a third k-space includes: extracting data of the occupied Cartesian trajectory in the central region from the first k-space and applying zero filling in regions outside the central region with reference to the sensitivity distribution information to obtain a plurality of Cartesian point sets having the same matrix size as the sensitivity distribution information, wherein the plurality of Cartesian point sets correspond to the acquisition of the receive coils; applying an inverse Fourier transform to the plurality of Cartesian point sets to obtain a plurality of third image data, multiplying the conjugate of the sensitivity distribution information with the third image data respectively; and merging according to the measurement channels corresponding to the receive coils to obtain second image data; and transforming based on the second image data to the third k-space.

[0012] Optionally, in the magnetic resonance imaging method, the synthesis of the data of the central region of the first k-space occupying a Cartesian trajectory and the second k-space to generate a k-space suitable for magnetic resonance imaging further includes: the synthesis of the second k-space and the third k-space to generate a k-space suitable for image reconstruction, wherein the data of the Cartesian trajectory occupied by the central region of the second k-space is replaced by the data of the Cartesian trajectory occupied by the corresponding region in the third k-space, and wherein the size of the central region of the second k-space is the same as that of the central region of the first k-space.

[0013] Optionally, the reconstructed first image data in the magnetic resonance imaging method is reconstructed by means of compressed sensing.

[0014] Optionally, the non-Cartesian trajectory in the magnetic resonance imaging method is a 3D radial trajectory.

[0015] Optionally, the acquisition of the sensitivity distribution information reflecting the receiving coil in the magnetic resonance imaging method includes: applying gridding processing to the first k-space; removing the high-frequency part from the gridded first k-space to extract and generate a low-resolution image; and calculating the sensitivity distribution information based on the low-resolution image.

[0016] Optionally, in the magnetic resonance imaging method, ESPIRiT is applied to the low-resolution image to obtain the sensitivity distribution information by calculating eigenvectors from the null space.

[0017] Another aspect of the present disclosure provides a magnetic resonance imaging system, including: at least one receiving coil; a magnetic resonance controller; and a memory storing a program, the program including instructions, the magnetic resonance controller being connected to the memory and configured to execute the instructions, wherein the instructions, when executed by the magnetic resonance controller, cause the magnetic resonance imaging system to execute the magnetic resonance imaging method according to the foregoing.

[0018] Another aspect of the present disclosure provides a computing device configured to be used in a magnetic resonance imaging method using a k-space trajectory in a magnetic resonance imaging system including at least one receiving coil, the k-space trajectory at least including a central region of a k-space fully sampled along a Cartesian trajectory by magnetic resonance signal data collected by the receiving coil and a peripheral region of the k-space undersampled along a non-Cartesian trajectory by the collected magnetic resonance signal data, the computing device including a memory and at least one processor, the memory containing instructions executed by the at least one processor, and executing the instructions causes the computing device to perform the magnetic resonance imaging method according to the foregoing.

[0019] Another aspect of the present disclosure provides a computer program product, including a program executed by at least one processor of a computing device, the program including instructions, wherein the execution of the instructions causes the at least one processor to execute the magnetic resonance imaging method according to the foregoing.

[0020] Another aspect of the present disclosure provides a computer-readable storage medium storing a program, the program including instructions, characterized in that when the instructions are executed by an electronic device, the electronic device is caused to execute the magnetic resonance imaging method according to the foregoing.

[0021] One advantage of the magnetic resonance imaging method provided by the present disclosure is that the k-space of the hybrid trajectory includes magnetic resonance signal data recorded by an undersampled non-Cartesian trajectory and magnetic resonance signal data acquired along a Cartesian trajectory. Considering applying at least compressed sensing reconstruction to the data part acquired by the undersampled non-Cartesian trajectory to reconstruct the image and transforming this part into the reconstructed k-space where data is recorded along the Cartesian trajectory, the central region of the k-space of the hybrid trajectory can be synthesized with this reconstructed k-space, and a resolution image suitable for medical diagnostic criteria can be obtained based on the synthesized k-space. This method utilizes the advantage of the short acquisition time of the MRI signal of the undersampled non-Cartesian trajectory.

[0022] Another advantage is that for the k-space of the hybrid trajectory acquired by multiple coils, considering applying parallel imaging and compressed sensing reconstruction in combination to the data part acquired by the undersampled non-Cartesian trajectory to obtain a resolution image suitable for medical diagnostic criteria, and parallel imaging can further improve the advantage of the short acquisition time of the MRI signal.

[0023] Another advantage is that for the k-space data acquired by the hybrid trajectory, parallel imaging and compressed sensing reconstruction cannot be directly applied. The present disclosure proposes to separately process the data (magnetic resonance signals) acquired by the Cartesian trajectory and the data acquired by the non-Cartesian trajectory in the k-space acquired by the hybrid trajectory to solve the above problem. And a technical solution is proposed to apply the coil sensitivity distribution information to the data of the Cartesian trajectory in the central region of the k-space for channel merging respectively and perform parallel imaging on the data of the non-Cartesian trajectory in the peripheral region.

[0024] Another advantage is that an improved third k-space is obtained in the k-space considering the hybrid trajectory by combining the fully sampled Cartesian trajectory part with the sensitivity distribution information about the receive coil, while the data acquired based on the non-Cartesian trajectory in the peripheral region is applied with Fourier transform after parallel imaging and compressed sensing reconstruction to obtain a second k-space, and the data of the second k-space is recorded in a Cartesian trajectory, enabling the second k-space and the third k-space to be suitable for synthesis processing. The corresponding region of the second k-space is replaced with the central region of the third k-space to obtain a k-space suitable for image reconstruction, further improving the quality and image resolution of the final MRI imaging based on the acquisition of magnetic resonance signals in a shorter duration.

[0025] Another advantage is that the magnetic resonance imaging method provided by the present disclosure can retain the advantage of ultra-low noise of the existing magnetic resonance scanning protocol based on the pulse sequence with an ultra-short echo time (TE), such as the PETRA sequence, etc., and can greatly improve the comfort of patients during the magnetic resonance scanning process. At the same time, the advantage of reducing the MRI signal acquisition duration by using the undersampled non-Cartesian trajectory in it enables the magnetic resonance scanning protocol based on the hybrid MRI signal acquisition trajectory to be more widely used. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings, making the above and other features and advantages of the present invention clearer to those of ordinary skill in the art. In the drawings:

[0027] Figure 1 A schematic diagram of a k-space obtained based on the execution of the PETRA pulse sequence is shown;

[0028] Figure 2 A schematic diagram of a magnetic resonance imaging system including a receive coil according to an exemplary embodiment is shown;

[0029] Figure 3 A flowchart schematically showing the steps of a magnetic resonance imaging method applied to an object to be examined in a magnetic resonance imaging system including at least one receive coil according to an embodiment of the present disclosure is shown;

[0030] Figure 4 A flowchart schematically showing the steps of a magnetic resonance imaging method applied to an object to be examined in a magnetic resonance imaging system including a plurality of receive coils according to an embodiment of the present disclosure is shown;

[0031] Figure 5 A flowchart schematically showing the steps of another magnetic resonance imaging method applied to an object to be examined in a magnetic resonance imaging system including a plurality of receive coils according to another embodiment of the present disclosure is shown;

[0032] Figure 6Shows a comparison schematic diagram of the image reconstruction of the magnetic resonance imaging method executed according to an embodiment of the present disclosure applied to a hybrid trajectory sequence and standard-based image reconstruction;

[0033] Figure 7 Shows a structural diagram of a computing device that can be applied to an exemplary embodiment.

[0034] Among them, the reference numerals are as follows:

[0035] 1 Magnetic resonance imaging system

[0036] 2 Examination area

[0037] 3 Receiver coil

[0038] 4 Bed

[0039] 5 First k-space

[0040] 5a Cartesian trajectory

[0041] 5b Radial trajectory

[0042] 5c Field of view

[0043] 5d Central region

[0044] 5e Peripheral region

[0045] 6 Gradient coil

[0046] 7 Radio frequency coil

[0047] 10 Magnet

[0048] 11 Magnetic resonance controller

[0049] 12 Gradient controller

[0050] 13 Radio frequency controller

[0051] 14 Recording unit

[0052] 15 Memory

[0053] 16 Computing unit

[0054] 17 Input unit

[0055] 18 Display

[0056] P Examination object Detailed implementation manners

[0057] For a clearer understanding of the technical features, objectives, and effects of the present disclosure, the detailed implementation manners of the present disclosure are now described with reference to the accompanying drawings, where the same reference numerals represent the same parts in each figure.

[0058] In this document, "schematic" means "serving as an instance, example, or illustration", and any illustration or implementation described as "schematic" in this document should not be construed as a more preferred or advantageous technical solution.

[0059] For the sake of simplicity of the drawings, only the parts related to the present invention are schematically shown in each figure, and they do not represent the actual structure of the product. In addition, for the sake of simplicity and easy understanding of the drawings, among the parts with the same structure or function in some figures, only one of them is schematically shown, or only one of them is labeled.

[0060] In this document, "a" not only means "only this one", but also can mean "more than one" situation. In this document, "first", "second", etc. are only used for distinguishing from each other, rather than indicating their importance degree, order, and the premise of mutual existence, etc. In addition, the term " / and" used in this disclosure covers any one of the listed items and all possible combination manners. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally indicates that the objects before and after are in an "or" relationship.

[0061] Pointwise Encoding Time Reduction With Radial Acquisition (PETRA) is a pulse sequence based on ultra-short echo time (Ultra-short TE), having a 3D (three-dimensional) radial acquisition trajectory and a Cartesian single-point feature at the center of k-space. The PETRA pulse sequence has the advantage of ultra-low noise and can improve the patient's comfort during the MRI image acquisition process, and is particularly widely used in the field of pediatric MRI. However, it takes a relatively long time (usually more than 6 minutes) to generate high-resolution pictures when executing the PETRA pulse sequence because the PETRA pulse sequence needs to sample a large number of radial views.

[0062] Parallel imaging (also known as parallel acquisition) is a technique that simultaneously acquires magnetic resonance signal data from multiple receive coils, thereby reducing the data sampling in k-space to accelerate the scan and maintain spatial resolution. Each coil presents different spatial sensitivity information (Profile), which serves as additional spatial encoding. The image is reconstructed by undersampling k-space and utilizing the sensitivity information to accelerate the acquisition. There are two main technical paths here: SENSE (SENSitivity Encoding) is based on explicit coil sensitivity; and GRAPPA (GeneRalized Autocalibrating Partial Parallel Acquisition), which is based on exploiting the known correlations in k-space. ESPIRiT is a technique based on eigenvector self-calibration that combines the advantages of SENSE and the robustness to specific errors similar to GRAPPA.

[0063] Compressed Sensing is a technique that uses sparse signals to recover highly undersampled data, suppressing the artifacts generated by undersampling in magnetic resonance images through the sparsity or sparse representation of magnetic resonance image data in the transform domain. The image reconstruction based on compressed sensing usually involves solving an underdetermined linear system.

[0064] Considering the above factors, the present disclosure provides a technique for accelerating imaging based on non-Cartesian parallel imaging and compressed sensing. The above two techniques use different prior information, and thus can be combined to obtain a higher acceleration factor, which is called PICS (Parallel imaging and Compressed sensing).

[0065] Figure 1 A schematic diagram of k-space obtained based on the execution of the PETRA pulse sequence is shown.

[0066] Reference Figure 1, the first k-space 5 acquired based on the execution of the PETRA pulse sequence consists of two parts, including: radial acquisition spokes in the peripheral region 5e that cover most of the 3D k-space except for the central region 5d, i.e., the radial trajectory 5b as a non-Cartesian trajectory, and multiple Cartesian acquisition single points or Cartesian trajectories 5a within the sphere covering the central region 5d of the 3D k-space. When obtaining the Cartesian acquisition single points or Cartesian trajectories 5a, the sphere of the central region 5d of the first k-space 5 is covered in a pointwise encoding (Pointwise Encoding - PE) manner, and this region is usually the part that is easily missing during non-Cartesian trajectory acquisition; when obtaining the magnetic resonance signals acquired along the radial trajectory 5b, when acquiring along a radial spoke, the gradient magnetic field is controlled to be set to a constant almost throughout the entire repetition sequence, and only slightly changed at the end of each repetition sequence. Since the number of spokes required for full sampling along the radial trajectory 5b or non-Cartesian trajectory is relatively large, usually about 50,000 radial trajectories or spokes need to be acquired for full acquisition to meet the image resolution requirements or be suitable for medical diagnosis requirements. This makes it time-consuming to obtain the hybrid trajectory k-space of the full-sampled radial trajectory 5b, while the number of points acquired in the central region 5d along the Cartesian trajectory 5a for meeting the image resolution is relatively small.

[0067] It should be noted that Figure 1 shows the acquisition trajectory of the first k-space 5 based on the execution of the PETRA pulse sequence in a 2D projection. Among them, the dotted circles represent the k-space acquired along the Cartesian trajectory 5a, and the solid squares represent the reconstructed field of view 5c (FoV - Field of View). The acquisition of the radial trajectory 5b occupies most of the scanning time. If the number of acquisitions related to the radial trajectory 5b can be reduced, the scanning time can be saved. However, according to the Nyquist criterion, undersampling of the radial trajectory 5b will result in image artifacts.

[0068] Figure 2 shows a schematic diagram of a magnetic resonance imaging system including a receiving coil in an exemplary embodiment.

[0069] Refer to Figure 2, describes a magnetic resonance (MR) system 1, using which an improved method for generating magnetic resonance images of an examination object P can be performed. The examination object P can be placed within the examination bore in the examination area 2 of the magnetic resonance imaging system 1. The magnetic resonance imaging system 1 includes a magnet 10 for generating a basic field B0. The examination object P is moved into the examination area 2 at the center of the magnet 10 such that the magnetic resonance imaging system 1 receives spatially encoded magnetic resonance signals from the examination area 2, which are acquired by at least one receiving coil 3. By, for example, emitting a radiofrequency (RF) pulse sequence with a radiofrequency coil 7 and switching the gradient magnetic field provided by the gradient coils 6, the nuclear spins in the examination area 2 can be deflected from the equilibrium position, and spin echoes and stimulated echo signals can be generated in the examination object P. The currents caused by the stimulated echo signals and spin echoes generated by the basic field B0 and returning to the equilibrium position can be converted into magnetic resonance signals for generating MR measurement data in the receiving coil 3. The MR measurement signals detected by each receiving coil 3 are associated with corresponding MR measurement channels. The general operating mode for generating magnetic resonance images and detecting magnetic resonance signals is not described here.

[0070] The MR system 1 includes a magnetic resonance (MR) controller 11 for controlling the MR system 1. The MR controller is adapted to execute a method of improved magnetic resonance imaging according to the present disclosure. The MR controller 11 further includes a gradient controller 12 and a radio frequency (RF) controller 13. The gradient controller 12 is used to control and switch the gradient magnetic field, and the RF controller 13 is used to control and transmit RF pulses for deflecting nuclear spins from the equilibrium position. In the memory 15, for example, the imaging sequence or pulse sequence for recording the acquisition of magnetic resonance images and the program for operating the MR system 1 can be stored. The recording unit 14 controls the image recording, thereby controlling the sequence of the gradient magnetic field and RF and the reception interval of the MR measurement signal as a function of a determined imaging sequence. The recording unit 14 also controls the gradient controller 12 and the RF controller 13. The magnetic resonance image that can be displayed on the display 18 can be calculated in the computing unit 16, where the operator operates the magnetic resonance imaging system 1 via the input unit 17. The memory 15 may have an imaging sequence (or pulse sequence) and program modules. When the instructions included in one of the program modules are executed in the computing unit 16, the imaging sequence and the instructions included in the program execute the magnetic resonance imaging method involved in the present disclosure. The following describes an MRI imaging method with reference to the accompanying drawings. The MRI imaging method is suitable for performing MRI imaging on the k-space acquired by a hybrid trajectory. The k-space acquired by the hybrid trajectory can fully acquire the magnetic resonance signals collected along the Cartesian trajectory to occupy the central region 5d of the first k-space 5, and under-sample the magnetic resonance signals collected along the non-Cartesian trajectory to occupy the peripheral region 5e of the first k-space 5. When applying this MRI imaging method, image resolution suitable for medical diagnosis can be obtained while reducing the time for collecting magnetic resonance signals.

[0071] Figure 3 A flowchart schematically showing the steps of performing a magnetic resonance imaging method applied to a subject in a magnetic resonance imaging system including at least one receiving coil according to an embodiment of the present disclosure

[0072] Reference Figure 3 , in step S1, the magnetic resonance signal data collected by at least one receiving coil 3 fully samples to occupy the central region 5d of the first k-space 5 along the Cartesian trajectory 5a under the control of the gradient magnetic field, and the collected magnetic resonance signal data under-samples to occupy the peripheral region 5e of the first k-space 5 along the non-Cartesian trajectory under the control of the gradient magnetic field.

[0073] Taking the execution of the PETRA sequence and recording the magnetic resonance signal data acquired from the receive coil 3 according to the PETRA acquisition trajectory as an example, under the control of the gradient magnetic field, the acquired magnetic resonance signal data is occupied in the center region 5d of the first k-space 5 along the Cartesian trajectory 5a point by point, and the center region 5d is spherical; under the control of the gradient magnetic field, the acquired magnetic resonance signal data is occupied in the peripheral region 5e of the first k-space 5 along the 3D radial trajectory 5b (or spoke trajectory - spoke) to obtain the first k-space 5 based on the PETRA acquisition trajectory. In addition, the non-Cartesian trajectory sampling for occupying the peripheral region 5e of the first k-space 5 can also be a spiral trajectory or the like.

[0074] In step S2, based on the data occupying the non-Cartesian trajectory 5b in the peripheral region 5e of the first k-space 5, the first image data is reconstructed to construct the second k-space. Among them, reconstructing the first image data at least includes suppressing the artifacts caused by the undersampling in the first image data during image reconstruction by using the sparse representation of the first image data in the transform domain, and among them, in the second k-space obtained by transforming the reconstructed first image data, the data occupies the Cartesian trajectory. Here, the reconstructed first image data can be in the second k-space obtained under Fourier transform or fast Fourier transform, and its data is recorded in the second k-space by occupying the Cartesian trajectory.

[0075] In reconstructing the first image data, first, for example, the data occupying the non-Cartesian trajectory 5b in the peripheral region 5e of the first k-space 5 is transformed to the image domain by inverse Fourier transform to obtain the first image data, then the filtering transform is applied to this first image data to perform the transformation to the sparse representation transform domain, and the artifacts introduced by the undersampling (non-Cartesian trajectory part) in the first image data are suppressed during image reconstruction. For this purpose, a solution satisfying the constraint conditions needs to be obtained when reconstructing the first image data, for example, minimizing the first image data under the l 1 norm and satisfying the regularization constraint between the k-space under Fourier transform and the measured k-space during the process of reconstructing the first image data, for example, minimizing under the l 2 norm. Here, the reconstructed first image data is reconstructed by means of compressive sensing, that is, by using the sparse representation of the first image data in the transform domain by compressive sensing to suppress the artifacts caused by undersampling in the first image during image reconstruction to obtain the reconstructed first image data. Applying Fourier transform to this reconstructed first image data can obtain the second k-space. Here, compressive sensing reconstruction can be applicable to iteration.

[0076] In step S3, data of the central region 5d of the first k-space 5 that occupies the Cartesian trajectory 5a is combined with the second k-space to generate a k-space suitable for magnetic resonance imaging. Specifically, data of the central region 5d of the first k-space 5 that occupies the Cartesian trajectory 5a can be extracted and used to replace the data of the corresponding central region that occupies the Cartesian trajectory in the second k-space to generate a k-space suitable for magnetic resonance imaging.

[0077] The present disclosure also provides a magnetic resonance imaging method, which can be applicable to the case of a k-space with a hybrid trajectory collected by multiple receive coils 3. Sampling multiple receive coils 3 can utilize the information reflecting the sensitivity distribution of the receive coils, and parallel imaging can be further used to reduce the acquisition time of MRI signals. The k-space with a hybrid trajectory can also be processed separately according to the regions where different acquisition trajectories are located to further improve it for obtaining a k-space suitable for magnetic resonance imaging.

[0078] Figure 4 A flowchart schematically shows the steps of a magnetic resonance imaging method performed on an object to be examined in a magnetic resonance imaging system including multiple receive coils according to an embodiment of the present disclosure.

[0079] Reference Figure 4 , in step S10, the magnetic resonance signal data collected by multiple receive coils is fully sampled and occupies the central region 5d of the first k-space 5 along the Cartesian trajectory 5a under the control of the gradient magnetic field, and the collected magnetic resonance signal data is undersampled and occupies the peripheral region 5e of the first k-space 5 along the non-Cartesian trajectory under the control of the gradient magnetic field. Here, the magnetic resonance signal data is collected through multiple receive coils 3 and corresponding measurement channels and is respectively recorded in multiple first k-spaces 5 in the above manner.

[0080] Here, parallel acquisition (Parallel sampling) can be applied to the magnetic resonance signal data collected by multiple receive coils 3.

[0081] In step S20, information reflecting the sensitivity distribution of multiple receive coils 3 is obtained based on the first k-space 5. Here, the sensitivity distribution information can provide phase or spatial position correction encoding for parallel imaging, so as to subsequently optimize the first k-space 5 or provide coil images for parallel imaging.

[0082] Here, methods such as SENSE, GRAPPA, SMASH, AUTO-SMASH, or SPIRiT can be used to calculate the sensitivity distribution information of the receive coils 3.

[0083] In step S30, the data of the central region 5d of the first k-space 5 that occupies the Cartesian trajectory 5a is applied with the sensitivity distribution information to construct a third k-space based on the sensitivity distribution information.

[0084] Here, data of the central region 5d of the first k-space 5 corresponding to the plurality of receiving coils 3 occupying the Cartesian trajectory 5a is transformed into the conjugate product of the image domain and the sensitivity distribution information and merged according to the measurement channels corresponding to the receiving coils 5 to obtain second image data for constructing the third k-space.

[0085] Specifically, interpolation such as zero-padding can be applied to regions other than the data occupying the Cartesian trajectory 5a in the central region 5d to obtain a set of Cartesian trajectories having the same matrix size as the sensitivity distribution information. Then, for example, the set of Cartesian trajectories is transformed into the image domain to be multiplied by the conjugate of the sensitivity distribution information, and then the image data is merged according to the measurement channels. Then, the third k-space is constructed by applying Fourier transform or fast Fourier transform to the merged image data.

[0086] In step S40, first image data is reconstructed based on the sensitivity distribution information for the data occupying non-Cartesian trajectories in the peripheral region 5e of the first k-space 5 to construct the second k-space. Here, reconstructing the first image data includes performing parallel imaging by means of the sensitivity distribution information and suppressing artifacts caused by undersampling in the first image data during image reconstruction by using the sparsity or sparse representation of the first image data in the transform domain to obtain the reconstructed first image data. And, in the second k-space, the data occupies the Cartesian trajectory and is recorded.

[0087] In an illustrated embodiment, in applying parallel imaging to the data occupying non-Cartesian trajectories such as the radial trajectory 5b in the peripheral region 5e of the first k-space 5 by means of the sensitivity distribution information to obtain the first image data, for example, first, data of the non-Cartesian trajectories occupied in the peripheral region 5e is extracted from the plurality of first k-spaces 5 corresponding to the receiving coils 3. The gridding is first applied to the plurality of data occupying non-Cartesian trajectories, and then the inverse Fourier transform is used to transform it into the image domain. In the image domain, the product can be applied to the matrices of the conjugates of the plurality of sensitivity distribution information respectively to merge and obtain the first image data. In this parallel imaging, iteration can be applied to make the obtained first image data satisfy the convergence condition. Then, in reconstructing the first image data, for example, it is transformed into the transform domain of sparse representation by performing a filtering transform, and artifacts introduced by undersampling in the first image data are suppressed during reconstruction. For this purpose, a solution satisfying the constraint conditions needs to be obtained when reconstructing the first image data, for example, minimizing the first image data under the l 1 norm and satisfying the regularization constraint between the k-space under Fourier transform and the measured k-space during the process of reconstructing the first image data, for example, minimizing under the l 2 norm. Here, the reconstructed first image data is reconstructed by means of compressed sensing, and compressed sensing reconstruction is also applicable to iteration.

[0088] In addition, a filtering transform or a non-linear filtering transform (such as a linear operator or a non-linear operator) applied when transforming data occupying a non-Cartesian trajectory, such as a radial trajectory 5b, in the peripheral region 5e of the first k-space 5 into a sparse representation transform domain can use, for example, a sparse transform such as a wavelet transform, a discrete wavelet transform, a discrete cosine transform, or a finite difference transform.

[0089] In addition, during the process of reconstructing image data using compressed sensing, coil sensitivity distribution information can be added as a weight in the optimization problem.

[0090] In addition, first image data can also be reconstructed using, for example, deep learning or a neural network. In addition, when reconstructing first image data in parallel imaging, regularizing data occupying a non-Cartesian trajectory to a regular k-space grid is not limited to applying gridding or re-gridding, or a non-uniform inverse Fourier transform (Non-uniform IFFT) can also be applied.

[0091] Here, in response to applying sensitivity distribution information to data occupying a Cartesian trajectory in the central region 5d of the first k-space 5 and merging according to the measurement channels corresponding to the receiving coils to obtain second image data, and based on the second image data being transformed into a third k-space, parallel imaging is applied to data occupying a non-Cartesian trajectory in the peripheral region 5e of the first k-space 5 based on the sensitivity distribution information to obtain first image data, and then the first image data is reconstructed to be transformed into a second k-space based on the first image data.

[0092] In step S50, a fourth k-space suitable for magnetic resonance imaging is generated based on the synthesis of the second k-space and the third k-space. Among them, during the process of generating the fourth k-space, data of the Cartesian trajectory occupied by the central region of the second k-space is replaced by data of the Cartesian trajectory occupied by the corresponding region in the third k-space, where the size of the central region of the second k-space is the same as the central region 5d of the first k-space 5.

[0093] Figure 5 A step flowchart schematically showing another magnetic resonance imaging method applied to an object to be examined in a magnetic resonance imaging system including a plurality of receiving coils according to another embodiment of the present disclosure is shown.

[0094] Reference Figure 5 , in the sampling stage S100 of this method, the acquired magnetic resonance signal data is recorded into the k-space of a mixed trajectory of corresponding Cartesian trajectories and non-Cartesian trajectories according to the gradient magnetic field, where the data recorded into the Cartesian trajectory is fully sampled, and the data recorded into the non-Cartesian trajectory can be irregular undersampled, undersampled, or randomly undersampled. Among them, step S101 in sampling stage S100 is the same as step S10 described above and will not be elaborated here.

[0095] In the stage S200 of obtaining the sensitivity distribution map for multiple receiving coils in this method, the following steps may be included:

[0096] In step S201, gridding processing is applied to the first k-space 5. Here, for example, the first k-space is gridded by convolving the first k-space with a grid kernel function. Using this gridding method, the acquisition data density of the first k-space 5 is weighted and convolved with a finite kernel function, and then resampled according to the grid, so that the original acquisition data or under this gridding is prepared in a regular k-space grid and transformed into the image domain by using the fast Fourier transform (FFT - Fast Fourier Transformation). Additionally, the first k-space 5 can also be directly transformed into the image domain by means of an inconsistent inverse Fourier transform.

[0097] In step S202, the high-frequency part is removed from the gridded first k-space 5 to extract and generate a low-resolution image. Here, for example, in the gridded first k-space 5, the high-frequency part is removed to extract the information representing the low-frequency part of the image in the central region of the first k-space, and after filling zeros in the high-frequency part, the inverse Fourier transform or the fast inverse Fourier transform is applied to obtain the low-resolution image.

[0098] In step S203, the ESPIRiT is applied based on the low-resolution image to calculate the sensitivity distribution map. Here, the ESPIRiT is applied to the low-resolution image to calculate the sensitivity distribution map. ESPIRiT can obtain the sensitivity distribution map by calculating the eigenvectors of the null space or the decomposition of the eigenvectors involved in parallel imaging.

[0099] In step S204, the sensitivity distribution map is obtained. This sensitivity distribution map records the sensitivity distribution information of the receiving coil 3.

[0100] In the stage S300 of constructing the third k-space based on the sensitivity distribution map in this method, in order to synthesize with the second k-space to obtain a k-space suitable for magnetic resonance imaging, the following steps are included:

[0101] In step S301, the first Cartesian point set represented by the data of the occupied Cartesian trajectory 5a in the central region 5d of the first k-space 5 is extracted, that is, a kind of k-space set.

[0102] In step S302, zero padding (Zero Padding or Zero Filling) is applied based on the region outside each first Cartesian point set to obtain a second Cartesian point set with the same size as the sensitivity distribution map (or the matrix of sensitivity distribution information).

[0103] In step S303, the inverse Fourier transform is applied to multiple second Cartesian point sets. Here, each second Cartesian point set corresponds to a respective receiving coil 3. Additionally, to accelerate the transformation of the second Cartesian point set from the k-space to the image domain, the inverse fast Fourier transform (Inverse FFT / iFFT) is typically applied.

[0104] In step S304, third image data corresponding to multiple receiving coils 3 is obtained. Here, the third image data reflects the position and / or phase encoding of the multi-coil image.

[0105] In step S305, the conjugate of the sensitivity profile is multiplied with the above-mentioned third image data respectively, and the second image data is obtained by combining based on the corresponding measurement channels.

[0106] Here, when multiplying the conjugate of the sensitivity profile with the above-mentioned image data respectively, it can be expressed as:

[0107] ∑ Coil IS * (1)

[0108] Wherein, in formula (1), "I" represents the third image data, "S * " represents the conjugate of the sensitivity profile, and the sum based on "Coil" represents combining the product of the conjugate of the sensitivity profile and the corresponding third image data according to the measurement channels corresponding to the receiving coil 3 to obtain the second image data.

[0109] In step S306, the fast Fourier transform (FFT) is applied to the second image data.

[0110] It should be noted that in the process of constructing the third k-space, there is no restriction on the order of optimizing the k-space represented as the second Cartesian point set with the sensitivity profile to provide the sensitivity profile of the receiving coil. For example, convolution can also be applied to the transform domain of the second Cartesian point set and the sensitivity profile first and then transformed to the image domain, and then the second image data is obtained by combining according to the measurement channels; then the Fourier transform is applied to the second image data to obtain the third k-space. This embodiment does not limit the above sequence.

[0111] In step S307, the third k-space is obtained, denoted as k Cartesian 。

[0112] In the first data image reconstruction stage S400 of this method, to construct the second k-space by parallel imaging and compressive sensing reconstruction using the sensitivity profile, and in the second k-space, data is recorded occupying a Cartesian trajectory to be combined with the third k-space to obtain a k-space suitable for magnetic resonance imaging, which includes the following steps:

[0113] In step S401, data of the occupied non-Cartesian trajectories in the peripheral region 5e is extracted from the first k-space 5. Here, taking PETRA raw data as an example, the non-Cartesian trajectory is a 3D radial trajectory 5b.

[0114] In step S402, parallel imaging is applied to the data of the first k-space 5 of the occupied non-Cartesian trajectory part based on the sensitivity distribution map to obtain first image data and the first image data is reconstructed by means of compressed sensing.

[0115] Here, for example, the data of the first k-space of the occupied non-Cartesian trajectory part is conjugated and multiplied with the sensitivity distribution map after being transformed into the image domain, and then first image data is obtained by means of parallel imaging, and then compressed sensing is applied to reconstruct the first image data. When applying compressed sensing reconstruction, for example, when the first image data is transformed into a transform domain of sparse representation under the execution of a filtering transform, the constraint conditions of minimization and regularization under the l 1 norm are satisfied to utilize the sparsity or sparse representation of the first image data in the transform domain to suppress the artifacts introduced by the undersampling of the first image data along the non-Cartesian trajectory in image reconstruction. The above compressed sensing reconstruction is suitable for iteration, and the finally reconstructed first image data can be obtained after multiple iterations.

[0116] In step S403, fast Fourier transform (FFT) is applied to the reconstructed first image data.

[0117] In step S405, a second k-space is obtained, denoted as k Radial . In the second k-space, data occupying the Cartesian trajectory has been recorded.

[0118] In the synthesis stage S500 of this method, it includes the following steps:

[0119] In step S501, based on the third k-space k Cartesian and the second k-space k Radial a fourth k-space suitable for magnetic resonance imaging is synthesized and generated. When synthesizing the third space k Cartesian and the second space k Radial , the data of the Cartesian trajectory occupied by the central region of the second k-space k Radial is replaced by the data of the Cartesian trajectory occupied by the corresponding region in the third k-space k Cartesian , and the size of the central region of the second k-space k Radial can be the same as that of the central region 5d of the first k-space 5.

[0120] In step S502, inverse fast Fourier transform (iFFT - Inverse FFT) is applied to the fourth k-space.

[0121] In step S503, the final magnetic resonance image is obtained.

[0122] Figure 6 FIG. shows a comparison schematic diagram of the image reconstruction of the magnetic resonance imaging method applied to the hybrid trajectory sequence and the standard-based image reconstruction according to an embodiment of the present disclosure.

[0123] Reference Figure 6 As shown, wherein Figure 6 A and B are images obtained by applying the standard non-uniform fast Fourier reconstruction method to the PETRA sequence with a radial trajectory of 25,000 and 10,000 spokes, respectively, which has a hybrid trajectory k-space (including a Cartesian trajectory with full acquisition in the central region). It should be noted that the radial trajectory in the outer region of this k-space is highly undersampled and sparse, and the time-consuming is 5 minutes and 2:10 minutes, respectively. Figure 6 C and D are the k-spaces obtained from the PETRA sequence with a radial trajectory of 25,000 and 10,000 spokes, respectively, and the final images are obtained by performing the magnetic resonance imaging method provided by the present disclosure, that is, image reconstruction based on the improved parallel imaging and compressive sensing method on the obtained k-space. Figure 6 G is the k-space obtained from the PETRA sequence with a radial trajectory of 50,000 spokes, the radial trajectory of which is fully acquired, the resolution is 256*256*256, and the time-consuming is about 9:42 minutes. The conclusion drawn by comparison is that when the acceleration factor is 2 (i.e., the radial trajectory of 25,000 spokes is acquired), the image quality obtained by the magnetic resonance imaging method provided by the present disclosure is very close to the fully acquired image; when the acceleration factor is 5 (i.e., the radial trajectory of 10,000 spokes is acquired), the image quality obtained by the magnetic resonance imaging method provided by the present disclosure is clinically acceptable and has a high image resolution, and the image quality is better than that obtained by the standard non-uniform fast Fourier image reconstruction (i.e., Figure 6 C).

[0124] Another aspect of the present disclosure provides a computer program product, including a program executed by at least one processor of a computing device, the program including instructions, wherein the execution of the instructions causes at least one processor to execute the magnetic resonance imaging method according to the foregoing.

[0125] Another aspect of the present disclosure provides a computer-readable storage medium storing a program, the program including instructions, which when executed by an electronic device, cause the electronic device to execute the magnetic resonance imaging method according to the foregoing.

[0126] Figure 7 FIG. shows a structural diagram of a computing device that can be applied to an exemplary embodiment.

[0127] See Figure 7 As shown, a computing device 2000 will now be described, which is an example of an electronic device to which various aspects of the present disclosure can be applied. The computing device 2000 can be any machine configured to perform processing and / or computations, and can be, but is not limited to, a workstation, a server, a desktop computer, a laptop computer, a tablet computer, a personal digital assistant, a robot, a smart phone, an in-vehicle computer, or any combination thereof. The above-described medical imaging method for detecting motion can be implemented in whole or at least in part by the computing device 2000 or a similar device or system.

[0128] The computing device 2000 can include (possibly via one or more interfaces) elements connected to or communicating with a bus 2002. For example, the computing device 2000 can include a bus 2002, one or more processors 2004, one or more input devices 2006, and one or more output devices 2008. The one or more processors 2004 can be any type of processor, and can include, but are not limited to, one or more general-purpose processors and / or one or more special-purpose processors (such as special processing chips). The input device 2006 can be any type of device capable of inputting information into the computing device 2000, and can include, but are not limited to, a mouse, a keyboard, a touch screen, a microphone, and / or a remote control. The output device 2008 can be any type of device capable of presenting information, and can include, but are not limited to, a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The computing device 2000 can also include a non-transitory storage device 2010 or be connected to the non-transitory storage device 2010. The non-transitory storage device can be any storage device that is non-transitory and can implement data storage, and can include, but are not limited to, a disk drive, an optical storage device, a solid-state memory, a floppy disk, a flexible disk, a hard disk, a magnetic tape, or any other magnetic medium, an optical disk or any other optical medium, a ROM (read-only memory), a RAM (random access memory), a cache memory, and / or any other memory chip or cartridge, and / or any other medium from which a computer can read data, instructions, and / or code. The non-transitory storage device 2010 can be removable from the interface. The non-transitory storage device 2010 can have data / programs (including instructions) / code for implementing the above-described methods and steps. The computing device 2000 can also include a communication device 2012. The communication device 2012 can be any type of device or system capable of enabling communication with external devices and / or with a network, and can include, but are not limited to, a modem, a network card, an infrared communication device, a wireless communication device, and / or a chipset, such as a BluetoothTM device, a 1302.11 device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0129] The computing device 2000 may also include a working memory 2014, which can be any type of working memory that can store programs (including instructions) and / or data useful for the operation of the processor 2004, and may include, but is not limited to, random access memory and / or read-only memory devices.

[0130] Software elements (programs) may be located in the working memory 2014, including but not limited to an operating system 2016, one or more application programs 2018, drivers, and / or other data and code. Instructions for performing the above methods and steps may be included in one or more application programs 2018, and the above magnetic resonance imaging method may be implemented by the processor 2004 reading and executing the instructions of one or more application programs 2018. More specifically, in the above magnetic resonance imaging method, steps S1 to S30, S10 to S50, and steps S101, steps S201 to S204, steps S301 to S307, steps S401 to S404, steps S501 to S503 may be implemented, for example, by the processor 2004 executing an application program 2018 having the steps S10 to S50, and steps S101, steps S201 to S204, steps S301 to S307, steps S401 to S404, steps S501 to S503. In addition, other steps in the above magnetic resonance imaging method may be implemented, for example, by the processor 2004 executing an application program 2018 having instructions for performing the corresponding steps. The executable code or source code of the instructions of the software elements (programs) may be stored in a non-transitory computer-readable storage medium (such as the above storage device 2010), and when executed, may be loaded into the working memory 2014 (possibly compiled and / or installed). The executable code or source code of the instructions of the software elements (programs) may also be downloaded from a remote location.

[0131] It should also be understood that various modifications can be made according to specific requirements. For example, custom hardware may also be used, and / or specific elements may be implemented using hardware, software, firmware, middleware, microcode, a hardware description language, or any combination thereof. For example, some or all of the disclosed methods and devices may be implemented by programming hardware (such as programmable logic circuits including field programmable gate arrays (FPGAs) and / or programmable logic arrays (PLAs)) using the logic and algorithms according to the present disclosure, with an assembly language or a hardware programming language (such as VERILOG, VHDL, C++).

[0132] It should also be understood that the foregoing method can be implemented in a server-client mode. For example, the client can receive data input by the user and send the data to the server. The client can also receive data input by the user, perform a part of the processing in the foregoing method, and send the processed data to the server. The server can receive data from the client, execute the foregoing method or another part of the foregoing method, and return the execution result to the client. The client can receive the execution result of the method from the server and, for example, present it to the user through an output device.

[0133] It should also be understood that the components of the computing device 2000 can be distributed over a network. For example, one processor can be used to perform some processing while another processor located away from the one processor can perform other processing. Other components of the computing system 2000 can be distributed similarly. In this way, the computing device 2000 can be interpreted as a distributed computing system that performs processing at multiple locations.

[0134] Although the embodiments or examples of the present disclosure have been described with reference to the accompanying drawings, it should be understood that the foregoing methods, systems and devices are merely exemplary embodiments or examples, and the scope of the present invention is not limited by these embodiments or examples, but is only defined by the authorized claims and their equivalent scope. Various elements in the embodiments or examples can be omitted or replaced by their equivalent elements. In addition, the steps can be executed in an order different from that described in the present disclosure. Further, various elements in the embodiments or examples can be combined in various ways. Importantly, with the evolution of technology, many of the elements described herein can be replaced by equivalent elements that emerge after the present disclosure.

Claims

1. A magnetic resonance imaging method, characterized in that, the method comprises the following steps: Magnetic resonance signal data collected by at least one receiving coil fully samples and occupies the central region of the first k-space along a Cartesian trajectory under the control of a gradient magnetic field, and the collected magnetic resonance signal data under-samples and occupies the peripheral region of the first k-space along a non-Cartesian trajectory under the control of a gradient magnetic field; Based on the data occupying the non-Cartesian trajectory in the peripheral region of the first k-space, reconstruct the first image data to construct a second k-space, wherein, the reconstruction of the first image data at least includes suppressing artifacts caused by the under-sampling in the first image data during image reconstruction by using the sparse representation of the first image data in the transform domain, and wherein, in the second k-space obtained by transforming the reconstructed first image data, the data occupies a Cartesian trajectory; Based on the data in the central region of the first k-space occupying the Cartesian trajectory and the second k-space, synthesize to generate a k-space suitable for magnetic resonance imaging.

2. The magnetic resonance imaging method according to claim 1, wherein, the magnetic resonance imaging method further comprises: Obtaining sensitivity distribution information reflecting multiple receiving coils.

3. The magnetic resonance imaging method according to claim 2, wherein, the extraction of the data occupying the non-Cartesian trajectory in the peripheral region of the first k-space to reconstruct the first image data to construct a second k-space includes: The reconstruction of the first image data further includes performing parallel imaging by means of the sensitivity distribution information.

4. The magnetic resonance imaging method according to claim 2, wherein, the extraction of the data occupying the non-Cartesian trajectory in the peripheral region of the first k-space to reconstruct the first image data to construct a second k-space includes: Performing parallel imaging on the data of multiple non-Cartesian trajectories occupied by the periphery of the first k-space to obtain first image data; Using the sparse representation of the first image data in the transform domain to suppress artifacts caused by the under-sampling in the first image during image reconstruction to obtain the reconstructed first image data; Applying Fourier transform based on the reconstructed first image data to obtain the second k-space.

5. The magnetic resonance imaging method according to claim 2, wherein, the synthesis of the data in the central region of the first k-space occupying the Cartesian trajectory and the second k-space to generate a k-space suitable for magnetic resonance imaging includes: Applying sensitivity distribution information to the data in the central regions of the first k-spaces corresponding to the receiving coils that occupy Cartesian trajectories to construct a third k-space.

6. The magnetic resonance imaging method according to claim 5, wherein, the application of sensitivity distribution information to the data in the central regions of the first k-spaces corresponding to the receiving coils that occupy Cartesian trajectories to construct a third k-space includes: Data corresponding to a central region of the first k-space occupied by the reception coils is transformed into the conjugate product of the image domain and the sensitivity distribution information, and the second image data is combined according to the measurement channels corresponding to the reception coils to construct the third k-space.

7. The magnetic resonance imaging method according to claim 5 or 6, wherein applying the sensitivity distribution information to the data occupying the Cartesian trajectory in the central region of the first k-space to construct the third k-space includes: extracting the data of the occupied Cartesian trajectory in the central region from the first k-space and applying zero filling in the regions outside the central region with reference to the sensitivity distribution information to obtain a plurality of Cartesian point sets having the same matrix size as the sensitivity distribution information, wherein the plurality of Cartesian point sets correspond to the acquisitions of the reception coils; applying an inverse Fourier transform to the plurality of Cartesian point sets to obtain a plurality of third image data, and multiplying the conjugate of the sensitivity distribution information with the third image data respectively; and combining according to the measurement channels corresponding to the reception coils to obtain the second image data; transforming to the third k-space based on the second image data.

8. The magnetic resonance imaging method according to any one of claims 5 or 6, wherein synthesizing the data occupying the Cartesian trajectory in the central region of the first k-space with the second k-space to generate a k-space suitable for magnetic resonance imaging further includes: synthesizing the second k-space and the third k-space to generate a k-space suitable for image reconstruction, wherein the data of the occupied Cartesian trajectory in the central region of the second k-space is replaced by the data of the occupied Cartesian trajectory in the corresponding region of the third k-space, and wherein the size of the central region of the second k-space is the same as that of the central region of the first k-space.

9. The magnetic resonance imaging method according to any one of claims 1 to 4, wherein the non-Cartesian trajectory is a 3D radial trajectory.

10. The magnetic resonance imaging method according to claim 1, wherein synthesizing the data occupying the Cartesian trajectory in the central region of the first k-space with the second k-space to generate a k-space suitable for magnetic resonance imaging includes: generating a k-space suitable for magnetic resonance imaging by replacing the data of the occupied Cartesian trajectory in the corresponding central region of the second k-space based on the extracted data of the occupied Cartesian trajectory in the central region of the first k-space.

11. The magnetic resonance imaging method according to claim 2, wherein obtaining the sensitivity distribution information reflecting the reception coils includes: applying gridding processing to the first k-space; removing the high-frequency part from the gridded first k-space to extract and generate a low-resolution image; calculating the sensitivity distribution map based on the low-resolution image.

12. The magnetic resonance imaging method according to claim 11, wherein applying ESPIRiT to the low-resolution image to calculate the eigenvectors from the null space to obtain the sensitivity distribution map.

13. A magnetic resonance imaging system (1), comprising: at least one reception coil (3); a magnetic resonance controller (11); and A memory (15) storing a program, the program including instructions, a magnetic resonance controller (11) being connected to the memory (15) and configured to execute the instructions, wherein the instructions, when executed by the magnetic resonance controller (11), cause the magnetic resonance imaging system (1) to perform the magnetic resonance imaging method according to any one of claims 1 to 12.

14. A computing device configured to be used in a magnetic resonance imaging system including at least one receive coil for a magnetic resonance imaging method with a k-space trajectory, the k-space trajectory at least including that magnetic resonance signal data acquired by the receive coil fully samples and occupies a central region of the k-space along a Cartesian trajectory and that the acquired magnetic resonance signal data under-samples and occupies a peripheral region of the k-space along a non-Cartesian trajectory, the computing device including a memory and at least one processor, the memory containing instructions executed by the at least one processor, execution of the instructions causing the computing device to perform the magnetic resonance imaging method according to any one of claims 1 to 12.

15. A computer program product including a program executed by at least one processor of a computing device, the program including instructions, wherein execution of the instructions causes the at least one processor to perform the magnetic resonance imaging method according to any one of claims 1 to 12.

16. A computer-readable storage medium storing a program, the program including instructions, characterized in that when the instructions are executed by an electronic device, the instructions cause the electronic device to perform the magnetic resonance imaging method according to any one of claims 1 to 12.

Images