CN113391251B - Magnetic resonance image reconstruction method, device and equipment - Google Patents
Magnetic resonance image reconstruction method, device and equipment Download PDFInfo
- Publication number
- CN113391251B CN113391251B CN202010169908.0A CN202010169908A CN113391251B CN 113391251 B CN113391251 B CN 113391251B CN 202010169908 A CN202010169908 A CN 202010169908A CN 113391251 B CN113391251 B CN 113391251B
- Authority
- CN
- China
- Prior art keywords
- space data
- sampling
- partial
- scanning
- phase
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
- G01R33/56—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
- G01R33/561—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution by reduction of the scanning time, i.e. fast acquiring systems, e.g. using echo-planar pulse sequences
- G01R33/5615—Echo train techniques involving acquiring plural, differently encoded, echo signals after one RF excitation, e.g. using gradient refocusing in echo planar imaging [EPI], RF refocusing in rapid acquisition with relaxation enhancement [RARE] or using both RF and gradient refocusing in gradient and spin echo imaging [GRASE]
- G01R33/5618—Echo train techniques involving acquiring plural, differently encoded, echo signals after one RF excitation, e.g. using gradient refocusing in echo planar imaging [EPI], RF refocusing in rapid acquisition with relaxation enhancement [RARE] or using both RF and gradient refocusing in gradient and spin echo imaging [GRASE] using both RF and gradient refocusing, e.g. GRASE
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/4818—MR characterised by data acquisition along a specific k-space trajectory or by the temporal order of k-space coverage, e.g. centric or segmented coverage of k-space
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
- G06T11/006—Inverse problem, transformation from projection-space into object-space, e.g. transform methods, back-projection, algebraic methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/30—Assessment of water resources
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- High Energy & Nuclear Physics (AREA)
- Condensed Matter Physics & Semiconductors (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Pure & Applied Mathematics (AREA)
- Algebra (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Signal Processing (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The embodiment of the invention discloses a method, a device and equipment for reconstructing a magnetic resonance image. The method comprises the following steps: acquiring two partial sampling K space data of a scanning object, wherein the filling directions of the two partial sampling K space data in the phase coding direction are opposite, and the two partial sampling K space data comprise data lines with the same phase coding in a preset proportion; weighting and combining the two partial sampling K space data to generate full sampling K space data corresponding to the scanning object; and carrying out image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanning object. By the technical scheme, the magnetic resonance image based on the partial Fourier technology is reconstructed more efficiently, and the signal-to-noise ratio of the reconstructed image is improved.
Description
Technical Field
Embodiments of the present invention relate to medical image processing technologies, and in particular, to a method, an apparatus, and a device for reconstructing a magnetic resonance image.
Background
The requirement of the magnetic resonance imaging technology on the scanning time is high, in order to reduce the scanning time, a partial fourier technology is generally used, and the purpose of rapid scanning is achieved by reducing the acquisition line number of the PE coding direction in the K space.
Currently, image reconstruction techniques for such partial K-space are to calculate and fill in missing K-space data to reconstruct an image. For example, the missing K-space data is filled using a zero padding method or a holomodyne technique.
However, the existing image reconstruction technology aiming at partial Fourier greatly blurs an image, and the Homodyne technology also generates artifacts due to inaccurate estimated image phase, so that the signal-to-noise ratio of the image is reduced.
Disclosure of Invention
The embodiment of the invention provides a magnetic resonance image reconstruction method, a device and equipment, which are used for realizing more efficient reconstruction of a magnetic resonance image based on a partial Fourier technology and improving the signal-to-noise ratio of the reconstructed image.
In a first aspect, an embodiment of the present invention provides a magnetic resonance image reconstruction method, including:
acquiring two partial sampling K space data of a scanning object, wherein the filling directions of the two partial sampling K space data in the phase coding direction are opposite, and the two partial sampling K space data comprise data lines with the same phase coding in a preset proportion;
weighting and combining the two partial sampling K space data to generate full sampling K space data corresponding to the scanning object;
And carrying out image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanning object.
In a second aspect, an embodiment of the present invention further provides a magnetic resonance image reconstruction apparatus, including:
the partial sampling K space data acquisition module is used for acquiring two partial sampling K space data of a scanning object, wherein the filling directions of the two partial sampling K space data in the phase coding direction are opposite, and the two partial sampling K space data comprise data lines with the same phase coding of a preset proportion;
the full sampling K space data generation module is used for carrying out weighting and combining processing on the two partial sampling K space data to generate full sampling K space data corresponding to the scanning object;
and the image reconstruction module is used for carrying out image reconstruction according to the full-sampling K space data and generating a magnetic resonance reconstruction image corresponding to the scanning object.
In a third aspect, an embodiment of the present invention further provides an apparatus, including:
a scanner having an accommodation space, and configured to perform scanning on a scanning object in the accommodation space;
one or more processors;
Storage means for storing one or more programs,
when the one or more programs are executed by the one or more processors, the one or more processors are caused to control the scanner to implement:
applying a first scanning sequence to the scanning object to obtain a first part of K space data, wherein the first scanning sequence comprises a first phase encoding gradient pulse;
applying a second scanning sequence to the scanning object to obtain second part of K space data, wherein the second scanning sequence comprises second phase encoding gradient pulses, and the gradient directions of the second phase encoding gradient pulses and the first phase encoding gradient pulses are opposite;
weighting and combining the first part of K space data and the second part of K space data to generate full-sampling K space data corresponding to the scanning object;
and carrying out image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanning object.
According to the embodiment of the invention, the two partial sampling K space data of the scanning object are obtained, wherein the filling directions of the two partial sampling K space data in the phase coding direction are opposite, and the two partial sampling K space data comprise data lines with the same phase coding in a preset proportion; weighting and combining the two partial sampling K space data to generate full sampling K space data corresponding to the scanning object; and carrying out image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanned object. The method realizes the weighted combination of two actually acquired partial sampling K space data to generate one real full sampling K space data of the scanning object, further obtains a magnetic resonance reconstruction image, solves the problem of image blurring caused by the estimation of the missing K space data of an image reconstruction algorithm of a partial Fourier technology, simplifies the magnetic resonance image reconstruction algorithm, and improves the reconstruction efficiency and the image signal to noise ratio of the magnetic resonance image.
Drawings
Fig. 1 is a flow chart of a method of reconstructing a magnetic resonance image according to a first embodiment of the present invention;
FIG. 2 is a schematic diagram showing the distribution of two partial samples of K-space data in K-space according to a first embodiment of the present invention;
fig. 3 is a flow chart of a magnetic resonance image reconstruction method in a second embodiment of the present invention;
FIG. 4 is a schematic diagram of a first phase encoding gradient pulse and a second phase encoding gradient pulse in a second embodiment of the present invention;
FIG. 5 is a schematic diagram of a preset magnetic resonance echo sequence for acquiring two partial samples of K-space data in accordance with a second embodiment of the present invention;
FIG. 6 is a schematic diagram of filling partial K space data by adopting a partial echo acquisition mode in a second embodiment of the present invention;
FIG. 7 is a graph of the results of magnetic resonance images reconstructed using different magnetic resonance image reconstruction methods in a second embodiment of the present invention;
fig. 8 is a schematic structural diagram of a magnetic resonance image reconstruction device according to a third embodiment of the present invention;
fig. 9a is a schematic structural view of an apparatus according to a fourth embodiment of the present invention;
fig. 9b is a schematic structural diagram of a console included in the apparatus in the fourth embodiment of the present invention.
Detailed Description
The invention is described in further detail below with reference to the drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting thereof. It should be further noted that, for convenience of description, only some, but not all of the structures related to the present invention are shown in the drawings.
Example 1
The magnetic resonance image reconstruction method provided by the embodiment can be suitable for partial sampling scanning magnetic resonance image reconstruction. The method may be performed by a magnetic resonance image reconstruction device, which may be implemented in software and/or hardware, which may be integrated in a device having image processing functions, such as an image processing device, typically in a magnetic resonance imaging system, like a notebook, desktop or server, etc. Referring to fig. 1, the method of this embodiment specifically includes the following steps:
s110, acquiring two part sampling K space data of a scanning object, wherein the filling directions of the two part sampling K space data in the phase coding direction are opposite, and the two part sampling K space data comprise data lines with the same phase coding in a preset proportion.
Wherein the K-space, also called fourier space, is the filling space of the original digital data of the MR signal with spatially localized encoded information. The spatial location coding information in the original digital data can be decoded by performing Fourier transform on the K-space data, and MR signals with different frequencies, phases and amplitudes are decomposed, wherein the different frequencies and phases represent different spatial positions, and the amplitudes represent the MR signal strength. MR digital information with different frequencies, phases and signal intensities is distributed to corresponding pixels, so that MR image data can be obtained, namely an MR image is reconstructed.
The partial sampling K-space data refers to magnetic resonance scan data in which a partial phase encoding data line is filled in the K-space, and may be, for example, K-space data in which slightly more than half of the K-space is filled along the phase encoding direction; the K-space data may be estimated based on some K-space filling algorithm, thereby supplementing the K-space data of slightly more than half of the K-space, and obtaining K-space data completely filled with the phase-encoded data line. The K space filling algorithm can be that the sensitivity of a radio frequency receiving coil for receiving magnetic resonance signals is obtained, then K space data of more than half of the K space is subjected to low-density acquisition filling, and then phase coding lines with missing K space are filled by using coil sensitivity information; or the K space data of slightly more than half of the K space can be acquired and filled at low density, and then the trained neural network is utilized to supplement the phase encoding line with the missing K space. Illustratively, the trained neural network may be obtained by: a sample K space may be pre-acquired, the sample K space comprising a plurality of sets of first K spaces and a plurality of sets of second K spaces, wherein slightly more than half of the first K spaces are completely filled; the corresponding part (slightly more than half area) of the second K space is filled with low density; the sample K space is input into the neural network for training to obtain a trained neural network, wherein the first K space is used as gold standard data, and the second K space is used as input data.
The preset proportion is a preset proportion of phase coding data in the K space to the whole K space data, and the preset proportion is used for controlling the number of data lines with the same phase coding in the two partial sampling K space data. Illustratively, the predetermined proportion is 20%. The purpose of this arrangement is to ensure the accuracy of the merging of the K-space data of the two subsequent partial samples.
Specifically, the partial fourier image reconstruction algorithm in the related art uses the acquired data in the K space to estimate the data of the K space portion of the data not acquired, so as to obtain the fully sampled K space data with the data filling the whole K space, and performs image reconstruction based on the fully sampled K space data. However, the K-space data estimated in this process is not accurate enough, resulting in a low signal-to-noise ratio of the reconstructed image. In view of this, in the embodiment of the present invention, the fully sampled K-space data is generated by using two actually acquired partially sampled K-space data, so that the authenticity of the basic data for image reconstruction can be ensured, and the image reconstruction error caused by estimating the K-space data can be reduced to a certain extent. Since the full-sample K-space data is derived from all of the two partial-sample K-space data, the union of the data filling ranges of the two partial-sample K-space data is required to cover the entire K-space, and thus, the filling directions of the two partial-sample K-space data in the phase encoding direction are required to be opposite. Taking the two-dimensional K space in fig. 2 as an example, the horizontal axes of the two K spaces in the figure represent the Kx direction, i.e., the frequency coding direction; the vertical axis represents the Ky direction, i.e. the phase encoding direction. When the phase encoding direction corresponding to one part of the sampling K space data (K_A) is from high to low, the phase encoding direction corresponding to the other part of the sampling K space data (K_B) is from low to high, and the two K spaces are partially filled by a plurality of data lines.
In addition, in order to reduce the scanning time as much as possible and ensure the accuracy of the subsequent data merging, a minimum number limit of data lines with the same phase code needs to be set for the two part sampling K-space data, that is, a preset proportion is set, so that a proper magnetic resonance scanning sequence can be designed to acquire the scanning data, and two part sampling K-data k_a and k_b with the same phase code between the data and the region 201 of the data lines with the same phase code occupying the whole K-space are obtained, that is, the proportion of the two part sampling K-data k_a and k_b with the same phase code is greater than or equal to the preset proportion, that is, the two K-space contains one or more data lines with the same phase code.
It should be noted that, the distribution of the two partial sampling K-space data k_a and k_b in the K-space may be non-mirror symmetry, as long as the two partial sampling K-space data have the same data lines in the phase encoding direction that meet the requirement, as shown in fig. 2 (B); in order to further reduce the scanning time, the distribution of the two partial sampling K-space data in the K-space is set to be mirror symmetrical, that is, the number of data lines in the two partial sampling K-space data k_a and k_b is identical, but the phase encoding directions are opposite, as shown in fig. 2 (a).
Taking fig. 2 (a) as an example, the first partial sample K-space data k_a is filled in as follows: the partial region is gradually filled from the side of the high K space phase encoding direction to the side of the low K space phase encoding direction from the space center. In this embodiment, ky= -127 is filled first, then ky= -126, … …, ky=0, and finally ky= +20. The second partial sample K-space data k_b is filled in a mirrored manner: ky= +128 is filled first, then ky= +127, … …, ky=0, and finally ky= -19.
In practical implementation, two parts of the scan object meeting the requirement need to acquire sampling K-space data before image reconstruction, which can be obtained by real-time scanning by using a designed magnetic resonance scanning sequence or can be obtained by directly reading from a storage device.
And S120, carrying out weighted combination processing on the two partial sampling K space data to generate full sampling K space data corresponding to the scanning object.
Specifically, in the process of merging two partial sampling K-space data, it is necessary to ensure that the difference between overlapping areas of two K-space data is as small as possible to ensure that the subsequently reconstructed image has a higher signal-to-noise ratio, so that the two partial sampling K-space data needs to be processed first to eliminate the scanning background noise (gaussian noise). In practice, the two partial sampled K-space data are weighted and summed to obtain a weighted and combined K-space data, which is the full sampled K-space data.
S130, performing image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanned object.
In particular, the fully sampled K-space data is converted into an image using an image reconstruction technique, such as inverse fourier transform, to obtain a magnetic resonance reconstructed image of the scan subject.
According to the technical scheme, two partial sampling K space data of a scanning object are obtained, wherein filling directions of the two partial sampling K space data in a phase encoding direction are opposite, and the two partial sampling K space data comprise data lines with the same phase encoding in a preset proportion; weighting and combining the two partial sampling K space data to generate full sampling K space data corresponding to the scanning object; and carrying out image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanned object. The method realizes the weighted combination of two actually acquired partial sampling K space data to generate one real full sampling K space data of the scanning object, further obtains a magnetic resonance reconstruction image, solves the problem of image blurring caused by the estimation of the missing K space data of an image reconstruction algorithm of a partial Fourier technology, simplifies the magnetic resonance image reconstruction algorithm, and improves the reconstruction efficiency and the image signal to noise ratio of the magnetic resonance image.
Example two
On the basis of the first embodiment, the present embodiment further optimizes "weighting and combining two partial sampling K-space data to generate full sampling K-space data corresponding to the scan object". On this basis, the relevant step of phase correction can be further added. Wherein the explanation of the same or corresponding terms as those of the above embodiments is not repeated herein. Referring to fig. 3, the magnetic resonance image reconstruction method provided in the present embodiment includes:
S210, acquiring two part sampling K space data of a scanning object, wherein the filling directions of the two part sampling K space data in the phase coding direction are opposite, and the two part sampling K space data comprise data lines with the same phase coding in a preset proportion.
Illustratively, acquiring two partial samples of K-space data of a scan object includes: applying a first scan sequence to the scan object to obtain a first portion of K-space data, wherein the first scan sequence comprises a first phase encoding gradient pulse; and applying a second scanning sequence to the scanning object to obtain second part of K space data, wherein the second scanning sequence comprises second phase encoding gradient pulses, and the second phase encoding gradient pulses have the same field intensity change as the first phase encoding gradient pulses and have opposite gradient directions. The first scan sequence, the second scan sequence may comprise, for example, a fast spin echo sequence, a spin echo sequence, or the like, imaging sequences having long echo chains. In addition, in order to further provide an imaging effect, the first scanning sequence and the second scanning sequence may also be provided with a diffusion sequence before the imaging sequence.
Specifically, referring to fig. 4 (a), for the first phase encoding gradient pulse, the gradient field at ky= -127 is set to be high on the left and low on the right in the phase encoding direction (Ky), and the gradient strength is the maximum value; the gradient field strength is gradually reduced after that, the high-low direction of the gradient field is kept unchanged; gradient strength at ky=0 is 0; subsequently, the gradient fields at the respective positions are set to be low on the left and high on the right, and the gradient strength is gradually increased. Referring to fig. 4 (b), for the second phase encoding gradient pulse, the gradient field at ky= +128 is set to be low to high in the left and high in the phase encoding direction (Ky), and the gradient strength is the maximum value; the gradient field strength is gradually reduced after that, the high-low direction of the gradient field is kept unchanged; gradient strength at ky=0 is 0; subsequently, the gradient fields at the respective positions are set to be high on the left and low on the right, and the gradient strength is gradually increased.
Illustratively, acquiring two partial samples of K-space data of a scan object includes: applying a first scanning sequence to the scanning object to obtain a first part of K space data, wherein the first scanning sequence comprises a first phase encoding gradient pulse and a first radio frequency excitation pulse; applying a second scanning sequence to the scanning object to obtain second part of K space data, wherein the second scanning sequence comprises a second phase encoding gradient pulse and a second radio frequency excitation pulse, the field intensity change of the second phase encoding gradient pulse is the same as that of the first phase encoding gradient pulse, the gradient direction is opposite, and the flip angle of the second radio frequency excitation pulse is smaller than that of the first radio frequency excitation pulse; and inputting the second part of K space data into the trained neural network to obtain corrected second part of K space data.
In particular, under the action of the radio frequency pulses, the macroscopic magnetization vector of the tissue will deviate from the equilibrium state by an angle called the deflection angle or excitation angle. The angle of deflection of the macroscopic magnetization vector depends on the energy of the radio frequency pulse, the greater the energy the greater the angle of deflection. However, considering that during a magnetic resonance scan, the rf pulses applied during the execution of the scan sequence may have a certain energy deposition in the scanned object, the energy absorbed by the human body per unit time in medical scanning is set to a strict limit. In the embodiment of the invention, the radio frequency excitation pulse with small flip angle is applied during the acquisition of the second part K space data, and the corrected second part K space data is acquired by using the neural network, so that the accumulation of radio frequency energy in a scanned object in the magnetic resonance scanning process can be effectively reduced, and the scanning safety is improved.
Illustratively, the trained neural network used to obtain the corrected second portion of K-space data is obtained by: acquiring a plurality of groups of first sample data and second sample data, wherein the flip angle of a radio frequency pulse corresponding to the first sample data acquisition is equal to the flip angle of a first radio frequency excitation pulse, the gradient coding comprises a second phase coding gradient pulse, and the first sample data is used as gold standard data; the flip angle of the radio frequency pulse corresponding to the second sample data acquisition is equal to the flip angle of the second radio frequency excitation pulse, the gradient coding comprises a second phase coding gradient pulse, and the second sample data is used as input data; and inputting a plurality of groups of first sample data and second sample data into the neural network for training to determine each node parameter of the neural network, thereby obtaining the trained neural network.
Illustratively, acquiring two partial samples of K-space data of a scan object includes: and carrying out magnetic resonance scanning on a scanning object by adopting a preset magnetic resonance echo sequence to obtain two part sampling K space data.
The preset magnetic resonance echo sequence is a pre-designed magnetic resonance scanning sequence, which may be a (fast) spin echo sequence, a planar echo sequence, a gradient echo sequence or the like.
Specifically, there is an echo train length parameter in the echo train, which refers to the number of echoes acquired with a 180 pulse after a 90 pulse. The longer the echo chain length (the greater the number of code lines in K-space), the more T2 decays and the more blurred the final reconstructed image. Therefore, when performing magnetic resonance scanning using an echo sequence (particularly, a long echo train sequence), the echo train length is shortened to perform partial sampling scanning on a scanning object in order to improve image clarity. The K space data required in the image reconstruction method of the embodiment of the invention is the K space data of partial sampling, so that the image reconstruction method of the embodiment of the invention can shorten the echo chain length of the preset magnetic resonance echo sequence, further improve the signal-to-noise ratio of the image and effectively reduce the problem of image blurring.
When two partial samples of K-space data are mirror images distributed in K-space as shown in fig. 2 (a) and obtained by real-time scanning with the fast spin echo sequence FSE, FES sequences corresponding to k_a and k_b are shown in fig. 5 (a) and 5 (B), respectively. It should be noted that the acquisition of two partial samples of K-space data of different K-space distribution can be achieved by varying the magnitude of the encoded gradient in the scan sequence.
Fig. 5 (a) is a schematic diagram of a first scan sequence in an embodiment of the present application. The sequence includes a diffusion module (disffusion module) and a first imaging sequence module. The first imaging sequence module in this embodiment is exemplified by a first Fast Spin Echo (FSE) module. Of course, the type of the first imaging sequence module is not particularly limited in this application, and in other embodiments, other sequence types with long echo chains may be used.
Wherein: RF means radio frequency pulse; g SS Representing a slice-selective gradient pulse; g PE Representing the phase encoding gradient; g RO Representing frequency coding; echo represents the acquired magnetic resonance signals acquired within the acquisition window. For a dispersion module, firstly, an excitation radio frequency pulse with a flip angle alpha is applied to a radio frequency signal RF; then, a stimulus at the alphaThe radio frequency pulse is transmitted and then a refocusing radio frequency pulse is applied, the flip angle of which may be 180 degrees or other values. At G SS A corresponding layer selection gradient is applied in the direction. Meanwhile, at G SS 、G PE And G RO The diffusion sensitive gradients are applied symmetrically on each side of at least one of the directions corresponding to the refocusing rf pulses. In FIG. 5 (a), at G SS 、G PE And G RO A diffusion sensitive gradient is applied in all three directions, wherein, at G SS In the direction of diffusion sensitivity gradient G SS,D1 And G SS,D2 The method comprises the steps of carrying out a first treatment on the surface of the At G PE In the direction of diffusion sensitivity gradient G PE,D1 And G PE,D2 The method comprises the steps of carrying out a first treatment on the surface of the At G RO In the direction of diffusion sensitivity gradient G RO,D1 And G RO,D2 . After the radio frequency pulse, the gradients in the corresponding directions and the diffusion sensitivity gradient are applied in the diffusion imaging pulse sequence, the diffusion sensitivity gradient G before the rephasing radio frequency pulse is heavy for static water molecules due to low diffusion capacity SS,D1 、G PE,D1 And/or G RO,D1 The resulting proton spin-diffusion phase is subject to a diffusion-sensitive gradient G after refocusing radio frequency pulses SS,D2 、G PE,D2 And/or G RO,D2 Complete refocusing, with signal strength unaffected; for moving water molecules, the diffusion sensitivity gradient G before the refocusing radio frequency pulse is high due to the strong diffusion capability SS,D1 、G PE,D1 And/or G RO,D1 The resulting proton spin-diffusion phase is away from the original position and thus cannot be separated by the diffusion sensitive gradient G after the refocusing radio frequency pulse SS,D2 、G PE,D2 And/or G RO,D2 Refocusing, resulting in a decay in its signal strength with the diffusion phase.
The first FSE module after the dispersion module specifically comprises: a 90 ° rf pulse followed by four 180 ° pulses; at G SS Correspondingly applying corresponding layer-selecting gradient pulse in the direction G RO Applying corresponding frequency coding gradient pulse in the direction, G PE Corresponding first phase encoding gradient pulses are applied in the direction, and 4 spin echoes (Echo 1) are generated after four 180 DEG pulses are excitedEcho 4) that are used to form the first portion of K-space data.
Please refer to fig. 5 (b), which is a diagram illustrating a first scan sequence in an embodiment of the present application. The sequence includes a diffusion module (disffusion module) and a second imaging sequence module. The first imaging sequence module in this embodiment is exemplified by a second Fast Spin Echo (FSE) module. Of course, the type of the second imaging sequence module is not particularly limited in this application, and in other embodiments, other sequence types with long echo chains may be used.
Wherein: RF means radio frequency pulse; g SS Representing a slice-selective gradient pulse; g PE Representing the phase encoding gradient; g RO Representing frequency coding; echo represents the acquired magnetic resonance signals acquired within the acquisition window. For a dispersion module, firstly, an excitation radio frequency pulse with a flip angle alpha is applied to a radio frequency signal RF; a refocusing radio frequency pulse is then applied after the excitation radio frequency pulse of alpha, the flip angle of which may be 180 degrees or other values. At G SS A corresponding layer selection gradient is applied in the direction. Meanwhile, at G SS 、G PE And G RO The diffusion sensitive gradients are applied symmetrically on each side of at least one of the directions corresponding to the refocusing rf pulses. In FIG. 5 (b), at G SS 、G PE And G RO A diffusion sensitive gradient is applied in all three directions, wherein, at G SS In the direction of diffusion sensitivity gradient G SS,D3 And G SS,D4 The method comprises the steps of carrying out a first treatment on the surface of the At G PE In the direction of diffusion sensitivity gradient G PE,D3 And G PE,D4 The method comprises the steps of carrying out a first treatment on the surface of the At G RO In the direction of diffusion sensitivity gradient G RO,D3 And G RO,D4 . After the radio frequency pulse, the gradients in the corresponding directions and the diffusion sensitivity gradient are applied in the diffusion imaging pulse sequence, the diffusion sensitivity gradient G before the rephasing radio frequency pulse is heavy for static water molecules due to low diffusion capacity SS,D3 、G PE,D3 And/or G RO,D3 The resulting proton spin-diffusion phase is subject to a diffusion-sensitive gradient G after refocusing radio frequency pulses SS,D4 、G PE,D4 And/or G RO,D4 Complete refocusing, with signal strength unaffected; for moving water molecules, the diffusion sensitivity gradient G before the refocusing radio frequency pulse is high due to the strong diffusion capability SS,D3 、G PE,D3 And/or G RO,D3 The resulting proton spin-diffusion phase is away from the original position and thus cannot be separated by the diffusion sensitive gradient G after the refocusing radio frequency pulse SS,D4 、G PE,D4 And/or G RO,D4 Refocusing, resulting in a decay in its signal strength with the diffusion phase.
The second FSE module after the dispersion module, in particular comprises: a 90 ° rf pulse followed by four 180 ° pulses; at G SS Correspondingly applying corresponding layer-selecting gradient pulse in the direction G RO Applying corresponding frequency coding gradient pulse in the direction, G PE Corresponding second phase encoding gradient pulses are applied in the direction, and after excitation of four 180 ° pulses, 4 spin echoes (Echo 1-Echo 4) are generated, which are used to form the second part of K-space data.
Illustratively, each partial sample K-space data covers a central region of the respective K-space. In particular, magnetic resonance scan sequences are very sensitive to motion, and it is often necessary to add a motion correction module to the scan sequence during the scan. Considering that respiratory motion of a human body generally brings about a low-frequency phase, a motion correction module for suppressing motion artifacts added to a scanning sequence in the related art is generally used for correcting image artifacts by repeatedly acquiring K-space signals passing through a K-space center and surrounding low-frequency regions, and an additional scanning process is added outside a formal scanning. However, in the image reconstruction process of the embodiment of the invention, two K-space data are provided, and the repeated acquisition of the central part of the K-space can be ensured as long as the two K-space data are set to cover the central area of the corresponding K-space. And by sampling the phase correction of the K space for the two parts (see the following description), the signal difference of the repeated signal acquisition part can be removed, so that the problem of inconsistent phase of the K space caused by motion can be eliminated, image artifacts are reduced, and the signal to noise ratio of the image is further improved.
The first partial K-space data and/or the second partial K-space data are filled in a partial echo acquisition manner. Specifically, referring to fig. 6, taking the partial echo acquisition mode of the first partial K-space data k_a as an example, for each phase encoding position, only half of the echo of the phase encoding position is acquired, so that the required time is shortened, and the echo time is shortened; as shown in fig. 6, the K space is filled only by half more in the frequency encoding direction (in the first diagram of fig. 6, along the frequency encoding direction, the solid line part is the area filled with data, and the broken line is the area not filled with data); further, for each phase encoding position, the remainder in the frequency encoding direction is analog-filled according to the symmetry principle. The advantage of this arrangement is that, due to the use of the partial echo technique, the echo Time (TE) can be shortened, so that the repetition Time (TR) can be shortened on the premise of ensuring the same number of layers, thereby indirectly shortening the acquisition time, or the number of acquired layers can be increased by maintaining the original TR, which is helpful for improving the acquisition speed and reducing the magnetic sensitivity artifacts.
S220, determining a phase correction value according to the two partial sampling K space data.
Specifically, in the process of acquiring the partial sampling K-space data, because the system eddy current and respiratory motion and other factors can cause a phase difference between overlapping areas of the two partial sampling K-space data, and the existence of the phase difference can reduce the signal to noise ratio of the image, in this embodiment, the two partial sampling K-space data needs to be subjected to phase correction first. In the related art, the estimated phase is usually introduced, but in this embodiment, the estimated phase is introduced to increase the error of the subsequent image reconstruction and reduce the signal-to-noise ratio of the image, so that in this embodiment, a complex phase correction algorithm is not adopted, but a phase difference between two partial sampling K-space data is calculated, and then the phase difference is used to correct the phase of one partial sampling K-space data, so as to obtain two partial sampling K-space data only including the normal phase difference caused by the tissue constitution of the scanned object.
In the implementation, first, partial fourier image reconstruction is performed based on overlapping regions of two partial sampling K spaces, respectively, to obtain two reconstructed images. For example, zero padding is performed on regions outside the overlapping region of the two partial sampling K-space data, so as to obtain corresponding zero-padded K-space data, and then inverse fourier transform is performed on the two zero-padded K-space data, so as to obtain two reconstructed images. Then, a phase difference corresponding to each pixel is acquired for the two reconstructed images, and the phase difference is a phase correction value.
S230, performing phase correction on any part of the sampling K space data according to the phase correction value so as to update the part of the sampling K space data subjected to the phase correction.
Specifically, firstly reconstructing any part of sampling K space data to obtain a reconstructed image; then subtracting the phase correction value from each pixel in one of the two reconstructed images to obtain a phase correction image of the reconstructed image; then, the phase correction image is converted into K space through inverse Fourier transform, so that partial sampling K space data after phase correction can be obtained, and the partial sampling K space data after phase correction is utilized to replace corresponding partial sampling K space data before phase correction, so that two partial sampling K space data which can be used for weighted combination can be obtained.
S240, carrying out weighted combination processing on the overlapping area of the two partial sampling K space data according to preset weights to generate overlapping area combined K space data.
The preset weight is a preset weight value used for weighting and combining. Illustratively, the preset weights include a first preset weight and a second preset weight that are applied to the two partial sampling K-space data, respectively, and the sum of the first preset weight and the second preset weight is 1. Specifically, in this embodiment, there are two sets of preset weight values, one set acts on one part of sampling K-space data, and the other set acts on the other part of sampling K-space data, and the sum of two preset weight values corresponding to the same PE encoding line in the two sets of preset weight values is 1. The aim of the arrangement is to ensure that the signal value of the PE line in the K space is more gentle, thereby ensuring the uniformity of the K space data and further improving the signal-to-noise ratio of the image reconstructed by the subsequent image. Illustratively, for each K-space encoding line within the overlapping region, the first preset weight and the second preset weight are linearly varied, and the direction of the linear variation is opposite. Specifically, on the basis that the sum of the first preset weight and the second preset weight is 1, the numerical value changes of the two groups of preset weight values are further set to be linear changes, and the changing directions in the phase encoding directions are opposite. Taking fig. 2 (a) as an example, the data lines with phase codes of ky= -19 to +20 in the first part K space data k_a and the data lines with phase codes of ky= -19 to +20 in the second part K space data k_b are the same data lines with phase codes in the two parts K space data. For the first preset weight corresponding to K_A, linearly and gradually reducing from 1 to 0 according to the sequence of Ky= -19 to +20; and for the second preset weight corresponding to k_b, it gradually increases from 0 to 1 in the order of ky= -19 to +20. The purpose of this arrangement is to further ensure the smoothness of the K-space data.
Specifically, for the overlapping region, weighting summation of the K space data is performed on the PE encoding lines according to preset weights, so that weighted and combined K space data corresponding to the overlapping region, namely, combined K space data of the overlapping region, can be obtained.
S250, combining the K space data corresponding to the non-overlapping region in each part of the sampling K space data and the overlapping region to generate full sampling K space data.
Specifically, the data of the non-overlapping area in the two partial sampling K space data are combined into the K space data of the overlapping area obtained above, so that the whole K space can be filled, and the full sampling K space data of which the data filled in the K space are all real scanning data obtained by actual scanning is generated.
And S260, performing image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanned object.
Referring to fig. 7, for head scan data obtained by fse_dwi sequence scanning, an image reconstruction method (abbreviated as dephasing partial fourier technique) and a zero-filled partial fourier method (abbreviated as zero-filled partial fourier technique) in the embodiment of the present invention are used to reconstruct images, respectively, so as to obtain corresponding magnetic resonance reconstructed images. As can be seen from the results of the reconstructed image, the degree of blurring of the reconstructed image (left two images) obtained by the dephasing partial Fourier technique of the embodiment of the invention is significantly lower than that of the reconstructed image obtained by the zero-padding partial Fourier technique (right two images).
According to the technical scheme of the embodiment, the phase correction value is determined by sampling K space data according to two parts; and performing phase correction on any part of the sampling K space data according to the phase correction value to update the part of the sampling K space data subjected to the phase correction. On the premise of not introducing estimated phases, the correction of other abnormal phase differences between the two partial sampling K space data except for normal phase differences caused by the tissue constitution of the scanning object is realized, so that the difference between the two partial sampling K space data is further reduced, the difficulty of subsequent data merging is further reduced, the accuracy of the full sampling K space data is improved, and the signal to noise ratio of the magnetic resonance reconstructed image is further improved. The overlapping area of the two partial sampling K space data is subjected to weighting and merging processing according to preset weights, so that overlapping area merging K space data is generated; and combining the K space data corresponding to the non-overlapping region in each part of the sampling K space data and the overlapping region to generate full sampling K space data. The high-efficiency fusion of the two partial sampling K space data is realized, and the accuracy of the full sampling K space data is further improved, so that the signal-to-noise ratio of the magnetic resonance reconstructed image is further improved.
Example III
The present embodiment provides a magnetic resonance image reconstruction apparatus, referring to fig. 8, specifically including:
a partial sampling K-space data obtaining module 810, configured to obtain two partial sampling K-space data of a scan object, where filling directions of the two partial sampling K-space data in a phase encoding direction are opposite, and the two partial sampling K-space data include data lines with the same phase encoding in a preset proportion;
the full sampling K-space data generating module 820 is configured to perform weighted combination processing on the two partial sampling K-space data, and generate full sampling K-space data corresponding to the scan object;
the image reconstruction module 830 is configured to perform image reconstruction according to the fully sampled K-space data, and generate a magnetic resonance reconstructed image corresponding to the scan object.
Optionally, the full sampling K-space data generating module 820 is specifically configured to:
according to preset weights, carrying out weighted combination processing on overlapping areas of the two partial sampling K space data to generate overlapping area combined K space data;
and combining the K space data corresponding to the non-overlapping region in each part of the sampling K space data and the overlapping region to generate full sampling K space data.
Optionally, on the basis of the above apparatus, the apparatus further includes a phase correction module for:
Before weighting and combining the two partial sampling K space data to generate full sampling K space data corresponding to a scanning object, determining a phase correction value according to the two partial sampling K space data;
and performing phase correction on any part of the sampling K space data according to the phase correction value to update the part of the sampling K space data subjected to the phase correction.
Optionally, the partial sampling K-space data acquisition module 810 is specifically configured to:
applying a first scan sequence to the scan object to obtain a first portion of K-space data, wherein the first scan sequence comprises a first phase encoding gradient pulse;
and applying a second scanning sequence to the scanning object to obtain second part of K space data, wherein the second scanning sequence comprises second phase encoding gradient pulses, and the second phase encoding gradient pulses have the same field intensity change as the first phase encoding gradient pulses and have opposite gradient directions.
Optionally, the partial sampling K-space data acquisition module 810 is specifically configured to:
applying a first scanning sequence to the scanning object to obtain a first part of K space data, wherein the first scanning sequence comprises a first phase encoding gradient pulse and a first radio frequency excitation pulse;
Applying a second scanning sequence to the scanning object to obtain second part of K space data, wherein the second scanning sequence comprises a second phase encoding gradient pulse and a second radio frequency excitation pulse, the field intensity change of the second phase encoding gradient pulse is the same as that of the first phase encoding gradient pulse, the gradient direction is opposite, and the flip angle of the second radio frequency excitation pulse is smaller than that of the first radio frequency excitation pulse;
and inputting the second part of K space data into the trained neural network to obtain corrected second part of K space data.
Optionally, the first scanning sequence and/or the second scanning sequence further comprise a diffusion module, the diffusion module comprising:
an excitation rf pulse and a subsequently applied refocusing rf pulse, which applies a diffusion-sensitive gradient symmetrically on both sides.
Optionally, the partial sampling K-space data acquisition module 810 is specifically configured to:
and carrying out magnetic resonance scanning on a scanning object by adopting a preset magnetic resonance echo sequence to obtain two part sampling K space data, wherein each part of sampling K space data covers the central area of the corresponding K space.
According to the magnetic resonance image reconstruction device, the method and the system for generating the full-sampling K space data of the scanned object by weighting and combining the two actually-acquired partial-sampling K space data are realized, so that the magnetic resonance reconstruction image is obtained, the problem of image blurring caused by the fact that the image reconstruction algorithm of the partial Fourier technology is used for estimating the missing K space data is solved, the magnetic resonance image reconstruction algorithm is simplified, and the reconstruction efficiency and the image signal to noise ratio of the magnetic resonance image are improved.
The magnetic resonance image reconstruction device provided by the embodiment of the invention can execute the magnetic resonance image reconstruction method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
It should be noted that, in the embodiment of the magnetic resonance image reconstruction device, each unit and module included are only divided according to the functional logic, but are not limited to the above-mentioned division, so long as the corresponding functions can be implemented; in addition, the specific names of the functional units are also only for distinguishing from each other, and are not used to limit the protection scope of the present invention.
Example IV
Referring to fig. 9a, the present embodiment provides an apparatus 900 comprising: a console 91 and a scanner 92, wherein the console 91 can receive or transmit information from/to the scanner 92. According to some embodiments of the present application, the console 91 may receive commands provided by, for example, a user, and adjust the superconducting magnet, the gradient coil, and the radio frequency coil to take an image of the object of interest according to the received commands. The controller may process different kinds of information received from different modules.
In this embodiment, scanner 92 may include superconducting magnets, gradient coils, and radio frequency coils, among other major components. In particular, the superconducting magnet may surround a bore that forms a receiving space, and the superconducting magnet may create a static magnetic field B0 during an imaging procedure. Gradient coils are disposed within the bore and may include X, Y and/or Z coils (not shown). In some embodiments, the Z coil may be designed based on a circular (Maxwell) coil, while the X and Y coils may be designed based on a saddle (Golay) coil configuration. The three sets of coils may perform a gradient pulse sequence to generate three different magnetic fields that are used for position encoding, e.g. the gradient coils may perform a gradient pulse sequence to form a gradient field in the readout direction, a gradient field in the phase encoding direction and a gradient field in the slice selection direction, respectively, to readout encode, phase encode and slice selection direction encode the signal. The radio frequency coils may comprise a radio frequency transmit coil and a radio frequency receive coil, wherein the radio frequency transmit coil performs a sequence of radio frequency pulses for transmitting RF signals to/from the object of interest; the radio frequency receive coil is for receiving magnetic resonance signals excited by the object of interest. RF coils can be classified into volume coils and local coils. In some embodiments of the present application, the volume coil may include a birdcage coil, a transverse electromagnetic coil, a surface coil, a saddle coil, or the like. In some embodiments of the present application, the local coil may comprise a birdcage coil, a solenoid coil, a saddle coil, a flexible coil, or the like.
Referring to fig. 9b, the console 91 includes: a processor 920, a storage device 910, an input device 930, and an output device 940; the number of processors 920 in the console 91 may be one or more, one processor 920 being taken as an example in fig. 9 b; the processor 920, the storage device 910, the input device 930, and the output device 940 in the console 91 may be connected by a bus or otherwise, for example, by a bus 950 in fig. 9 b.
The storage device 910 is used as a computer readable storage medium, and may be used to store a software program, a computer executable program, and modules, such as program instructions/modules corresponding to a magnetic resonance image reconstruction method in an embodiment of the present invention (for example, a partially sampled K-space data acquisition module, a fully sampled K-space data generation module, and an image reconstruction module in a magnetic resonance image reconstruction device).
The storage device 910 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, at least one application program required for a function; the storage data area may store data created according to the use of the terminal, etc. In addition, the storage 910 may include high-speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid-state storage device. In some examples, the storage 910 may further include memory remotely located relative to the processor 920, which may be connected to the device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input means 930 may be used to receive input numeric or character information and to generate key signal inputs related to user settings and function control of the device. The output device 940 may include a display device such as a display screen.
In one embodiment, console 91 may control scanner 92 to apply a first scan sequence to the scan object to obtain a first portion of K-space data, wherein the first scan sequence includes a first phaseThe bit encodes a gradient pulse. As shown in fig. 5a, the first scanning sequence in the embodiment of the present application includes: a dispersion module (disffusion module) and a first Fast Spin Echo (FSE) module. RF in fig. 5a represents the radio frequency pulses emitted by the radio frequency coil; g SS A slice-selective gradient pulse representing the formation of gradient coils; g PE Representing the phase encoding gradient formed by the gradient coils; g RO Frequency encoding representing gradient coil formation; echo represents the acquired magnetic resonance signals acquired within the acquisition window. The first phase encoding gradient pulse is G in FIG. 5a PE Gradient pulses applied in the direction and four echo signals (echo 1-echo 4) are acquired after the application of the first scanning sequence.
The console 91 may control the scanner 92 to apply a second scan sequence to the scan object, wherein the second scan sequence comprises second phase encoding gradient pulses having the same field strength variation as the first phase encoding gradient pulses and opposite gradient directions, to obtain a second portion of K-space data. Specifically, as shown in fig. 5b, the second scan sequence in the embodiment of the present application includes: a dispersion module (disffusion module) and a second Fast Spin Echo (FSE) module. The diffusion module of the second scanning sequence is the same as the radio frequency pulse and gradient pulse contained in the diffusion module of the first scanning sequence. The second FSE module comprises 90-degree radio frequency pulse and four 180-degree radio frequency pulses which are applied subsequently, and gradient G is selected SS Phase encoding gradient G PE And a frequency encoding gradient G RO Corresponding gradient pulses are applied in the directions respectively. The second phase encoding gradient pulse is G in FIG. 5b PE The gradient pulses applied in the direction are of opposite polarity to the gradient of the first phase encoding gradient pulse at the corresponding timing, i.e., the gradient coils of scanner 92 each generate phase gradient pulses of opposite polarity at different time periods.
The console 91 can control the scanner 92 to perform weighting and combining processing on the two partial sampling K-space data to generate full sampling K-space data corresponding to the scanning object; and, the console 91 may control the scanner 92 to reconstruct an image according to the fully sampled K-space data, and generate a magnetic resonance reconstructed image corresponding to the scanned object.
Example five
The present embodiment provides a storage medium containing computer executable instructions which, when executed by a computer processor, are for performing a magnetic resonance image reconstruction method comprising:
acquiring two partial sampling K space data of a scanning object, wherein the filling directions of the two partial sampling K space data in the phase coding direction are opposite, and the two partial sampling K space data comprise data lines with the same phase coding in a preset proportion;
Weighting and combining the two partial sampling K space data to generate full sampling K space data corresponding to the scanning object;
and carrying out image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanned object.
Of course, the storage medium containing the computer executable instructions provided in the embodiments of the present invention is not limited to the above method operations, and may also perform the related operations in the magnetic resonance image reconstruction method provided in any embodiment of the present invention.
From the above description of embodiments, it will be clear to a person skilled in the art that the present invention may be implemented by means of software and necessary general purpose hardware, but of course also by means of hardware, although in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, etc., and include several instructions for causing a device (which may be a personal computer, a server, or a network device, etc.) to perform the magnetic resonance image reconstruction method provided by the embodiments of the present invention.
Note that the above is only a preferred embodiment of the present invention and the technical principle applied. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, while the invention has been described in connection with the above embodiments, the invention is not limited to the embodiments, but may be embodied in many other equivalent forms without departing from the spirit or scope of the invention, which is set forth in the following claims.
Claims (9)
1. A method of magnetic resonance image reconstruction, comprising:
acquiring two partial sampling K space data of a scanning object, wherein the filling directions of the two partial sampling K space data in the phase coding direction are opposite, the two partial sampling K space data comprise data lines with the same phase coding of a preset proportion, the two partial sampling K space data are obtained by exciting the two K spaces through FSE sequences respectively, and the filling directions of the data lines belonging to one partial sampling K space data in the frequency coding direction are the same; weighting and combining the two partial sampling K space data to generate full sampling K space data corresponding to the scanning object;
Performing image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanning object;
the method further comprises the following steps before the weighted combination processing of the two partial sampling K space data:
performing partial Fourier image reconstruction based on the overlapping areas of the two partial sampling K spaces respectively to obtain two reconstructed images;
respectively acquiring phase differences corresponding to each pixel for the two reconstructed images, and taking the phase differences as phase correction values;
and carrying out phase correction on any part of the sampling K space data according to the phase correction value so as to update the part of the sampling K space data.
2. The method of claim 1, wherein the weighting and combining the two partial sample K-space data to generate the full sample K-space data corresponding to the scan object comprises:
according to preset weights, carrying out weighted combination processing on the overlapping areas of the two partial sampling K space data to generate overlapping area combined K space data;
and combining the K space data corresponding to the non-overlapping region in each piece of the partial sampling K space data and the overlapping region to generate the full sampling K space data.
3. The method of claim 1, wherein the acquiring two partial sample K-space data of the scan object comprises:
applying a first scanning sequence to the scanning object to obtain a first part of K space data, wherein the first scanning sequence comprises a first phase encoding gradient pulse;
and applying a second scanning sequence to the scanning object to obtain second part of K space data, wherein the second scanning sequence comprises second phase encoding gradient pulses, and the second phase encoding gradient pulses have the same field intensity change as the first phase encoding gradient pulses and have opposite gradient directions.
4. The method of claim 1, wherein the acquiring two partial sample K-space data of the scan object comprises:
applying a first scanning sequence to the scanning object to obtain a first part of K space data, wherein the first scanning sequence comprises a first phase encoding gradient pulse and a first radio frequency excitation pulse;
applying a second scanning sequence to the scanning object to obtain second part of K space data, wherein the second scanning sequence comprises a second phase encoding gradient pulse and a second radio frequency excitation pulse, the second phase encoding gradient pulse has the same field intensity change as the first phase encoding gradient pulse and opposite gradient directions, and the flip angle of the second radio frequency excitation pulse is smaller than that of the first radio frequency excitation pulse;
And inputting the second part of K space data into the trained neural network to obtain corrected second part of K space data.
5. The method according to claim 3 or 4, wherein the first and/or second scan sequence further comprises a diffusion module comprising:
an excitation rf pulse and a subsequently applied refocusing rf pulse, which applies a diffusion-sensitive gradient symmetrically on both sides.
6. The method of claim 1, wherein the acquiring two partial sample K-space data of the scan object comprises:
and carrying out magnetic resonance scanning on the scanning object by adopting a preset magnetic resonance echo sequence to obtain two partial sampling K space data, wherein each partial sampling K space data covers the central area of the corresponding K space.
7. A magnetic resonance image reconstruction apparatus, comprising:
the partial sampling K space data acquisition module is used for acquiring two partial sampling K space data of a scanning object, wherein the filling directions of the two partial sampling K space data in the phase coding direction are opposite, and the two partial sampling K space data comprise data lines with the same phase coding of a preset proportion; the full sampling K space data generation module is used for carrying out weighting and combining processing on the two partial sampling K space data to generate full sampling K space data corresponding to the scanning object, the two partial sampling K space data are respectively obtained through FSE sequence scanning, and the filling directions of data lines belonging to one partial sampling K space data in the frequency coding direction are the same;
The image reconstruction module is used for carrying out image reconstruction according to the fully sampled K space data and generating a magnetic resonance reconstruction image corresponding to the scanning object;
the method further comprises the following steps before the weighted combination processing of the two partial sampling K space data:
performing partial Fourier image reconstruction based on the overlapping areas of the two partial sampling K spaces respectively to obtain two reconstructed images;
respectively acquiring phase differences corresponding to each pixel for the two reconstructed images, and taking the phase differences as phase correction values;
and carrying out phase correction on any part of the sampling K space data according to the phase correction value so as to update the part of the sampling K space data.
8. An apparatus, the apparatus comprising:
a scanner having an accommodation space, and configured to perform scanning on a scanning object in the accommodation space;
one or more processors;
storage means for storing one or more programs,
when the one or more programs are executed by the one or more processors, the one or more processors are caused to control the scanner to implement:
applying a first scanning sequence to the scanning object to obtain a first part of K space data, wherein the first scanning sequence comprises a first phase encoding gradient pulse;
Applying a second scanning sequence to the scanning object to obtain second part of K space data, wherein the second scanning sequence comprises second phase encoding gradient pulses, and the gradient directions of the second phase encoding gradient pulses and the first phase encoding gradient pulses are opposite;
the first part of K space data and the second part of K space data are weighted and combined to generate full-sampling K space data corresponding to the scanning object, the first part of K space data and the second part of K space data are respectively obtained through FSE sequence scanning, and the filling directions of data lines belonging to one part of sampling K space data in the frequency coding direction are the same;
performing image reconstruction according to the fully sampled K space data to generate a magnetic resonance reconstruction image corresponding to the scanning object;
before the weighted combination processing is performed on the first part of K space data and the second part of K space data, the method further comprises the following steps:
performing partial Fourier image reconstruction based on the overlapping areas of the two partial sampling K spaces respectively to obtain two reconstructed images;
respectively acquiring phase differences corresponding to each pixel for the two reconstructed images, and taking the phase differences as phase correction values;
And carrying out phase correction on any part of the sampling K space data according to the phase correction value so as to update the part of the sampling K space data.
9. The apparatus of claim 8, wherein the processor further controls the scanner implementation to:
before the scanner applies the first phase encoding gradient pulse, applying an excitation radio frequency pulse and a rephasing radio frequency pulse after the excitation radio frequency pulse, wherein the two sides of the rephasing radio frequency pulse symmetrically apply diffusion sensitivity gradients;
and/or applying an excitation radio frequency pulse and a refocusing radio frequency pulse after the excitation radio frequency pulse before the scanner applies the second phase encoding gradient pulse, the refocusing radio frequency pulse symmetrically applying a diffusion sensitive gradient on both sides.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010169908.0A CN113391251B (en) | 2020-03-12 | 2020-03-12 | Magnetic resonance image reconstruction method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010169908.0A CN113391251B (en) | 2020-03-12 | 2020-03-12 | Magnetic resonance image reconstruction method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113391251A CN113391251A (en) | 2021-09-14 |
| CN113391251B true CN113391251B (en) | 2023-05-26 |
Family
ID=77615606
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010169908.0A Active CN113391251B (en) | 2020-03-12 | 2020-03-12 | Magnetic resonance image reconstruction method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113391251B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114325528B (en) * | 2021-12-25 | 2024-01-02 | 沈阳工业大学 | Magnetic resonance imaging method and related equipment |
| CN115656900B (en) * | 2022-11-02 | 2023-06-23 | 佛山瑞加图医疗科技有限公司 | Method and device for reducing influence of system error on magnetic resonance imaging |
| CN116012472A (en) * | 2022-12-02 | 2023-04-25 | 中国科学院深圳先进技术研究院 | Quick magnetic resonance imaging method and device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1803092A (en) * | 2005-11-29 | 2006-07-19 | 东南大学 | Artifact correction method based on even marker in plane echo imaging technique |
| CN101172036A (en) * | 2006-11-02 | 2008-05-07 | 西门子公司 | Method for phase correction |
| CN102440778A (en) * | 2010-09-30 | 2012-05-09 | 株式会社东芝 | Magnetic resonance imaging apparatus |
| CN102772210A (en) * | 2011-05-10 | 2012-11-14 | 西门子公司 | Diffusion-weighted magnetic resonance imaging |
| CN104013405A (en) * | 2014-06-09 | 2014-09-03 | 深圳先进技术研究院 | Dynamic myocardium activity detection method and system |
| CN104614767A (en) * | 2014-12-11 | 2015-05-13 | 中国石油大学(华东) | Method for correcting seismic time-varying wavelet phase based on sectional prolongation |
| CN104918545A (en) * | 2013-02-19 | 2015-09-16 | 株式会社东芝 | Magnetic resonance imaging device |
| CN108697369A (en) * | 2016-02-17 | 2018-10-23 | 三星电子株式会社 | MR imaging apparatus and its method |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US4740748A (en) * | 1986-12-03 | 1988-04-26 | Advanced Nmr Systems, Inc. | Method of high-speed magnetic resonance imaging |
| CN101153896A (en) * | 2006-09-29 | 2008-04-02 | 西门子（中国）有限公司 | Image reconstruction method for echo wave plane imaging sequence |
| US8085043B2 (en) * | 2009-04-03 | 2011-12-27 | Siemens Aktiengesellschaft | Method for image data acquisition with a magnetic resonance device |
| DE102013201616B3 (en) * | 2013-01-31 | 2014-07-17 | Siemens Aktiengesellschaft | TSE-based MR multilayer excitation insensitive to local B0 field variations |
| CN104714199B (en) * | 2013-12-17 | 2018-04-24 | 西门子(深圳)磁共振有限公司 | A kind of MR imaging method and device |
| CN105738846B (en) * | 2014-12-12 | 2019-01-25 | 西门子(深圳)磁共振有限公司 | K space data acquisition method and its MR imaging method |
| US20170016972A1 (en) * | 2015-07-13 | 2017-01-19 | Siemens Medical Solutions Usa, Inc. | Fast Prospective Motion Correction For MR Imaging |
| CN106361336B (en) * | 2015-07-23 | 2020-12-04 | 上海联影医疗科技股份有限公司 | Magnetic resonance imaging method and system |
| CN109658471B (en) * | 2018-12-20 | 2023-07-25 | 上海联影医疗科技股份有限公司 | Medical image reconstruction method and system |
| CN109597012B (en) * | 2018-12-24 | 2020-08-04 | 厦门大学 | Single-scanning space-time coding imaging reconstruction method based on residual error network |
| CN110346743B (en) * | 2019-07-22 | 2021-09-14 | 上海东软医疗科技有限公司 | Magnetic resonance diffusion weighted imaging method and device |
-
2020
- 2020-03-12 CN CN202010169908.0A patent/CN113391251B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1803092A (en) * | 2005-11-29 | 2006-07-19 | 东南大学 | Artifact correction method based on even marker in plane echo imaging technique |
| CN101172036A (en) * | 2006-11-02 | 2008-05-07 | 西门子公司 | Method for phase correction |
| CN102440778A (en) * | 2010-09-30 | 2012-05-09 | 株式会社东芝 | Magnetic resonance imaging apparatus |
| CN102772210A (en) * | 2011-05-10 | 2012-11-14 | 西门子公司 | Diffusion-weighted magnetic resonance imaging |
| CN104918545A (en) * | 2013-02-19 | 2015-09-16 | 株式会社东芝 | Magnetic resonance imaging device |
| CN104013405A (en) * | 2014-06-09 | 2014-09-03 | 深圳先进技术研究院 | Dynamic myocardium activity detection method and system |
| CN104614767A (en) * | 2014-12-11 | 2015-05-13 | 中国石油大学(华东) | Method for correcting seismic time-varying wavelet phase based on sectional prolongation |
| CN108697369A (en) * | 2016-02-17 | 2018-10-23 | 三星电子株式会社 | MR imaging apparatus and its method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113391251A (en) | 2021-09-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113391251B (en) | Magnetic resonance image reconstruction method, device and equipment | |
| RU2592039C2 (en) | Formation of magnetic resonance images using multiple-point dixon method | |
| US7633291B2 (en) | Magnetic resonance system and method for determining a diffusion-weighted image | |
| CN110244246B (en) | Magnetic resonance imaging method, magnetic resonance imaging apparatus, computer device, and storage medium | |
| JP5127841B2 (en) | Magnetic resonance imaging apparatus and magnetic susceptibility weighted imaging method | |
| US9746539B2 (en) | MR imaging with suppresion of flow artifacts | |
| US10234528B2 (en) | Method and apparatus to correct noise effects in quantitative techniques in magnetic resonance imaging | |
| CN110869790B (en) | MR imaging using star-stack acquisition with variable contrast | |
| JP2012236022A (en) | Method for generating diffusion-weighted magnetic resonance data, magnetic resonance system, and computer-readable storage medium | |
| US10466330B2 (en) | Magnetic resonance diffusion weighted imaging method and apparatus | |
| JPH0956694A (en) | Mr imaging device | |
| CN107076819B (en) | Dixon MR imaging with suppression of flow artifacts | |
| US11327133B2 (en) | Dixon-type water/fat separation MR imaging | |
| Mooiweer et al. | Combining a reduced field of excitation with SENSE‐based parallel imaging for maximum imaging efficiency | |
| CN115639510A (en) | Magnetic resonance imaging method, spectroscopic imaging method, apparatus, device, and storage medium | |
| CN109983358B (en) | PROPELLER MR imaging | |
| US9772390B2 (en) | Magnetic resonance imaging device and method for generating image using same | |
| JP2006501919A (en) | Magnetic resonance method and apparatus | |
| JP7128972B2 (en) | EPI MR image with distortion correction | |
| CN111164444B (en) | Dixon-type water/fat separation MR imaging with improved fat displacement correction | |
| Ponce et al. | Auto‐calibration approach for k–t SENSE | |
| US20230366962A1 (en) | Dixon-type water/fat separation mr imaging | |
| EP4343356A1 (en) | Mr imaging with water/fat/b0 mapping | |
| Daval‐Frérot et al. | Deep learning‐assisted model‐based off‐resonance correction for non‐Cartesian SWI | |
| Yun et al. | High‐resolution fMRI with higher‐order generalized series imaging and parallel imaging techniques (HGS‐parallel) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |