WO2020037814A1

WO2020037814A1 - Equal voxel magnetic resonance diffusion imaging method and apparatus based on multi-plate simultaneous excitation

Info

Publication number: WO2020037814A1
Application number: PCT/CN2018/112353
Authority: WO
Inventors: 戴二鹏; 郭华; 吴宥萱
Original assignee: 清华大学
Priority date: 2018-08-23
Filing date: 2018-10-29
Publication date: 2020-02-27
Also published as: CN109212443B; CN109212443A

Abstract

An equal voxel magnetic resonance diffusion imaging method and apparatus (10) based on multi-plate simultaneous excitation. The method comprises the following steps: step S1: performing excitation on a target to be tested for many times by means of a multi-plate simultaneous excitation pulse, and in the process of each excitation, performing signal acquisition on the target to be tested by means of a multi-channel coil, thereby obtaining k space data acquired by means of down sampling in each excitation; step S2: by a multi-excitation diffusion imaging reconstruction algorithm uniting a k space and an image domain, recovering k space position data which is not acquired in each excitation; and S3, correcting an edge artifact by means of an improved NPEN algorithm, thereby obtaining an imaged image. According to the method, high-resolution images are acquired and a relatively high signal-to-noise ratio can still be kept at the same time; interference of three-dimensional navigation echo errors is reduced, quality and stability of reconstructed images are improved, and signal-to-noise ratio efficiency and scanning efficiency are improved on the basis of guaranteeing the quality of images.

Description

Method and device for equal voxel magnetic resonance diffusion imaging based on simultaneous excitation of multiple plates

Cross-reference to related applications

This application claims the priority of Chinese Patent Application No. “201810967348.6” filed by Tsinghua University on August 23, 2018, with the invention name of “Isovoxel-based magnetic resonance diffusion imaging method and device based on simultaneous excitation of multiple plates”.

Technical field

The present invention relates to the technical field of magnetic resonance diffusion imaging, and in particular, to a method and device for iso-voxel magnetic resonance diffusion imaging based on simultaneous excitation of multiple plates.

Background technique

Magnetic resonance diffusion imaging technology is currently the only imaging method for measuring the diffusion motion of water molecules in vivo. It detects the microstructure of tissues by applying a diffusion gradient to sense the microscopic motion of water molecules. It can obtain both structural information and functional information. Therefore, This technology has developed rapidly in the past two decades and has gradually become an important routine clinical examination and scientific research tool. At present, the conventional magnetic resonance diffusion imaging technology in neuroscience and clinical diagnosis is single-shot EPI (echoplanar imaging). The single-shot EPI completes the acquisition of the entire k-space after a single-layer excitation of RF (radiofrequency, radio frequency) pulses, which has the advantages of fast imaging speed, insensitivity to motion, and relatively simple reconstruction algorithm. However, limited by the imaging principle, diffusion imaging based on single-shot EPI has the disadvantages of large image distortion and low resolution.

In order to reduce image distortion and improve image resolution, multiple excitation diffusion imaging has been proposed in recent years. Multiple excitations By reducing the number of phase-encoded lines acquired for each excitation and increasing the acquisition bandwidth, image distortion can be effectively reduced, a larger acquisition matrix can be obtained, and the spatial resolution within the layer can be improved. However, using two-dimensional multiple excitation techniques cannot effectively improve the interlayer resolution.

Related technologies show that high-resolution voxel diffusion imaging technology is of great significance in neuroscience research, especially in detecting the microstructure and regional connections of the brain. For example, in terms of microstructures, high-resolution voxel-diffusion imaging can help more accurately distinguish between different nerve fiber structures; better detect gray matter boundaries; and more accurately detect complex nerve fiber structures such as bends and Crossed nerve fibers.

The main challenge of high-resolution voxel diffusion imaging is how to ensure that the diffusion image has a sufficient signal-to-noise ratio without reducing the imaging efficiency. Currently there are two main solutions: the first is to use ultra-high field magnetic resonance scanners (such as the 7T platform). The ultra-high field magnetic resonance scanner can increase the strength of the signal itself, and is used to compensate for the loss of signal-to-noise ratio caused by high resolution. However, the ultra-high field magnetic resonance scanner also has its own limiting factors: the B ₀ / B ₁ magnetic field is more uneven, the T ₂ / T ₂ ^* decay time constant becomes shorter, the energy deposition effect increases, and so on.

The second scheme is to use a high SNR efficiency acquisition strategy. At present, there are mainly two types of hybrid 2D / 3D acquisition strategies that can guarantee high SNR efficiency. The first category is the SMS (simultaneous multi-slice) technology. Assuming that the R _SMS layer is excited at the same time, the TR time can be shortened to 1 / R _SMS of the traditional two-dimensional imaging method. The second type of method is the three-dimensional multi-plate acquisition technology: firstly, a "thick layer" is excited, and then three-dimensional Fourier coding is used to reconstruct a high-resolution diffusion image such as voxels. In three-dimensional multi-plate diffusion imaging, since the plate excited each time is relatively thick, there are fewer N _slices (or N _slabs ) in a single TR (repetition time) and the TR time is relatively short.

Summary of the Invention

The present invention aims to solve at least one of the technical problems in the related technology.

To this end, an object of the present invention is to propose an iso-voxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates, which can obtain high-resolution images while still maintaining a high signal-to-noise ratio and reducing three-dimensional The interference of the navigation echo error improves the quality and stability of the reconstructed image, and improves the signal-to-noise ratio efficiency and scanning efficiency on the basis of ensuring the image quality.

Another object of the present invention is to provide an iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates.

In order to achieve the above object, an embodiment of the present invention provides an iso-voxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates, including the following steps: Step S1: Exciting pulses simultaneously through multiple plates to perform multi-target detection. For each excitation, and during each excitation, signal acquisition is performed on the measured target through a multi-channel coil to obtain k-space data for each excitation reduction; step S2: through the combined k-space and image domain The multiple-excitation diffusion imaging reconstruction algorithm recovers the k-space position data that was not collected for each excitation; step S3: correct the edge artifacts by using the improved NPEN algorithm to obtain imaging images.

The isovoxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates according to the embodiment of the present invention can acquire high-resolution images while maintaining a high signal-to-noise ratio while using a three-dimensional multi-plate acquisition technology; the proposed combination K-space and image domain phase correction algorithms reduce the interference of 3D navigation echo errors and improve the quality and stability of reconstructed images; combined with the use of multi-plate simultaneous excitation technology, on the basis of ensuring image quality, improve the signal-to-noise ratio efficiency and scanning effectiveness.

In addition, the iso-voxel magnetic resonance diffusion imaging method based on the simultaneous excitation of multiple plates according to the above embodiment of the present invention may also have the following additional technical features:

Further, in an embodiment of the present invention, the step S1 further includes: applying different phase compensation to the sub-pulses of different plates in the multi-plate to eliminate the phase error generated by the plate interval; and / or During each excitation process, navigation echoes are collected to obtain phase information for each excitation, wherein the simultaneous excitation pulses of the multiple plates are represented as:

RF is a single-layer excited RF pulse, R _SMS is the number of simultaneous excitation layers, i is the i-th layer that is excited simultaneously, ω _i is the frequency modulation, ω _i = γ · (i-1) · G · d, and γ is the spin Magnetic ratio, G is the gradient of layer selection, and d is the center distance between two adjacent layers.

It is the phase error generated by the plate interval in the gradient encoding process.

Further, in an embodiment of the present invention, the step S2 further comprises: restoring the initial diffusion image for each excitation through a 2D CK-GRAPPA algorithm; and reconstructing by POCSMUSE to introduce coil sensitivity and phase smoothing constraints to reduce residual artifact And noise, the reconstruction diffusion image is estimated for each excitation, and is continuously updated in each iteration to obtain the final diffusion image for each excitation.

Further, in an embodiment of the present invention, the improved NPEN algorithm is to obtain a non-linear optimization problem according to the contour coding of the plate level, and to solve the non-linear optimization problem by inverse of a non-linear equation, and in the process of solving , Iteratively optimizes the slice-level contours to correct the edge artifacts.

Further, in an embodiment of the present invention, the target equation of plate edge artifact correction is:

Wherein, n represents the n-th sector, N _grp is the number of groups, each group is defined as a single unit R _SMS blocks are simultaneously excited, R _SMS is a multi-plate speedup excited simultaneously, so the total number of blocks of N = R _SMS N _grp . P is the operator that selects the corresponding collected k-space position, F is the Fourier transform, and C is the coil sensitivity code.

It is the slice contour of simultaneous excitation pulses of multiple plates, μ is the image to be reconstructed.

In order to achieve the above object, an embodiment of another aspect of the present invention proposes an iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates, including: an acquisition module for simultaneously exciting pulses through multiple plates to perform a measurement on a target to be measured. Multiple excitations, and during each excitation, signal acquisition is performed on the measured target through a multi-channel coil to obtain k-space data that is reduced for each excitation; a reconstruction module is used to pass the joint k-space The multiple-excitation diffusion imaging reconstruction algorithm in the image field and the image domain recovers the k-space position data that is not collected for each excitation; a correction module is used to correct edge artifacts through an improved NPEN algorithm to obtain an imaging image.

The iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates according to the embodiment of the present invention can acquire high-resolution images while still maintaining a high signal-to-noise ratio by using three-dimensional multi-plate acquisition technology; the proposed combination K-space and image domain phase correction algorithms reduce the interference of 3D navigation echo errors and improve the quality and stability of reconstructed images; combined with the use of multi-plate simultaneous excitation technology, on the basis of ensuring image quality, improve the signal-to-noise ratio efficiency and scanning effectiveness.

In addition, the iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates according to the above embodiment of the present invention may also have the following additional technical features:

Further, in an embodiment of the present invention, the acquisition module is further configured to apply different phase compensation to the sub-pulses of different plates in the multi-plate to eliminate phase errors generated by the plate interval; and / or During each excitation process, navigation echoes are collected to obtain phase information for each excitation.

Further, in an embodiment of the present invention, the reconstruction module is further configured to recover an initial diffusion image for each excitation through a 2D CK-GRAPPA algorithm; and reconstructing by POCSMUSE to introduce coil sensitivity and phase smoothing constraints to reduce residual artifacts And noise, the reconstruction diffusion image is estimated for each excitation, and is continuously updated in each iteration to obtain the final diffusion image for each excitation.

Additional aspects and advantages of the present invention will be given in part in the following description, part of which will become apparent from the following description, or be learned through the practice of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the following description of the embodiments with reference to the accompanying drawings, in which:

FIG. 1 is a flowchart of an isotope voxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates according to an embodiment of the present invention;

2 is a sequence diagram of simultaneous excitation magnetic resonance diffusion imaging of a multi-plate with navigation echo based on spin echo based on an embodiment of the present invention;

3 is a schematic diagram of the shapes of excitation pulses and first refocusing pulses in 4 different kz planes according to an embodiment of the present invention;

4 is an overall reconstruction flowchart according to an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of an isotope magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates according to an embodiment of the present invention.

detailed description

Hereinafter, embodiments of the present invention will be described in detail. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary and are intended to explain the present invention, but should not be construed as limiting the present invention.

The method and device for iso-voxel magnetic resonance diffusion imaging based on simultaneous excitation of multiple plates according to the embodiment of the present invention will be described below with reference to the drawings. MRI diffusion imaging method.

FIG. 1 is a flowchart of an iso-voxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates according to an embodiment of the present invention.

As shown in FIG. 1, the iso-voxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates includes the following steps:

Step S1: Simultaneous excitation of pulses through multiple plates to perform multiple excitations on the measured target, and during each excitation, signal acquisition is performed on the measured targets through multi-channel coils to obtain the k-space for each excitation reduction. data.

In an embodiment of the present invention, step S1 further includes: applying different phase compensation to the sub-pulses of different plates in the multiple plates to eliminate the phase error generated by the plate interval; and / or during each excitation, Acquire navigation echoes to obtain phase information for each shot.

More specifically, and as multilayer simultaneous excitation, while the multi-section excitation RF pulse is obtained from R _SMS number of RF pulse frequency modulation are added,

RF represents a single-layer excited RF pulse, ω _i represents frequency modulation, ω _i = γ · (i-1) · G · d, γ is a gyromagnetic ratio (2π × 42.575MHz for H ¹ ), and i indicates simultaneous excitation. For the i-th layer, G is the selection gradient, d is the center interval between two adjacent layers, and R _SMS is the number of simultaneous excitation layers.

When using 3D multi-plate acquisition, a second phase encoding gradient needs to be applied in the layer selection direction. The phase generated by the gradient encoding is related to the position. When multi-layer simultaneous excitation is further combined, the plate spacing will bring additional phase error, and its size is:

Among them, d _slab represents the interval between two adjacent plates, k _z is the index of the collected k space in the layer direction, R _SMS is the number of plates excited at the same time, N is the number of layers to be reconstructed for each plate, and Δz is the desired High resolution of reconstructed layer orientation.

Therefore, when designing multi-plate excitation pulses at the same time, it is necessary to apply different phase compensation to the sub-pulses corresponding to different plates in order to eliminate the phase error caused by the plate interval during the acquisition process, that is:

The meaning of each symbol is the same as that of formulas 1 and 2.

The spin-echo based multi-plate simultaneous magnetic resonance diffusion imaging sequence with navigation echo is shown in Figure 2. Among them, the excitation pulse and the first refocusing pulse (for generating imaging echo) can be designed according to Equation 3.

According to Equation 2, for different kz,

The sizes are different, that is, the RF _SMSlab shapes corresponding to different kz are not exactly the same. Figure 3 shows the shape of the excitation pulse and the first refocusing pulse in 4 different kz planes.

It is worth noting that the above is the basic composition of the multi-plate excitation pulse. In practical applications, for the optimization of the RF pulse level profile, more advanced SLR designs can be used for basic RF pulses. In order to reduce the energy deposition during imaging and the maximum B1 value, multi-plate excitation pulses can use more advanced designs, such as VERSE, PINS, etc. Ω _i and

Optimized so that the final RF _SMSlab only has amplitude information, which can achieve better performance on some systems.

If the multi-plate simultaneous excitation technology is used for magnetic resonance diffusion imaging, a navigation echo needs to be collected for each excitation pair to record the phase information of each excitation. Because navigation echoes are usually collected using a single-shot EPI, phase interference caused by plate spacing cannot be resolved by the RF pulse coding method described above, so three-dimensional navigation echoes cannot be directly collected. At the same time, it is considered that the phase change along the layer direction in each plate is not large, so a two-dimensional navigation echo is collected each time it is excited. It is worth noting that the shape of the second refocusing pulse does not match the shape of the first refocusing pulse. Assuming the first refocusing pulse is shown in Equation 3, the second refocusing pulse is:

In the embodiment of the present invention, the signal acquisition may use an imaging sequence with multiple excitations with navigation data, such as, but not limited to, multiple excitation EPI diffusion imaging with navigation data, propeller diffusion imaging, or multiple excitation spiral diffusion imaging . The embodiments of the present invention are not limited to the types of imaging methods with multiple excitations. The navigation data may be self-navigation data (such as VDS (variable density spiral)), or it may be navigation data acquired in addition.

To estimate the coil sensitivity C required in steps S2 and S3, a set of low-resolution images is acquired in advance ("calibration scan 1"). The pulse sequence can be different from the above-mentioned real image acquisition, and the traditional two-dimensional acquisition method is usually used. Before the final reconstruction, the coil sensitivity is transformed to the corresponding position of the real image through operations such as interpolation.

In order to estimate the RF pulse slice profile required in step S3, an additional set of non-diffusion-coded calibration data is acquired in advance ("calibration scan 2"). Oversampling 2 times in the kz direction, using the same pulse shape as the real image acquisition, and keeping the TR consistent.

Step S2: The k-space position data that is not collected for each excitation is restored by a multiple excitation diffusion imaging reconstruction algorithm combining k-space and image domain.

It can be understood that the embodiment of the present invention proposes a multiple-excitation diffusion imaging reconstruction technique that combines k-space and image domains, recovers data of k-space positions that are not collected for each excitation, and corrects phase changes between different excitations. The complete k-space data of each excitation is recovered to perform inverse Fourier transform, and the images of each excitation are complexly combined.

Specifically, firstly, the reconstruction technology of parallel imaging is used to restore the navigation echo data of each excitation to a complete k-space. The recovery process can use k-space algorithms (such as GRAPPA). For example, the k-space-based data interpolation process can be expressed by the following formula:

Where d _j (m, n, p) is the k-space data corresponding to the j-th channel of the data point to be restored, and (m, n, p) is the kx-ky-kz coordinate of the point to be interpolated; d _{j '} ( m ', n', p ') is the k-space data collected by the interpolation kernel on the j'th channel, and (m', n ', p') is the kx- of the points collected by the interpolation kernel (represented by K) ky-kz coordinates; j, j'∈ (1, N _c ), where N _c is the total number of channels. w ₁ (j ′, m ′, n ′, p ′) is the weight coefficient corresponding to the j′-th channel and data point (m ′, n ′, p ′), which can be estimated from the data of calibration scan 1.

Suppose in calibration scan 1, the point collected at the j-th channel, (m, n, p) position is b _j (m, n, p), and the j'-th channel, (m ', n', p ' ) At the point a _{j ′} (m ′, n ′, p ') collected, the equation corresponding to calibration scan 1 and equation 5 is:

Write the above equation as a vector,

b = aw ₁ (7),

Using the least square method or other linear equation solving algorithms, the process of solving the weight matrix w ₁ can be expressed as:

w ₁ = [a ^H a] ^-1 [a ^H b] (8).

For imaging echo data without diffusion coding, an image can be obtained by directly inverse Fourier change. If diffusion coding is applied, the phase change between different excitations needs to be corrected. Here, a multiple excitation diffusion imaging reconstruction method combining k-space and image domain is proposed, which mainly includes the following two steps.

Further, in an embodiment of the present invention, step S2 further includes: restoring the initial diffusion image for each excitation by a 2D CK-GRAPPA algorithm; and reconstructing by POCSMUSE to introduce coil sensitivity and phase smoothing constraints to reduce residual artifacts and noise , The reconstructed diffusion image for each excitation is estimated and continuously updated in each iteration to obtain the final diffusion image for each excitation.

Specifically, step 1: The specific interpolation process using the 2D CK-GRAPPA algorithm can be expressed by the following formula:

Among them, d _{i, j} (m, n, p) is the k-space data corresponding to the i-th excitation and j-th channel of the data point to be recovered, and (m, n, p) is the kx-ky- kz coordinates; d _{i ', j'} (m ', n', p ') is the k-space data collected by the interpolation kernel at the i'th excitation and j'th channel, (m', n ', p' ) Is the kx-ky-kz coordinates of the points collected by the interpolation kernel (represented by K); i, i'∈ (1, N _shot ), N _shot is the total number of shots; j, j'∈ (1, N _c ), N _c is the total number of channels. w ₂ is a weight matrix, which can be calculated from the restored navigation echo (formula 5), and the specific process is similar to that of formulas 6 to 8.

Using 2D CK-GRAPPA, the diffusion image of each excitation can be roughly recovered. However, there are two problems here: (1) In the three-dimensional multi-plate acquisition, the equivalent multiplying reduction factor for each excitation is very large. Using the 2D CK-GRAPPA algorithm alone, the image quality recovered for each excitation is also limited. (2) Since the TE (echo time, echo time) collected by the 3D navigation echo in different kz planes is different, affected by the non-uniformity of the magnetic field, different kz planes may have a certain phase error, which is directly returned by the navigation. The accuracy of the phase diagram estimated by the wave will be reduced.

Step 2: On the basis of 2D CK-GRAPPA, plus POCSMUSE reconstruction, by introducing coil sensitivity and phase smoothing constraints, the residual artifacts and noise are further reduced. The initial phase map is estimated from the diffusion image of each excitation reconstructed by 2DCK-GRAPPA, and is continuously updated in each iteration. The termination condition of POCSMUSE is that the error between two successive iterations is less than a preset threshold, or the number of iterations is equal to a preset value. Among them, the overall reconstruction process is shown in Figure 4.

Step S3: The edge artifact is corrected by the improved NPEN algorithm to obtain an imaging image.

In one embodiment of the present invention, the improved NPEN algorithm is to obtain a non-linear optimization problem according to the contour coding of the plate level, and solve the non-linear optimization problem by inverse of a non-linear equation, and in the process of solving, iteratively optimize the plate Layer contours to correct edge artifacts.

Specifically, another challenge of 3D multi-plate acquisition is plate edge artifacts, which are mainly caused by non-ideal slice contours of RF pulses. The ideal RF pulse level profile should be rectangular, and its layer selection width is equal to the layer thickness, but it requires an infinitely long pulse time, so it is impossible to achieve in practice. In practice, due to various practical factors such as the truncation effect in the time domain, RF pulses always produce non-ideal layer contours, such as Gibbs ringing artifacts, transition bands, etc. Wait. In 3D multi-plate imaging, if the reconstructed "thin layer" is located in the transition zone area, the actual signal amplitude is lower than the theoretical value. At the same time, the size of the field-of-view (FOV) of the spatial coding of the layer selection direction is generally smaller than the excited FOV, resulting in aliasing artifacts. There may be overlap between adjacent plates, causing crosstalk between plates.

For the three-dimensional multi-plate imaging technology (not combined with the multi-layer simultaneous excitation technology), some correction algorithms for plate edge artifacts have been proposed from the perspective of image acquisition and reconstruction. Here we mainly introduce two algorithms based on the reconstruction angle. The basic principle is It is considered that the plate edge aliasing artifacts are similar to the parallel reduced mining aliasing, except that the coil sensitivity coding is replaced by the plate level contours. Therefore, the idea similar to SENSE can be used to solve the plate edge artifacts. According to the linearity of solving the objective equation, it can be divided into PEN (slab profile coding) method and NPEN (nonlinear inversion for slab profile coding) method. PEN believes that plate edge artifacts are linear coding problems, which can be solved by linear equation inversion. PEN can effectively solve the problem of aliasing between plates, but there may be residual plate crosstalk artifacts, especially at short TR. NPEN is based on PEN. Plate-level contour coding is considered as a non-linear optimization problem, which can be solved using the method of inverse of nonlinear equations, and iteratively optimizes the plate-level contour during the solution process. The objective equation of NPEN can be written as follows:

E (x) = d, x = [μ, S] ^T (10),

E (x) = (PFCs ₁ u PFCs ₂ u ... PFCs _N u) (11),

Among them, μ is the image to be reconstructed, S = [s ₁ s ₂ ... s _N ] ^T represents the RF pulse slice profile of different plates, which can be estimated from the calibration scan 2, N is the total number of plates, and d = [d ₁ d ₂ ... d _N ] ^T is the acquired k-space signal, E is a non-linear coding matrix, and is composed of three parts: C is the coil sensitivity code, which can be estimated from calibration scan 1, F is the Fourier transform, P is an operator that selects the corresponding acquired k-space position.

The above equations can be solved by several commonly used linear or non-linear optimization algorithms, such as Gauss-Newton algorithm, gradient descent algorithm, conjugate gradient algorithm and so on. In order to improve the reconstruction effect, different regular term constraints can be added according to the needs during the solution process, such as the smoothness of the pulse level contour within the layer, and the plate edge artifacts are periodic in the image domain.

When using multiple plate simultaneous excitation technology, the main challenge is that the aliasing form of plate edge artifacts will change: from intra-slab aliasing to inter-slab aliasing, so the original plate The objective equation of edge artifact correction needs corresponding improvement, as shown in the following formula. The algorithm for solving the objective equation can be as before.

In summary, the embodiment of the present invention uses magnetic resonance diffusion imaging as an example to introduce the proposed image acquisition and reconstruction strategy, but is not limited to diffusion imaging. The related image acquisition and reconstruction methods can be applied to other imaging modes, such as functions Magnetic resonance imaging (fMRI). When used in fMRI, the acquisition module is as described above. Based on the imaging sequence of two-dimensional multiple excitations, a three-dimensional multi-plate diffusion imaging sequence is designed to acquire signals. The reconstruction module is simpler and there is no phase inconsistency between different excitations. For problems, you can directly use traditional parallel imaging reconstruction algorithms (such as 2D GRAPPA). On this basis, the multi-layer simultaneous excitation technology and the three-dimensional multi-plate imaging technology are further combined to achieve high-resolution voxel imaging while improving imaging efficiency and signal-to-noise ratio efficiency.

According to the iso-voxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates according to the embodiment of the present invention, by using a three-dimensional multi-plate acquisition technology, a high-resolution image can be acquired while maintaining a high signal-to-noise ratio; Phase correction algorithm that combines k-space and image domain, reduces the interference of 3D navigation echo error, improves the quality and stability of reconstructed image; combined with the use of multi-plate simultaneous excitation technology, improves the signal-to-noise efficiency on the basis of ensuring image quality And scanning efficiency.

Next, an iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates according to an embodiment of the present invention will be described with reference to the drawings.

FIG. 5 is a schematic structural diagram of an iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates according to an embodiment of the present invention.

As shown in FIG. 5, the iso-voxel magnetic resonance diffusion imaging apparatus 10 based on simultaneous excitation of multiple plates includes: an acquisition module 100, a reconstruction module 200, and a correction module 300.

Among them, the acquisition module 100 is used for simultaneously stimulating pulses through multiple plates to perform multiple excitations on the measured target, and during each excitation, signal acquisition is performed on the measured target through a multi-channel coil to obtain each excitation drop. Collected k-space data. The reconstruction module 200 is configured to recover k-space position data that is not collected for each excitation through multiple excitation diffusion imaging reconstruction algorithms combining k-space and image domain. The correction module 300 is configured to correct edge artifacts by using an improved NPEN algorithm to obtain an imaging image. The device 10 according to the embodiment of the present invention can maintain a high signal-to-noise ratio while acquiring a high-resolution image, and reduce the interference of three-dimensional navigation echo errors, improve the quality and stability of the reconstructed image, and ensure the quality of the image. In order to improve the signal-to-noise ratio efficiency and scanning efficiency.

Further, in an embodiment of the present invention, the acquisition module 100 is further configured to apply different phase compensation to the sub-pulses of different plates in the multiple plates to eliminate the phase error generated by the plate interval; and / or at each excitation In the process of collecting navigation echoes to obtain phase information of each excitation, wherein the simultaneous excitation pulses of the multiple plates are expressed as:

Further, in one embodiment of the present invention, the reconstruction module 200 is further configured to recover the initial diffusion image for each excitation by the 2D CK-GRAPPA algorithm; and reconstruct the POCSMUSE to introduce coil sensitivity and phase smoothing constraints to reduce residual artifacts and Noise, reconstructed diffusion image estimates for each excitation, and continuously update in each iteration to obtain the final diffusion image for each excitation.

Further, in one embodiment of the present invention, the improved NPEN algorithm is to obtain a non-linear optimization problem according to the contour coding of the plate level, and to solve the non-linear optimization problem by inverse of the non-linear equation, and in the process of solving, iterative optimization Plate-level contours to correct edge artifacts.

It should be noted that the foregoing explanation of the embodiment of the isovoxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates is also applicable to the isovoxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates in this embodiment, here No longer.

According to the embodiment of the present invention, an iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates can acquire high-resolution images while maintaining a high signal-to-noise ratio by using three-dimensional multi-plate acquisition technology. Phase correction algorithm that combines k-space and image domain, reduces the interference of 3D navigation echo error, improves the quality and stability of reconstructed image; combined with the use of multi-plate simultaneous excitation technology, improves the signal-to-noise efficiency on the basis of ensuring image quality And scanning efficiency.

In addition, the terms "first" and "second" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "a plurality" is at least two, for example, two, three, etc., unless it is specifically and specifically defined otherwise.

In the description of this specification, the description with reference to the terms “one embodiment”, “some embodiments”, “examples”, “specific examples”, or “some examples” and the like means specific features described in conjunction with the embodiments or examples , Structure, material, or characteristic is included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms are not necessarily directed to the same embodiment or example. Moreover, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. In addition, without any contradiction, those skilled in the art may combine and combine different embodiments or examples and features of the different embodiments or examples described in this specification.

Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limitations on the present invention. Those skilled in the art can interpret the above within the scope of the invention Embodiments are subject to change, modification, substitution, and modification.

Claims

An isotope-based magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates is characterized in that it includes the following steps:

Step S1: Simultaneous excitation of pulses through multiple plates to perform multiple excitations on the measured target, and during each excitation, signal acquisition is performed on the measured target through a multi-channel coil to obtain the mined mined per excitation. k-space data

Step S2: recovering the k-space position data that is not collected for each excitation through the multiple-excitation diffusion imaging reconstruction algorithm that combines k-space and image domain; and

Step S3: The edge artifact is corrected by the improved NPEN algorithm to obtain an imaging image.
The method according to claim 1, wherein the step S1 further comprises:

Applying different phase compensation to the sub-pulses of different plates in the multi-plate to eliminate the phase error generated by the plate interval; and / or

During the process of each excitation, a navigation echo is collected to obtain phase information of each excitation, wherein:

The simultaneous excitation pulses of the multiple plates are expressed as:

RF is a single-layer excited RF pulse, R SMS is the number of simultaneous excitation layers, i is the i-th layer that is excited simultaneously, ω i is the frequency modulation, ω i = γ · (i-1) · G · d, and γ is the spin Magnetic ratio, G is the gradient of layer selection, and d is the center distance between two adjacent layers.
It is the phase error generated by the plate interval in the gradient encoding process.
The method of claim 1, wherein the step S2 further comprises:

The 2D CK-GRAPPA algorithm is used to recover the initial diffusion image for each shot;

The POCSMUSE reconstruction is used to introduce coil sensitivity and phase smoothing constraints to reduce residual artifacts and noise. The reconstructed diffusion image estimates for each excitation are continuously updated in each iteration to obtain the final diffusion image for each excitation.
The iso-voxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates according to claim 1, characterized in that the improved NPEN algorithm is to obtain a nonlinear optimization problem according to the contour encoding of the plate level, and to obtain the nonlinear optimization problem through a nonlinear equation The non-linear optimization problem is solved inversely, and in the process of solving, the contour of the plate level is iteratively optimized to correct the edge artifacts.
The iso-voxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates according to claim 4, wherein the target equation of plate edge artifact correction is:

Wherein, n represents the n-th sector, N grp is the number of groups, each group is defined as a single unit R SMS blocks are simultaneously excited, R SMS is a multi-plate speedup excited simultaneously, so the total number of blocks of N = R SMS N grp , P is the operator that selects the corresponding collected k-space position, F is the Fourier transform, and C is the coil sensitivity code.
It is the slice contour of simultaneous excitation pulses of multiple plates, μ is the image to be reconstructed.
An iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates is characterized in that it includes:

An acquisition module is configured to simultaneously excite pulses through multiple plates to excite the measured target multiple times, and during each excitation process, perform signal acquisition on the measured target through a multi-channel coil to obtain each excitation drop. Collected k-space data;

A reconstruction module for reconstructing and recovering k-space position data that is not collected for each excitation through the multiple excitation diffusion imaging algorithm combining the k-space and image domain; and

A correction module is used to correct edge artifacts through an improved NPEN algorithm to obtain an imaging image.
The iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates according to claim 6, wherein the acquisition module is further configured to apply different phase compensation to the sub-pulses of different plates in the multiple plates, To eliminate the phase error generated by the plate interval; and / or collect navigation echoes during each excitation to obtain the phase information of each excitation, wherein the simultaneous excitation pulses of the multiple plates are represented as:

RF is a single-layer excited RF pulse, R SMS is the number of simultaneous excitation layers, i is the i-th layer that is excited simultaneously, ω i is the frequency modulation, ω i = γ · (i-1) · G · d, and γ is the spin Magnetic ratio, G is the gradient of layer selection, and d is the center distance between two adjacent layers.
It is the phase error generated by the plate interval in the gradient encoding process.
The iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates according to claim 6, wherein the reconstruction module is further configured to recover an initial diffusion image for each excitation by a 2D CK-GRAPPA algorithm; and by POCSMUSE Reconstruction introduces coil sensitivity and phase smoothing constraints to reduce residual artifacts and noise. The reconstructed diffusion image estimates for each excitation are continuously updated in each iteration to obtain the final diffusion image for each excitation.
The iso-voxel magnetic resonance diffusion imaging device based on simultaneous excitation of multiple plates according to claim 6, characterized in that the improved NPEN algorithm is to obtain a nonlinear optimization problem according to the contour encoding of the plate level, and to obtain the nonlinear optimization problem through a nonlinear equation The non-linear optimization problem is solved inversely, and in the process of solving, the contour of the plate level is iteratively optimized to correct the edge artifacts.
The iso-voxel magnetic resonance diffusion imaging method based on simultaneous excitation of multiple plates according to claim 8, wherein the target equation of plate edge artifact correction is:

Wherein, n represents the n-th sector, N grp is the number of groups, each group is defined as a single unit R SMS blocks are simultaneously excited, R SMS is a multi-plate speedup excited simultaneously, so the total number of blocks of N = R SMS N grp , P is the operator that selects the corresponding collected k-space position, F is the Fourier transform, and C is the coil sensitivity code.
It is the slice contour of simultaneous excitation pulses of multiple plates, μ is the image to be reconstructed.