Image reconstruction model generation and image reconstruction method, device, equipment and medium

Abstract

The embodiment of the present invention discloses an image reconstruction model generation and image reconstruction method, device, equipment and medium, wherein the image reconstruction model generation method includes: obtaining the full-sampled K-space data of a dynamic magnetic resonance image sequence, and obtaining the under-sampled K-space data corresponding to the full-sampled K-space data based on a preset under-sampled model; inputting the under-sampled K-space data into, according to the sub-problem obtained by solving the magnetic resonance image reconstruction problem under the constraints of low-rank characteristics and sparse characteristics for the full-sampled K-space data, establishing an image reconstruction model, and training the model; when the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets the preset conditions, the model training is completed to obtain the target image reconstruction model. The technical solution of this embodiment realizes the improvement of the image reconstruction speed, can reconstruct images according to the sparse prior and low-rank prior characteristics of the sampling data, and improves the image quality.

Description

Technical Field

[0001] Embodiments of the present invention relate to the technical field of medical image processing, and in particular to an image reconstruction model generation and image reconstruction method, device, equipment and medium. Background Art

[0002] Magnetic resonance cardiac cine imaging is a non-invasive imaging technology that can be used to evaluate cardiac function, ventricular wall motion abnormalities, etc., providing rich information for clinical cardiac diagnosis. However, due to the limitations of magnetic resonance physics and hardware, and the duration of the cardiac motion cycle, magnetic resonance cardiac cine imaging often has limited temporal and spatial resolution and cannot accurately evaluate some heart diseases, such as arrhythmia and other cardiac function conditions. Therefore, it is necessary to use fast imaging methods to improve the temporal and spatial resolution of magnetic resonance cardiac cine imaging while ensuring imaging quality.

[0003] Currently, commonly used methods to accelerate magnetic resonance cardiac cine imaging include parallel imaging (PI), compressed sensing (CS) technology, low-rank matrix decomposition and deep learning methods.

[0004] However, traditional parallel imaging or compressed sensing technology does not use big data priors, and the parameters of this iterative optimization method are difficult to select and time-consuming. Although the neural network method based on deep learning can avoid the iterative solution step and speed up the reconstruction time, these deep learning methods only rely on the sparse priors of big data, which limits the improvement of image reconstruction effect. Summary of the invention

[0005] The embodiments of the present invention provide an image reconstruction model generation and image reconstruction method, device, equipment and medium, so as to achieve the goal of accelerating the image reconstruction speed while making full use of the low rank and sparse characteristics of the sampling data to establish an image reconstruction model to improve the reconstructed image quality.

[0006] In a first aspect, an embodiment of the present invention provides a method for generating an image reconstruction model, the method comprising:

[0007] Acquire full-sampled K-space data of a dynamic magnetic resonance image sequence, and obtain under-sampled K-space data corresponding to the full-sampled K-space data based on a preset under-sampling model;

[0008] The under-sampled K-space data is input into an image reconstruction model established according to a sub-problem obtained by solving a magnetic resonance image reconstruction problem under the constraints of low-rank characteristics and sparse characteristics for the fully sampled K-space data, and the model is trained;

[0009] When the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets a preset condition, the model training is completed and the target image reconstruction model is obtained.

[0010] Optionally, the process of establishing an image reconstruction model according to a sub-problem obtained by solving a magnetic resonance image reconstruction problem under low-rank and sparse characteristics constraints for the fully sampled K-space data includes:

[0011] According to the temporal periodicity of the motion of the magnetic resonance scanning object, the magnetic resonance image reconstruction problem of the full-sampled K-space data is modeled under the constraints of low-rank characteristics and sparse characteristics;

[0012] Setting auxiliary variables of a representation model of the magnetic resonance image reconstruction problem, and determining a penalty function of the representation model based on the auxiliary variables;

[0013] Determine three sub-problems of the representation model according to the penalty function;

[0014] The three sub-problems are solved respectively, and the image reconstruction model is obtained through networking.

[0015] Optionally, solving the three sub-problems respectively and obtaining the image reconstruction model through networking includes:

[0016] Optimize and rewrite the three sub-problems respectively;

[0017] The approximate gradient method is used to solve the three sub-problems after optimization rewriting, and the iterative representation of the solution of each sub-problem is obtained;

[0018] The iterative representation process is networked to obtain the image reconstruction model.

[0019] Optionally, optimizing the three sub-problems respectively includes:

[0020] The three sub-problems are optimized respectively according to the result of the penalty function being expanded and represented by the preset numerical value of the data fidelity term.

[0021] Optionally, the image reconstruction model includes a low-rank network module, a sparse network module and a data consistency module.

[0022] In a second aspect, an embodiment of the present invention further provides an image reconstruction method, the method comprising:

[0023] Acquiring under-sampled K-space data of a dynamic magnetic resonance image sequence based on a preset under-sampling model;

[0024] The under-sampled K-space data is input into a target image reconstruction model obtained by the image reconstruction model generation method described in any embodiment to obtain a reconstructed image corresponding to the under-sampled K-space data.

[0025] In a third aspect, an embodiment of the present invention further provides an image reconstruction model generation device, the device comprising:

[0026] A data preprocessing module, used for acquiring full-sampled K-space data of a dynamic magnetic resonance image sequence, and obtaining under-sampled K-space data corresponding to the full-sampled K-space data based on a preset under-sampling model;

[0027] A data input module is used to input the under-sampled K-space data into an image reconstruction model established according to a sub-problem obtained by solving a magnetic resonance image reconstruction problem under the constraints of low-rank characteristics and sparse characteristics for the fully sampled K-space data, and train the model;

[0028] The model generation module is used to complete the model training and obtain the target image reconstruction model when the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets the preset conditions.

[0029] Optionally, the image reconstruction model generating device also includes a model construction module, which is used to establish an image reconstruction model based on sub-problems obtained by solving the magnetic resonance image reconstruction problem under the constraints of low rank and sparse characteristics for the fully sampled K-space data.

[0030] Optionally, the model construction module is specifically used to:

[0031] According to the temporal periodicity of the motion of the magnetic resonance scanning object, the magnetic resonance image reconstruction problem of the full-sampled K-space data is modeled under the constraints of low-rank characteristics and sparse characteristics;

[0032] Setting auxiliary variables of a representation model of the magnetic resonance image reconstruction problem, and determining a penalty function of the representation model based on the auxiliary variables;

[0033] Determine three sub-problems of the representation model according to the penalty function;

[0034] The three sub-problems are solved respectively, and the image reconstruction model is obtained through networking.

[0035] Optionally, the model construction module may also be used to:

[0036] Optimize and rewrite the three sub-problems respectively;

[0037] The approximate gradient method is used to solve the three sub-problems after optimization rewriting, and the iterative representation of the solution of each sub-problem is obtained;

[0038] The iterative representation process is networked to obtain the image reconstruction model.

[0039] Optionally, the model construction module can also be used to optimize the three sub-problems separately, specifically including:

[0040] The three sub-problems are optimized respectively according to the result of the penalty function being expanded and represented by the preset numerical value of the data fidelity term.

[0041] Optionally, the image reconstruction model includes a low-rank network module, a sparse network module and a data consistency network module.

[0042] In a fourth aspect, an embodiment of the present invention further provides an image reconstruction device, the device comprising:

[0043] A data acquisition module, used for acquiring under-sampling K-space data of a dynamic magnetic resonance image sequence obtained based on a preset under-sampling model;

[0044] An image reconstruction module is used to input the under-sampled K-space data into a target image reconstruction model obtained by the image reconstruction model generation method described in any embodiment to obtain a reconstructed image corresponding to the under-sampled K-space data.

[0045] In a fifth aspect, an embodiment of the present invention further provides a computer device, the computer device comprising:

[0046] one or more processors;

[0047] A memory for storing one or more programs;

[0048] When the one or more programs are executed by the one or more processors, the one or more processors implement the image reconstruction model generation method or image reconstruction method provided by any embodiment of the present invention.

[0049] In a sixth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements an image reconstruction model generation method or an image reconstruction method as provided in any embodiment of the present invention.

[0050] The embodiments of the above invention have the following advantages or beneficial effects:

[0051] The embodiment of the present invention processes the fully sampled K-space data samples of the acquired dynamic magnetic resonance image sequence according to a preset under-sampling model to obtain under-sampled K-space data corresponding to the fully sampled K-space data; then, the under-sampled K-space data is input into an image reconstruction model established based on a sub-problem obtained by solving the magnetic resonance image reconstruction problem under the constraints of low-rank characteristics and sparse characteristics for the fully sampled K-space data, and the model is trained; when the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the fully sampled K-space data meets the preset conditions, the model training is completed to obtain the target image reconstruction model. The problem in the prior art that the image reconstruction time is long or the sampling data characteristics cannot be fully utilized for image reconstruction is solved, and the low-rank characteristics and sparse characteristics of the sampling data can be fully utilized while speeding up the image reconstruction speed to establish an image reconstruction model to improve the quality of the reconstructed image. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a flow chart of a method for generating an image reconstruction model provided in Embodiment 1 of the present invention;

[0053] Figure 2 is a schematic diagram of a network structure of an image reconstruction network provided by Embodiment 1 of the present invention;

[0054] Figure 3 is a flow chart of an image reconstruction method provided by Embodiment 2 of the present invention;

[0055] Figure 4 is a structural schematic diagram of an image reconstruction model generation device provided by Embodiment 3 of the present invention;

[0056] Figure 5 is a structural schematic diagram of an image reconstruction device provided by Embodiment 4 of the present invention;

[0057] Figure 6 It is a structural diagram of a computer device provided in Embodiment 5 of the present invention. DETAILED DESCRIPTION

[0058] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0059] Embodiment 1

[0060] Figure 1This is a flow chart of a method for generating an image reconstruction model provided in Embodiment 1 of the present invention. This embodiment is applicable to the case where the image reconstruction model is trained using the full sampling image of the dynamic magnetic resonance image as a sample. The method can be executed by an image reconstruction model generation device, which can be implemented by software and / or hardware and integrated into an electronic device with application development function.

[0061] like Figure 1 As shown, the image reconstruction model generation method includes the following steps:

[0062] S110, acquiring full-sampled K-space data of a dynamic magnetic resonance image sequence, and obtaining under-sampled K-space data corresponding to the full-sampled K-space data based on a preset under-sampling model.

[0063] The full-sampled K-space data of the dynamic magnetic resonance image sequence is pre-collected sample data, and a high-resolution magnetic resonance dynamic image can be reconstructed based on the full-sampled K-space data of the dynamic magnetic resonance image sequence. For example, the dynamic magnetic resonance image sequence can be each frame of magnetic resonance cardiac movie imaging. The preset undersampling model can be determined by the preset undersampling operator, Fourier transform operator and coil sensitivity parameters according to the requirements of the sampling acceleration multiple.

[0064] Specifically, the preset undersampling model can be expressed as y=MFCX(1); wherein M is an undersampling operator, F is a Fourier transform operator, C is a coil sensitivity matrix (sensitivity maps), X is a one-dimensional vector formed by stretching each frame of a fully sampled dynamic magnetic resonance image sequence, and each column is spliced ​​together as a column in the order of the image frame to form an image matrix, and y is the undersampling K-space data. In this step, first, multiple columns of the fully sampled K-space data corresponding to each frame of the dynamic magnetic resonance image sequence are spliced ​​into a single-column data in the order of the columns; then, the single-column data corresponding to all the image frames are spliced ​​in the order of the image frames to obtain X. C can be estimated by algorithms such as ESPIRiT or Walsh, F is a Fourier transform operator, and M is a vector composed of 0 and 1, which determines which points are collected and which points are not collected, and the value of the uncollected points is 0.

[0065] S120. Input the under-sampled K-space data into an image reconstruction model established based on a sub-problem obtained by solving the magnetic resonance image reconstruction problem under the constraints of low-rank and sparse characteristics for the fully sampled K-space data, and train the model.

[0066] The image reconstruction model is an image reconstruction model that is modeled and trained based on the sparse prior and low-rank prior characteristics of dynamic magnetic resonance data. Specifically, the construction process of the image reconstruction model is as follows:

[0067] First, a low-rank matrix is ​​constructed for X, and the matrix X can be decomposed into a background component L and a dynamic component S. Because there is a large amount of relevant information between each frame in the background component, the L matrix has a low rank property. At the same time, in cardiac movie imaging, the area where the heart beats is small. After removing the background component L, the remaining S component itself is sparse. In a preferred embodiment, the dynamic component S can be further sparsely transformed, and its sparsity can be further improved. The stronger the sparsity of S, the better the effect of reconstructing the image. Coefficient transformation can be achieved by methods such as Fourier transform, wavelet transform or discrete cosine transform. For example, a one-dimensional Fourier transform is performed on S in the time direction, and the transformed matrix is ​​sparser than S. It should be noted here that L, S and X are data matrices of the same dimension.

[0068] Furthermore, the heartbeat has periodicity in the time direction. Therefore, the MRI reconstruction problem under the constraints of low rank and sparse characteristics can be written as

[0069]

[0070] Among them, λ L and λ S are the regularization factors for low-quality constraints and sparse constraints, respectively. A is the encoding matrix, y is the undersampled k-space data, and D is the sparse transformation matrix. ||||* represents the matrix nuclear norm, which is the sum of the non-zero singular values ​​of the matrix. In mathematics, the sum norm is often used to approximate the matrix rank.

[0071] In order to solve the MRI reconstruction problem, an auxiliary variable X is introduced in this embodiment to represent the cumulative sum of L and S. Then the above optimization problem can be written as

[0072]

[0073] The introduction of auxiliary variables X allows L and S to be searched inexactly in the initial iteration step after the model is successfully constructed, so that the algorithm converges to the optimal L and S more quickly. Fast convergence is very important for the networking of iterative algorithms, because when networking, we need to use a small number of iterative network modules to replace the original large number of iterative steps. In order to understand the equation optimization problem in formula (3), determine the penalty function of formula (3):

[0074]

[0075] Furthermore, the penalty function can be solved by an alternating minimization algorithm to obtain the following three sub-problems. Finally, the target image reconstruction model can be obtained by network-based operation of the following three sub-problems.

[0076]

[0077] In a preferred embodiment, since solving (A * A+ρI) -1 It is time-consuming, and an inexact method can be used to solve the X subproblem. Specifically, the penalty function J is added to the data fidelity term exist The linear expansion is used to obtain the X sub-optimization problem

[0078]

[0079] in, represents the result after J linearization processing, In the representation of , the angle brackets represent the inner product of the two matrices in the angle brackets. This is equivalent to the 0th order term in the Taylor expansion of the function, but here it is a matrix, so it must be written in the form of an inner product.

[0080]

[0081] Furthermore, the subproblem in formula (5) can be rewritten as

[0082]

[0083] Each sub-problem can be solved using the approximate gradient method, based on The above sub-problems can be written as the following iterative steps:

[0084]

[0085] in, is the singular value soft threshold operator

[0086]

[0087] in, is a soft threshold operator, X=UΣV * is the singular value decomposition of X, Σ is the singular value matrix, U and V are the left and right matrices respectively, and * indicates conjugate transpose; is an approximation operator associated with the sparse transformation D. If D is an orthogonal transformation, is a soft threshold operator. If D is a non-orthogonal transformation, then the form of P cannot be directly written as an analytical solution; γ is the update step size, γ = 1 / (1+ηρ). In each iteration step, By cutting the L The singular value of makes L have low rank; in the X iteration step in formula (9), It is the residual in k-space between the reconstructed image and the input k-space data, which is used for data consistency correction.

[0088] Furthermore, although the iterative solution can be expressed as shown in Formula 9, there are still three problems to be solved: First, the iteration takes a long time to converge and the reconstruction speed is slow, especially after the calculation of singular value decomposition is added to the iterative process; second, the regularization parameter is difficult to adjust and requires a lot of time to try, especially when L+S has two components. If λ L Compared to S If λ is too large, the S part of the reconstructed image will not be sparse enough and contain a large number of static components. L Compared to S If it is too small, the L part of the reconstructed image will contain a lot of dynamic information, which is contrary to the model; third, only when the sparse transformation D is orthogonal, The explicit solution of can only be written, otherwise the specific calculation method cannot be obtained. However, the orthogonal sparse transform will greatly limit the selection range of sparse transform, which may lead to limited reconstruction performance.

[0089] Therefore, we network the iterative process in formula (9) and expand it into a low-rank + sparse network (L+S-Net). L+S-Net contains multiple network blocks, each of which contains Figure 2 The three network modules shown are low-rank modules L k , sparse module S k , Data consistency module X k :

[0090]

[0091] Low-rank module L k+1 X k -S k A learnable singular value soft threshold operation (Learned Singular Value Threshold, LSVT) is performed, which is consistent with that shown in formula (10), but the threshold is replaced by a learnable variable:

[0092]

[0093] The threshold of the soft threshold operator is

[0094] T(β,Σ)=β·σ1 (13)

[0095] Where σ1 is the largest singular value and β is a learnable threshold factor. In the sparse module in formula (11), a learnable convolutional neural network C is used to replace the approximation operator In each network block, the threshold factor β, convolutional neural network C, and update step size γ are all independently learnable. During network training, X0 and L0 are initialized to the zero-filled map corresponding to the under-sampled k-space data y, and S0 is initialized with all 0s. These learnable parameters are continuously optimized, and finally high-quality reconstruction results are obtained to generate an image reconstruction model.

[0096] S130. When the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets a preset condition, the model training is completed to obtain a target image reconstruction model.

[0097] The reconstructed image corresponding to the full-sampled K-space data is a standard image with high reconstruction quality. The process of image reconstruction model training is the process of learning and updating the parameters in the model. The optimization of the parameters makes the image output by the model closer and closer to the standard image. When the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets the preset conditions, that is, when the model converges, the target image reconstruction model can be obtained, and the model training process is completed.

[0098] The technical solution of this embodiment processes the fully sampled K-space data samples of the acquired dynamic magnetic resonance image sequence according to a preset under-sampling model to obtain under-sampled K-space data corresponding to the fully sampled K-space data; then, the under-sampled K-space data is input into an image reconstruction model established according to a sub-problem obtained by solving the magnetic resonance image reconstruction problem under the constraints of low-rank characteristics and sparse characteristics for the fully sampled K-space data, and the model is trained; when the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the fully sampled K-space data meets the preset conditions, the model training is completed to obtain the target image reconstruction model. The problem in the prior art that the image reconstruction time is long or the sampling data characteristics cannot be fully utilized for image reconstruction is solved, and the low-rank characteristics and sparse characteristics of the sampling data can be fully utilized while speeding up the image reconstruction speed to establish an image reconstruction model to improve the quality of the reconstructed image.

[0099] Embodiment 2

[0100] Figure 2 This is a flowchart of an image reconstruction method provided in Embodiment 2 of the present invention. This embodiment is applicable to the case of medical image reconstruction. The method can be performed by an image reconstruction device, which can be implemented by software and / or hardware and integrated into a computer device with application development function.

[0101] like Figure 2 As shown, the image reconstruction method includes the following steps:

[0102] S210 , obtaining under-sampled K-space data of a dynamic magnetic resonance image sequence based on a preset under-sampling model.

[0103] Specifically, during the magnetic resonance scanning sampling process, the sampling model can be set according to the sampling speed requirement, which is determined by the preset undersampling operator, Fourier transform operator and coil sensitivity parameter. The preset undersampling operator determines the doubling speed of sampling.

[0104] Specifically, the preset undersampling model can be expressed as y=MFCX(1); wherein M is an undersampling operator, F is a Fourier transform operator, C is a coil sensitivity matrix (sensitivity maps), X is a one-dimensional vector formed by stretching each frame of a fully sampled dynamic magnetic resonance image sequence, and each column is spliced ​​together as a column in the order of the image frame to form an image matrix, and y is the undersampling K-space data. In this step, first, multiple columns of the fully sampled K-space data corresponding to each frame of the dynamic magnetic resonance image sequence are spliced ​​into a single-column data in the order of the columns; then, the single-column data corresponding to all the image frames are spliced ​​in the order of the image frames to obtain X. C can be estimated by algorithms such as ESPIRiT or Walsh, F is a Fourier transform operator, and M is a vector composed of 0 and 1, which determines which points are collected and which points are not collected, and the value of the uncollected points is 0.

[0105] When performing magnetic resonance scanning, undersampled K-space data can be obtained through the above sampling model, for example, obtaining magnetic resonance sampling data of cardiac dynamics.

[0106] S220: Input the under-sampled K-space data into a target image reconstruction model obtained by the image reconstruction model generation method described in any embodiment to obtain a reconstructed image corresponding to the under-sampled K-space data.

[0107] Among them, the target image reconstruction model is a model that performs image reconstruction under the constraints of low-rank characteristics and sparse characteristics, and can reconstruct high-quality images.

[0108] The technical solution of this embodiment obtains a reconstructed image by sampling according to a preset sampling model and inputting the sampled data into a trained image reconstruction model. This solves the problem in the prior art that image reconstruction takes a long time or that the characteristics of the sampled data cannot be fully utilized for image reconstruction. This achieves the goal of speeding up image reconstruction while fully utilizing the low-rank and sparse characteristics of the sampled data to establish an image reconstruction model to improve the quality of the reconstructed image.

[0109] Embodiment 3

[0110] Figure 4This is a schematic diagram of the structure of an image reconstruction model generation device provided in Embodiment 3 of the present invention. This embodiment is applicable to the case where the image reconstruction model is trained using the fully sampled images of the magnetic resonance dynamic image as samples.

[0111] like Figure 4 As shown, the image reconstruction model generation device includes a data preprocessing module 310 , a data input module 320 and a model generation module 330 .

[0112] Among them, the data preprocessing module 310 is used to obtain the full-sampled K-space data of the dynamic magnetic resonance image sequence, and obtain the under-sampled K-space data corresponding to the full-sampled K-space data based on a preset under-sampling model; the data input module 320 is used to input the under-sampled K-space data into, according to the sub-problem obtained by solving the magnetic resonance image reconstruction problem under the constraints of low rank characteristics and sparse characteristics for the full-sampled K-space data, establish an image reconstruction model, and train the model; the model generation module 330 is used to complete the model training and obtain the target image reconstruction model when the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets the preset conditions.

[0113] The technical solution of this embodiment processes the fully sampled K-space data samples of the acquired dynamic magnetic resonance image sequence according to a preset under-sampling model to obtain under-sampled K-space data corresponding to the fully sampled K-space data; then, the under-sampled K-space data is input into an image reconstruction model established according to a sub-problem obtained by solving the magnetic resonance image reconstruction problem under the constraints of low-rank characteristics and sparse characteristics for the fully sampled K-space data, and the model is trained; when the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the fully sampled K-space data meets the preset conditions, the model training is completed to obtain the target image reconstruction model. The problem in the prior art that the image reconstruction time is long or the sampling data characteristics cannot be fully utilized for image reconstruction is solved, and the low-rank characteristics and sparse characteristics of the sampling data can be fully utilized while speeding up the image reconstruction speed to establish an image reconstruction model to improve the quality of the reconstructed image.

[0114] Optionally, the image reconstruction model generating device also includes a model construction module, which is used to establish an image reconstruction model based on sub-problems obtained by solving the magnetic resonance image reconstruction problem under the constraints of low rank and sparse characteristics for the fully sampled K-space data.

[0115] Optionally, the model construction module is specifically used to:

[0116] According to the temporal periodicity of the motion of the magnetic resonance scanning object, the magnetic resonance image reconstruction problem of the full-sampled K-space data is modeled under the constraints of low-rank characteristics and sparse characteristics;

[0117] Setting auxiliary variables of a representation model of the magnetic resonance image reconstruction problem, and determining a penalty function of the representation model based on the auxiliary variables;

[0118] Determine three sub-problems of the representation model according to the penalty function;

[0119] The three sub-problems are solved respectively, and the image reconstruction model is obtained through networking.

[0120] Optionally, the model construction module may also be used to:

[0121] Optimize and rewrite the three sub-problems respectively;

[0122] The approximate gradient method is used to solve the three sub-problems after optimization rewriting, and the iterative representation of the solution of each sub-problem is obtained;

[0123] The iterative representation process is networked to obtain the image reconstruction model.

[0124] Optionally, the model construction module can also be used to optimize the three sub-problems separately, specifically including:

[0125] The three sub-problems are optimized respectively according to the result of the penalty function being expanded and represented by the preset numerical value of the data fidelity term.

[0126] Optionally, the image reconstruction model includes a low-rank network module, a sparse network module and a data consistency network module. The image reconstruction model generation device provided in the embodiment of the present invention can execute the image reconstruction model generation method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0127] Embodiment 4

[0128] Figure 5 This is a structural schematic diagram of an image reconstruction device provided in Embodiment 4 of the present invention, and this embodiment is applicable to the following situations.

[0129] like Figure 5 As shown, the image reconstruction device includes a data acquisition module 410 and an image reconstruction module 420 .

[0130] Among them, the data acquisition module 410 is used to obtain under-sampled K-space data of a dynamic magnetic resonance image sequence obtained based on a preset under-sampling model; the image reconstruction module 420 is used to input the under-sampled K-space data into a target image reconstruction model obtained by the image reconstruction model generation method described in any embodiment, to obtain a reconstructed image corresponding to the under-sampled K-space data.

[0131] The technical solution of this embodiment obtains a reconstructed image by sampling according to a preset sampling model and inputting the sampled data into a trained image reconstruction model. This solves the problem in the prior art that image reconstruction takes a long time or that the characteristics of the sampled data cannot be fully utilized for image reconstruction. This achieves the goal of speeding up image reconstruction while fully utilizing the low-rank and sparse characteristics of the sampled data to establish an image reconstruction model to improve the quality of the reconstructed image.

[0132] The image reconstruction device provided in the embodiment of the present invention can execute the image reconstruction method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0133] Embodiment 5

[0134] Figure 6 A schematic diagram of the structure of a computer device provided in Embodiment 5 of the present invention. Figure 6 A block diagram of an exemplary computer device 12 suitable for use in implementing embodiments of the present invention is shown. Figure 6 The computer device 12 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention. The computer device 12 can be any terminal device with computing capabilities, such as an intelligent controller and server, a mobile phone and other terminal devices.

[0135] like Figure 6 As shown, the computer device 12 is in the form of a general-purpose computing device. The components of the computer device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 that connects various system components (including the system memory 28 and the processing unit 16).

[0136] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0137] The computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0138] The system memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The computer device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 6 not shown, usually called a "hard drive"). Although Figure 6 Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The system memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.

[0139] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. Program modules 42 generally perform the functions and / or methods of the embodiments described herein.

[0140] The computer device 12 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), one or more devices that enable a user to interact with the computer device 12, and / or any device that enables the computer device 12 to communicate with one or more other computing devices (e.g., network cards, modems, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the computer device 12 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with the other modules of the computer device 12 via the bus 18. It should be understood that although Figure 6Not shown, other hardware and / or software modules may be used in conjunction with computer device 12, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0141] The processing unit 16 executes various functional applications and data processing by running the program stored in the system memory 28, for example, implementing the steps of an image reconstruction model generation method provided in the embodiment of the present invention, the method comprising:

[0142] Acquire full-sampled K-space data of a dynamic magnetic resonance image sequence, and obtain under-sampled K-space data corresponding to the full-sampled K-space data based on a preset under-sampling model;

[0143] The under-sampled K-space data is input into an image reconstruction model established according to a sub-problem obtained by solving a magnetic resonance image reconstruction problem under the constraints of low-rank characteristics and sparse characteristics for the fully sampled K-space data, and the model is trained;

[0144] When the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets a preset condition, the model training is completed and the target image reconstruction model is obtained.

[0145] Alternatively, an image reconstruction method provided by an embodiment of the present invention may be implemented, the method comprising:

[0146] Acquiring under-sampled K-space data of a dynamic magnetic resonance image sequence based on a preset under-sampling model;

[0147] The under-sampled K-space data is input into a target image reconstruction model obtained by the image reconstruction model generation method described in any embodiment to obtain a reconstructed image corresponding to the under-sampled K-space data.

[0148] Embodiment 6

[0149] This embodiment 6 provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the image reconstruction model generation method provided in any embodiment of the present invention is implemented, including:

[0150] Acquire full-sampled K-space data of a dynamic magnetic resonance image sequence, and obtain under-sampled K-space data corresponding to the full-sampled K-space data based on a preset under-sampling model;

[0151] The under-sampled K-space data is input into an image reconstruction model established according to a sub-problem obtained by solving a magnetic resonance image reconstruction problem under the constraints of low-rank characteristics and sparse characteristics for the fully sampled K-space data, and the model is trained;

[0152] When the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets a preset condition, the model training is completed and the target image reconstruction model is obtained.

[0153] Alternatively, an image reconstruction method provided by an embodiment of the present invention may be implemented, the method comprising:

[0154] Acquiring under-sampled K-space data of a dynamic magnetic resonance image sequence based on a preset under-sampling model;

[0155] The under-sampled K-space data is input into a target image reconstruction model obtained by the image reconstruction model generation method described in any embodiment to obtain a reconstructed image corresponding to the under-sampled K-space data.

[0156] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0157] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0158] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0159] Computer program code for performing the operation of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0160] It should be understood by those skilled in the art that the modules or steps of the present invention described above can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, optionally, they can be implemented by a program code executable by a computer device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0161] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for generating an image reconstruction model, characterized in that: include: Acquire full-sampled K-space data of a dynamic magnetic resonance image sequence, and obtain under-sampled K-space data corresponding to the full-sampled K-space data based on a preset under-sampling model; The under-sampled K-space data is input into an image reconstruction model to train the model; wherein the image reconstruction model is established based on sub-problems obtained by solving the magnetic resonance image reconstruction problem under the constraints of low-rank and sparse characteristics for the fully sampled K-space data; the process of establishing an image reconstruction model based on sub-problems obtained by solving the magnetic resonance image reconstruction problem under the constraints of low-rank and sparse characteristics for the fully sampled K-space data comprises: modeling the magnetic resonance image reconstruction problem under the constraints of low-rank and sparse characteristics for the fully sampled K-space data according to the periodicity of the movement of the magnetic resonance scanning object in time; setting auxiliary variables of the representation model of the magnetic resonance image reconstruction problem, and determining the penalty function of the representation model based on the auxiliary variables; determining three sub-problems of the representation model according to the penalty function; solving the three sub-problems respectively, and obtaining the image reconstruction model through networking; When the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets a preset condition, the model training is completed and the target image reconstruction model is obtained.

2. The method according to claim 1, characterized in that The three sub-problems are solved respectively and the image reconstruction model is obtained through networking, including: Optimize and rewrite the three sub-problems respectively; The approximate gradient method is used to solve the three sub-problems after optimization rewriting, and the iterative representation of the solution of each sub-problem is obtained; The iterative representation process is networked to obtain the image reconstruction model.

3. The method according to claim 2, characterized in that The three sub-problems are optimized respectively, including: The three sub-problems are optimized respectively according to the result of the penalty function being expanded and represented by the preset numerical value of the data fidelity term.

4. The method according to any one of claims 1 to 3, characterized in that: The image reconstruction model includes a low-rank network module, a sparse network module and a data consistency network module.

5. An image reconstruction method, characterized in that: include: Acquiring under-sampled K-space data of a dynamic magnetic resonance image sequence based on a preset under-sampling model; The under-sampled K-space data is input into a target image reconstruction model obtained by the image reconstruction model generation method according to any one of claims 1 to 4 to obtain a reconstructed image corresponding to the under-sampled K-space data.

6. An image reconstruction model generation device, characterized in that: include: A data preprocessing module, used for acquiring full-sampled K-space data of a dynamic magnetic resonance image sequence, and obtaining under-sampled K-space data corresponding to the full-sampled K-space data based on a preset under-sampling model; A data input module is used to input the under-sampled K-space data into an image reconstruction model to train the model; wherein the image reconstruction model is established based on sub-problems obtained by solving the magnetic resonance image reconstruction problem under the constraints of low-rank and sparse characteristics for the fully sampled K-space data; the process of establishing the image reconstruction model based on the sub-problems obtained by solving the magnetic resonance image reconstruction problem under the constraints of low-rank and sparse characteristics for the fully sampled K-space data comprises: modeling the magnetic resonance image reconstruction problem under the constraints of low-rank and sparse characteristics for the fully sampled K-space data according to the periodicity of the movement of the magnetic resonance scanning object in time; setting auxiliary variables of the representation model of the magnetic resonance image reconstruction problem, and determining the penalty function of the representation model based on the auxiliary variables; determining three sub-problems of the representation model according to the penalty function; solving the three sub-problems respectively, and obtaining the image reconstruction model through networking; The model generation module is used to complete the model training and obtain the target image reconstruction model when the mean square error between the reconstructed image generated by the image reconstruction model and the reconstructed image corresponding to the full-sampled K-space data meets the preset conditions.

7. An image reconstruction device, characterized in that: include: A data acquisition module, used for acquiring under-sampling K-space data of a dynamic magnetic resonance image sequence obtained based on a preset under-sampling model; An image reconstruction module is used to input the under-sampled K-space data into a target image reconstruction model obtained by the image reconstruction model generation method according to any one of claims 1 to 4 to obtain a reconstructed image corresponding to the under-sampled K-space data.

8. A computer device, characterized in that: The computer device comprises: one or more processors; A memory 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 implement the image reconstruction model generation method or the image reconstruction method as described in any one of claims 1 to 5.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the image reconstruction model generation method or the image reconstruction method as described in any one of claims 1 to 5 is implemented.

Images