A method, apparatus, device, and medium for optimizing magnetic resonance imaging

Abstract

The application discloses a method, device, equipment and medium for optimizing magnetic resonance imaging, relates to the field of medical image processing, and solves the problem of poor magnetic resonance imaging effect caused by the fact that the frame similarity at the edges based on multi-scale gradient directions between adjacent frames is not considered in the image reconstruction process of the prior art. Technical scheme points are as follows: a T frame image is extracted from an image data stream generated by magnetic resonance scanning, wherein T is a positive integer greater than or equal to 3; starting from the third frame image, a reconstructed image of a second frame image is used as a pre-reconstruction image of the third frame image to serve as a reference object for image registration until a reconstructed image of a T-1 frame image is used as a reference object for the pre-reconstruction image of the T frame image, and the pre-reconstruction image of the T frame image is subjected to image registration according to the reconstructed image of the T-1 frame image to obtain a registration result of the T frame image; and the registration result of the T frame image is subjected to image reconstruction to obtain a magnetic resonance imaging result.

Description

Technical Field

[0001] This application relates to the field of medical image processing, and more specifically, to a method, apparatus, device, and medium for optimizing magnetic resonance imaging. Background Technology

[0002] Magnetic resonance imaging (MRI) utilizes the principle of nuclear magnetic resonance, based on the varying attenuation of released energy within different structural environments of matter. By applying an external gradient magnetic field, it detects the emitted electromagnetic waves to create images of the object's internal structure. It not only provides rich tissue contrast but is also harmless to the human body, thus becoming a powerful tool for clinical medical diagnosis. Dynamic MRI aids in computer-aided diagnosis and vision-guided surgery, greatly promoting the rapid development of medicine, neurophysiology, and cognitive neuroscience.

[0003] Compressed sampling techniques can reduce the amount of measurement data while reconstructing the original image, and can be applied to dynamic magnetic resonance imaging (MRI). Dynamic MRI based on compressed sampling can be divided into offline and online methods. Offline methods utilize the sparsity and low rank of all frames, achieving relatively high reconstruction performance, but their limitation lies in the time-consuming reconstruction process. Online methods, on the other hand, do not rely on subsequent frames for reconstruction, instead utilizing prior information from adjacent frames. However, due to the lack of additional information from more frames and the impact of error accumulation, the accuracy of image reconstruction is relatively low. Existing methods (such as DTV) primarily use the first frame to reconstruct the remaining images frame by frame, relying on the assumption of similar intensity distributions between adjacent frames. They do not consider the frame similarity at edges based on multi-scale gradient directions between adjacent frames, leading to poor MRI results. Summary of the Invention

[0004] The purpose of this application is to provide a method, apparatus, device and medium for optimizing magnetic resonance imaging. This application solves the problem that the existing technology does not take into account the frame similarity at the edges of adjacent frames based on multi-scale gradient directions during the image reconstruction process, resulting in poor magnetic resonance imaging effect.

[0005] A first aspect of this application provides a method for optimizing magnetic resonance imaging, the method comprising:

[0006] T frames of images are extracted from the image data stream generated by magnetic resonance scanning; where T is a positive integer greater than or equal to 3.

[0007] Starting from the third frame, the reconstructed image of the second frame is used as the reference object for image registration of the pre-reconstructed image of the third frame, until the reconstructed image of the (T-1)th frame is used as the reference object for the pre-reconstructed image of the Tth frame. The pre-reconstructed image of the Tth frame is then registered based on the reconstructed image of the (T-1)th frame to obtain the registration result of the Tth frame.

[0008] Image reconstruction is performed on the registration results of the T-th frame image to obtain the magnetic resonance imaging results.

[0009] In one implementation of the first aspect of this application, before image registration is performed using the reconstructed image of the second frame as the reference object for the pre-reconstructed image of the third frame, starting from the third frame image, until the reconstructed image of the (T-1)th frame image is used as the reference object for the pre-reconstructed image of the Tth frame image, and image registration is performed on the pre-reconstructed image of the Tth frame image based on the reconstructed image of the (T-1)th frame image to obtain the registration result of the Tth frame image, the method further includes:

[0010] Image pre-reconstruction is performed on the first frame image and the second frame image respectively to obtain the pre-reconstructed image of the first frame image and the pre-reconstructed image of the second frame image;

[0011] The pre-reconstructed image of the first frame is used as the reference object for image registration of the pre-reconstructed image of the second frame. The pre-reconstructed image of the second frame is registered based on the pre-reconstructed image of the first frame to obtain the registration result of the second frame.

[0012] The registration results of the second frame image are used to reconstruct the image, resulting in the reconstructed image of the second frame image.

[0013] In one implementation of the first aspect of this application, the second to T-th frames are sequentially processed by image pre-reconstruction, image registration and image reconstruction.

[0014] In one implementation of the first aspect of this application, the expression for image pre-reconstruction processing is: Where X represents all frames; B represents the measurement value of all frames; F U Let represent the spatiotemporal sensing matrix, ε represent the upper bound of the measurement data fitting, and Φ represent the sparse transformation.

[0015] In one implementation of the first aspect of this application, the expression for image registration processing is: Where G represents an isotropic Gaussian filter,

[0016] σ

[0017] n represents the number of iterations. σ represents the displacement field. 2This represents a constant controlling the deformation intensity and speed, where x1 represents the first frame image, x... k Let P represent the image of the k-th frame. n This represents the value of the nth iteration.

[0018] In one implementation of the first aspect of this application, the expression for image reconstruction processing is: Among them, F U，k Let X represent the spatiotemporal perception matrix of the k-th frame image. k B represents the image matrix of the k-th frame. k This represents the measurement value of the k-th frame image. λ1 and λ2 represent parameters that control the intensity of regularization, and SC represents a similarity constraint.

[0019] The expression for the similarity constraint is: Where i takes the value 1 or 2. When i is 1, M1(X) represents the multi-scale gradient operator in the horizontal direction. When i is 2, M2(X) represents the multi-scale gradient operator in the vertical direction. D represents the transpose of the matrix.

[0020] In one implementation of the first aspect of this application, similarity constraints are defined based on multi-scale gradient vectors, and the expression for the multi-scale gradient vectors is: Where G1(X) and G2(X) represent the tangent direction and normal direction at a certain edge position, respectively;

[0021] The expressions for the multi-scale gradient operators in the horizontal and vertical directions are: Where J represents the scaling factor, r and s represent the row index and column index of the image matrix, respectively, vec represents matrix vectorization operation, X(r+j,s) represents the image matrix with row index r+j and column index s, X(r,s+j) represents the image matrix with row index r and column index s+j, and X(r,s) represents the image matrix with row index r and column index s.

[0022] A second aspect of this application provides an apparatus for optimizing magnetic resonance imaging, the apparatus comprising:

[0023] The image extraction module is used to extract T frames of images from the image data stream generated by magnetic resonance scanning; where T is a positive integer greater than or equal to 3.

[0024] The image registration module is used to perform image registration starting from the third frame image, using the reconstructed image of the second frame image as the reference object for the pre-reconstructed image of the third frame image, until the reconstructed image of the (T-1)th frame image is used as the reference object for the pre-reconstructed image of the Tth frame image, and to perform image registration on the pre-reconstructed image of the Tth frame image based on the reconstructed image of the (T-1)th frame image to obtain the registration result of the Tth frame image.

[0025] The image reconstruction module is used to reconstruct the image from the registration results of the T-th frame to obtain the magnetic resonance imaging results.

[0026] A third aspect of this application provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein the computer program, when executed by the processor, implements the steps of an optimized magnetic resonance imaging method as provided in the first aspect of this application.

[0027] A fourth aspect of this application provides a computer-readable storage medium storing a computer program, wherein when executed by a processor, the computer program implements the steps of an optimized magnetic resonance imaging method as provided in the first aspect of this application.

[0028] Compared with the prior art, this application has the following beneficial effects:

[0029] In the method for optimizing magnetic resonance imaging provided in this application, the method considers the frame similarity between adjacent frames based on multi-scale edge orientation. The reconstructed image of the previous frame is used as a reference image to improve the reconstruction of the next frame. That is, the first frame image obtained by sampling is used as a reference image to improve the reconstruction of the second frame. Then, the reconstructed image of the second frame obtained earlier is used as a reference image to improve the reconstruction of the third frame, and so on, until the reconstructed image of the (T-1)th frame is used as a reference image to improve the reconstruction of the Tth frame, thereby making the reconstruction effect of magnetic resonance imaging better. Attached Figure Description

[0030] The accompanying drawings, which are included to provide a further understanding of the embodiments of this application and form part of this application, do not constitute a limitation on the embodiments of this application. In the drawings:

[0031] Figure 1 A flowchart illustrating an optimized magnetic resonance imaging method provided in this application embodiment;

[0032] Figure 2 A schematic diagram of the framework of the magnetic resonance imaging process provided in the embodiments of this application;

[0033] Figure 3The reconstruction result of the 7th frame in the brain imaging experiment of rigid motion provided in the embodiments of this application;

[0034] Figure 4 RLNE value map of reconstructed all frames in the brain imaging experiment provided in the embodiments of this application;

[0035] Figure 5 The reconstruction result of the 7th frame in the non-rigid motion cardiac imaging experiment provided in the embodiments of this application;

[0036] Figure 6 RLNE value map of reconstructed all frames in a cardiac imaging experiment provided in this application embodiment;

[0037] Figure 7 A schematic diagram of an optimized magnetic resonance imaging apparatus provided in an embodiment of this application;

[0038] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0039] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the embodiments and accompanying drawings. The illustrative embodiments and descriptions of this application are only for explaining this application and are not intended to limit this application.

[0040] It should be noted that the terms "comprising" or "may include" used in the various embodiments of this application indicate the presence of the claimed function, operation, or element, and do not limit the addition of one or more functions, operations, or elements. Furthermore, as used in the various embodiments of this application, the terms "comprising," "having," and their cognates are intended only to indicate a specific feature, number, step, operation, element, component, or combination of the foregoing, and should not be construed as primarily excluding the presence of one or more other features, numbers, steps, operations, elements, components, or combinations of the foregoing, or adding one or more combinations of the foregoing.

[0041] It should be understood that terms such as "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0042] Please refer to Figure 1 , Figure 1A flowchart illustrating an optimized magnetic resonance imaging method provided in this application embodiment is shown below. Figure 1 As shown, the method includes:

[0043] 101. Extract T frames of images from the image data stream generated by magnetic resonance scanning; where T is a positive integer greater than or equal to 3.

[0044] In this embodiment, the image data stream is obtained by sampling the magnetic resonance scan using a high-sampling method, which is a conventional technique in this field and will not be described in detail here. Similarly, extracting T-frame images from the image data stream is also a conventional technique and will not be described in detail here.

[0045] 102. Starting from the third frame, the reconstructed image of the second frame is used as the reference object for image registration of the pre-reconstructed image of the third frame, until the reconstructed image of the (T-1)th frame is used as the reference object for the pre-reconstructed image of the Tth frame. Based on the reconstructed image of the (T-1)th frame, the pre-reconstructed image of the Tth frame is registered to obtain the registration result of the Tth frame.

[0046] like Figure 2 As shown, at least three frames are required to complete image reconstruction from one frame, which serves as a reference image to improve the reconstruction of the next frame. This is why T is a positive integer greater than or equal to 3. Image reconstruction is... Figure 2 The CoS reconstruction in this embodiment considers the problem of frame similarity between adjacent frames based on multi-scale edge orientation. Starting from the third frame, the previous frame is used as a reference image to improve the reconstruction of the next frame. First, the reconstructed image of the second frame obtained by high sampling is used as a reference image to improve the reconstruction of the third frame. Then, the reconstructed image of the third frame is used as a reference image to improve the reconstruction of the fourth frame, and so on, until the reconstructed image of the (T-1)th frame is used as a reference image to improve the reconstruction of the Tth frame.

[0047] In some embodiments, before performing step 102, the method further includes: performing image pre-reconstruction on the first frame image and the second frame image respectively to obtain a pre-reconstructed image of the first frame image and a pre-reconstructed image of the second frame image; using the pre-reconstructed image of the first frame image as a reference object for image registration on the pre-reconstructed image of the second frame image, and performing image registration on the pre-reconstructed image of the second frame image based on the pre-reconstructed image of the first frame image to obtain a registration result of the second frame image; and performing image reconstruction on the registration result of the second frame image to obtain a reconstructed image of the second frame image.

[0048] Specifically, the pre-reconstructed image of the first frame is used as a reference image to improve the reconstruction of the second frame. Since the first frame does not have a previous frame, this embodiment uses the image obtained through image pre-reconstruction as the second reference image to improve the reconstruction of the second frame. For the third frame, the pre-reconstructed image of the first frame provided in this embodiment is used as a reference image to improve the reconstruction of the second frame, and the CoS reconstructed image obtained from the second frame is used as a reference image to improve the reconstruction of the third frame.

[0049] Finally, by Figure 2 As can be seen from the content shown, the method provided in this embodiment requires image pre-reconstruction, image registration and image reconstruction processing to be performed sequentially on each frame of the image from the second frame to the Tth frame.

[0050] Accordingly, image pre-reconstruction is based on traditional CS (Continuous Reconstruction) for pre-reconstructing each frame. The expression for its pre-reconstruction model is: Where X represents all frames; B represents the measurement value of all frames; F U Let represent the spatiotemporal sensing matrix, ε represent the upper bound of the measurement data fitting, and Φ represent the sparse transformation.

[0051] For the image registration process, the first frame image x1 obtained from the pre-reconstruction is used to register other frame images x1. k Registration is performed. Since the dynamic image sequence has consistent contrast, image registration uses a demos algorithm based on grayscale intensity, which can be expressed as: Where G represents an isotropic Gaussian filter

[0052] σ

[0053] Filter, where n represents the number of iterations. σ represents the displacement field. 2 This represents a constant controlling the deformation intensity and speed, where x1 represents the first frame image, x... k Let P represent the image of the k-th frame. n This represents the value of the nth iteration.

[0054] For the CoS reconstruction processing part, the expression for image reconstruction processing is:

[0055] Among them, F U，k Let X represent the spatiotemporal perception matrix of the k-th frame image. k B represents the image matrix of the k-th frame. k This represents the measurement value of the k-th frame image. λ1 and λ2 represent parameters that control the intensity of regularization, and SC represents a similarity constraint.

[0056] The expression for the similarity constraint is: Where i takes the value 1 or 2. When i is 1, M1(X) represents the multi-scale gradient operator in the horizontal direction. When i is 2, M2(X) represents the multi-scale gradient operator in the vertical direction. D represents the transpose of the matrix.

[0057] Since this invention focuses more on gradient direction than on the intensity distribution of existing technologies, similarity constraints are defined based on multi-scale gradient vectors, where the expression for the multi-scale gradient vector is:

[0058] Where G1(X) and G2(X) represent the tangent direction and normal direction at a certain edge position, respectively;

[0059] The expressions for the multi-scale gradient operators in the horizontal and vertical directions are:

[0060] Where J represents the scaling factor, r and s represent the row index and column index of the image matrix, respectively, vec represents matrix vectorization operation, X(r+j,s) represents the image matrix with row index r+j and column index s, X(r,s+j) represents the image matrix with row index r and column index s+j, and X(r,s) represents the image matrix with row index r and column index s.

[0061] The multi-scale scheme provided in this embodiment can capture the gradient response on the planar region of the image. At the same time, the noise in the image is also suppressed to a certain extent. Therefore, the reconstruction effect after CoS reconstruction provided in this embodiment is better.

[0062] 103. The registration results of the T-th frame image are used to reconstruct the image and obtain the magnetic resonance imaging results.

[0063] The method for optimizing magnetic resonance imaging provided in the above embodiments considers the frame similarity between adjacent frames based on multi-scale edge orientation. The reconstructed image of the previous frame is used as a reference image to improve the reconstruction of the next frame. That is, the first frame image obtained by sampling is used as a reference image to improve the reconstruction of the second frame. Then, the reconstructed image of the second frame obtained earlier is used as a reference image to improve the reconstruction of the third frame, and so on, until the reconstructed image of the (T-1)th frame is used as a reference image to improve the reconstruction of the Tth frame, thereby making the reconstruction effect of magnetic resonance imaging better.

[0064] Secondly, such as Figure 3 As shown, where, Figure 3Figure (a) is the standard image, (b) is the reconstructed image obtained using the method in Reference 1, (e) is the error map corresponding to this algorithm, (c) is the reconstructed image obtained using Reference 2, (f) is the error map corresponding to this algorithm, (d) is the reconstructed image obtained using the method proposed in this invention, and (g) is the error map corresponding to this algorithm. Comparative Analysis Figure 3 The reconstruction maps (b), (c), (d) and error maps (e), (f), (g) show that the reconstruction effect of the method proposed in this invention is better. This is because the method further considers the frame similarity between adjacent frames based on multi-scale edge orientation.

[0065] like Figure 4 As shown, by this Figure 4 As can be seen, the method proposed in this embodiment has a lower RLNE value, and therefore the reconstruction effect is better.

[0066] like Figure 5 As shown, (a) is the standard image, (b) is the reconstructed image obtained using the method in Reference 1, (e) is the error map corresponding to this algorithm, (c) is the reconstructed image obtained using Reference 2, (f) is the error map corresponding to this algorithm, (d) is the reconstructed image obtained using the method proposed in this invention, and (g) is the error map corresponding to this algorithm. Comparative analysis of the reconstructed images (b), (c), and (d) and the error maps (e), (f), and (g) reveals that the method proposed in this invention achieves superior reconstruction results. Figure 6 As shown in the figure, the method proposed in this invention has a lower RLNE value.

[0067] References 1 and 2 mentioned above are as follows: Reference 1: [1] C. Chen, Y. Li, L. Axel, et al. Real time dynamic MRI by exploiting spatial and temporal sparsity[J]. Magnetic Resonance Imaging, 2016, 34(4):473–482. Reference 12: [2] Liu Liangyou, Li Zhaotong, Zhang Zeru, et al. Compressed sensing magnetic resonance diffusion tensor imaging based on reference images[J], Chinese Journal of Medical Physics, 2021, 38(3):323-326.

[0068] Please refer to Figure 7 , Figure 7 A schematic diagram of an optimized magnetic resonance imaging apparatus provided in this application embodiment is shown below. Figure 7 As shown, the device includes:

[0069] The image extraction module 710 is used to extract T-frame images from the image data stream generated by magnetic resonance scanning; where T is a positive integer greater than or equal to 3.

[0070] The image registration module 720 is used to perform image registration starting from the third frame image, using the reconstructed image of the second frame image as the reference object for the pre-reconstructed image of the third frame image, until the reconstructed image of the (T-1)th frame image is used as the reference object for the pre-reconstructed image of the Tth frame image, and to perform image registration on the pre-reconstructed image of the Tth frame image based on the reconstructed image of the (T-1)th frame image to obtain the registration result of the Tth frame image;

[0071] The image reconstruction module 730 is used to reconstruct the image from the registration result of the T-th frame image to obtain the magnetic resonance imaging result.

[0072] Accordingly, the device for optimizing magnetic resonance imaging provided in this application considers the frame similarity between adjacent frames based on multi-scale edge orientation. The reconstructed image of the previous frame is used as a reference image to improve the reconstruction of the next frame. That is, the first frame image obtained by sampling is used as a reference image to improve the reconstruction of the second frame. Then, the reconstructed image of the second frame obtained earlier is used as a reference image to improve the reconstruction of the third frame, and so on, until the reconstructed image of the (T-1)th frame is used as a reference image to improve the reconstruction of the Tth frame, thereby making the reconstruction effect of magnetic resonance imaging better.

[0073] Please refer to Figure 8 , Figure 8 This is a schematic diagram of an electronic device provided in an embodiment of this application. The electronic device 800 includes a processor 810, a memory 820, a communication interface 830, and at least one communication bus for connecting the processor 810, the memory 820, and the communication interface 830. The memory 820 includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (PROM), or portable read-only memory (CD-ROM), and is used for related instructions and data.

[0074] The communication interface 830 is used to receive and send data. The processor 810 can be one or more CPUs. If the processor 810 is a CPU, the CPU can be a single-core CPU or a multi-core CPU. The processor 810 in the electronic device 800 is used to read one or more programs 821 stored in the memory 820 and perform the following operations: extract T frames of images from the image data stream generated by magnetic resonance scanning; where T is a positive integer greater than or equal to 3; starting from the third frame of images, use the reconstructed image of the second frame of images as the reference object for image registration of the pre-reconstructed image of the third frame of images, until the reconstructed image of the (T-1)th frame of images is used as the reference object for the pre-reconstructed image of the Tth frame of images, and perform image registration of the pre-reconstructed image of the Tth frame of images based on the reconstructed image of the (T-1)th frame of images to obtain the registration result of the Tth frame of images; perform image reconstruction on the registration result of the Tth frame of images to obtain the magnetic resonance imaging result.

[0075] It should be noted that the specific implementation of each operation can be described above. Figure 1 The corresponding description of the method embodiments shown indicates that the electronic device 800 can be used to execute an optimized magnetic resonance imaging method according to the above method embodiments of this application, which will not be described in detail here.

[0076] In embodiments of this disclosure, a computer-readable storage medium is also provided. This computer-readable storage medium is a memory device in a computer device used to store programs and data. It is understood that the computer-readable storage medium here can include both built-in storage media in the computer device and extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the operating system of a terminal. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. One or more instructions stored in the computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the method for optimizing magnetic resonance imaging in the above embodiments. Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0077] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above description is only a specific embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for optimizing magnetic resonance imaging, characterized in that, The methods include: Extract T-frame images from the image data stream generated by magnetic resonance scanning; Where T is a positive integer greater than or equal to 3; Image pre-reconstruction is performed on the first frame image and the second frame image respectively to obtain the pre-reconstructed images of the first frame image and the second frame image; the pre-reconstructed image of the first frame image is used as the reference object for image registration of the pre-reconstructed image of the second frame image, and image registration is performed on the pre-reconstructed image of the second frame image based on the pre-reconstructed image of the first frame image to obtain the registration result of the second frame image; image reconstruction is performed on the registration result of the second frame image to obtain the reconstructed image of the second frame image. For the images from the second frame to the Tth frame, image pre-reconstruction, image registration, and image reconstruction are performed sequentially; the expression for image reconstruction is as follows: ,in, This represents the spatiotemporal perception matrix of the k-th frame image. The image matrix represents the image of the k-th frame. This represents the measurement value of the k-th frame image. λ1 and λ2 represent parameters that control the intensity of regularization, and SC represents a similarity constraint. The expression for the similarity constraint is: Where i takes the value 1 or 2, when i is 1, This represents the multi-scale gradient operator in the horizontal direction, when i is 2. This represents the multi-scale gradient operator in the vertical direction; D represents the transpose of the matrix. Starting from the third frame, the reconstructed image of the second frame is used as the reference object for image registration of the pre-reconstructed image of the third frame, until the reconstructed image of the (T-1)th frame is used as the reference object for the pre-reconstructed image of the Tth frame. The pre-reconstructed image of the Tth frame is then registered based on the reconstructed image of the (T-1)th frame to obtain the registration result of the Tth frame. Image reconstruction is performed on the registration results of the T-th frame image to obtain the magnetic resonance imaging results.

2. The method for optimizing magnetic resonance imaging according to claim 1, characterized in that, The expression for image pre-reconstruction processing is: ,in, Indicates all frames; This represents the measurements from all frames. Represents the spatiotemporal perception matrix. Φ represents the upper bound of the fit to the measurement data, and Φ represents the sparse transformation.

3. The method for optimizing magnetic resonance imaging according to claim 1, characterized in that, The expression for image registration processing is: ,in, This represents an isotropic Gaussian filter, where n represents the number of iterations. Represents the displacement field. Represents the constants that control the deformation strength and rate. This represents the first frame of the image. This represents the image of the k-th frame. This represents the value of the nth iteration.

4. The method for optimizing magnetic resonance imaging according to claim 1, characterized in that, Similarity constraints are defined based on multi-scale gradient vectors. The expression for the multi-scale gradient vector is as follows: ,in, and These represent the tangent direction and the normal direction at a certain edge location, respectively. The expressions for the multi-scale gradient operators in the horizontal and vertical directions are: Where J represents the scaling factor, and r and s represent the row and column indices of the image matrix, respectively. This represents matrix vectorization operations. This represents the image matrix with row index r+j and column index s. This represents the image matrix with row index r and column index s+j. This represents the image matrix with row index r and column index s.

5. An apparatus for optimizing magnetic resonance imaging, characterized in that, The device includes: The image extraction module is used to extract T-frame images from the image data stream generated by magnetic resonance scanning; Where T is a positive integer greater than or equal to 3; The image registration module is used to perform image pre-reconstruction on the first frame image and the second frame image respectively, to obtain pre-reconstructed images of the first frame image and the second frame image; using the pre-reconstructed image of the first frame image as the reference object for image registration of the pre-reconstructed image of the second frame image, and performing image registration on the pre-reconstructed image of the second frame image based on the pre-reconstructed image of the first frame image, to obtain the registration result of the second frame image; performing image reconstruction on the registration result of the second frame image, to obtain the reconstructed image of the second frame image; performing image pre-reconstruction, image registration, and image reconstruction processing sequentially on the second frame to the Tth frame image; wherein, the expression for the image reconstruction processing is: ,in, This represents the spatiotemporal perception matrix of the k-th frame image. The image matrix represents the image of the k-th frame. This represents the measurement value of the k-th frame image. λ1 and λ2 represent parameters that control the intensity of regularization, and SC represents a similarity constraint. The expression for the similarity constraint is: Where i takes the value 1 or 2, when i is 1, This represents the multi-scale gradient operator in the horizontal direction, when i is 2. denoted by , which represents the multi-scale gradient operator in the vertical direction; D represents the transpose of the matrix; starting from the third frame image, the reconstructed image of the second frame image is used as the reference object for image registration of the pre-reconstructed image of the third frame image, until the reconstructed image of the (T-1)th frame image is used as the reference object for the pre-reconstructed image of the Tth frame image, and image registration of the pre-reconstructed image of the Tth frame image is performed based on the reconstructed image of the (T-1)th frame image to obtain the registration result of the Tth frame image; The image reconstruction module is used to reconstruct the image from the registration results of the T-th frame to obtain the magnetic resonance imaging results.

6. An electronic device, characterized in that, The electronic device includes a processor, a memory, and a computer program stored in the memory and executable by the processor, wherein when the computer program is executed by the processor, it implements the steps of an optimized magnetic resonance imaging method as claimed in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, it implements the steps of a method for optimizing magnetic resonance imaging as described in any one of claims 1 to 4.

Images