Medical MRI (Magnetic Resonance Imaging) image denoising method, device, system, equipment, medium and product

Abstract

The invention discloses a medical MRI image denoising method, device, system and equipment, a medium and a product, and relates to the technical field of image enhancement. The method comprises the following steps: firstly, converting a medical MRI image signal into an image domain to obtain three-dimensional image data, then selecting three-dimensional image sub-blocks, searching M three-dimensional image similar sub-blocks which are most similar in a predefined neighborhood, and then carrying out vectorization straightening processing and superposition on the sub-blocks to obtain an original matrix Y; then assuming that an original matrix Y is equal to X + N + S and solving the minimization objective function to obtain a de-noising matrix X, reconstructing according to the de-noising matrix X to obtain new sub-blocks corresponding to the sub-blocks, and replacing the new and old sub-blocks to obtain new three-dimensional image data subjected to image de-noising. Therefore, the problem that the denoised image is blurred when the prior art is applied to MRI denoising can be solved, the quality of the medical MRI image is improved, and subsequent image analysis is facilitated.

Description

Technical Field

[0001] The present invention belongs to the field of image enhancement technology, and specifically relates to a medical MRI image denoising method, device, system, equipment, medium and product. Background Art

[0002] Magnetic Resonance Imaging (MRI) has the characteristics of no ionization, no radiation and high contrast for soft tissues, and plays an important role in human medicine. With the development of technology, multi-nuclear MRI has also been developed accordingly. In multi-nuclear MRI, the radionuclides commonly used are 23 As one of the most important electrolytes in human physiology, sodium ion ( 23 Na) plays a vital role in osmotic regulation and cell physiology. It is maintained across the membrane by active transport via the sodium-potassium pump (Na+ / K+-ATPase). 23 The sodium concentration gradient is more than 10-fold different between the intracellular (10-15 mM) and extracellular (140-150 mM) compartments. Under pathological conditions, impaired cellular metabolism or membrane integrity may disrupt sodium pump function, leading to tissue-specific 23 Disrupted Na homeostasis and abnormal ion concentrations. These changes in tissue sodium concentration (TSC) reflect underlying metabolic dysfunction and have been observed in a variety of neurological diseases, including brain tumors, stroke, multiple sclerosis, and epilepsy. 23 Na MRI can noninvasively measure TSC, providing a valuable tool for studying these pathologies. However, compared to proton ( 1 H) Imaging compared to 23 Na MRI faces greater technical challenges due to its inherently low signal-to-noise ratio (SNR) resulting from its lower gyromagnetic ratio, lower in vivo concentration, and shorter transverse relaxation time.

[0003] In order to further improve the quality of magnetic resonance images, especially to improve 23 To improve the quality of MRI, various post-processing and reconstruction strategies have been proposed. Broadly speaking, these denoising methods fall into two categories: deep learning-based methods and traditional non-learning techniques.

[0004] Deep learning models use the nonlinear mapping and automatic feature extraction capabilities of neural networks to discover latent structures in large datasets. For example, Adlung et al. used a u-net-based architecture to reconstruct highly undersampled images of patients with acute ischemic stroke. 23Na MRI data, reducing acquisition time while improving signal-to-noise ratio and TSC quantification accuracy. Similarly, Bakeret et al. trained a Convolutional Neural Network (CNN) on synthetically corrupted low-SNR data generated by adding Gaussian noise to the high-quality 1H k-space of the fastMRI dataset. However, these supervised networks require a large number of paired noise-free datasets and often have difficulty generalizing. Although self-supervised denoising methods have been developed that are trained only on noisy images, they rely heavily on predefined noise models. In multi-channel MRI, phased array coils are often used to improve the inherently low signal-to-noise ratio; however, inter-channel coupling introduces complex noise patterns that violate the noise assumptions of self-supervised denoising methods. To date, no such methods have been successfully applied to MRI denoising in complex scenarios, especially 23 Na MRI denoising.

[0005] Traditional non-learning-based denoising techniques usually rely on manual priors to formulate noise suppression algorithms, which perform well in conventional MRI denoising. However, when extended to multi-core MRI denoising scenarios, the denoising capability is insufficient. For example, Madelin et al. were the first to apply Compressed Sensing (CS) to 23 One group of researchers working on Na MRI has halved the acquisition time while retaining some degree of reconstruction accuracy. However, this approach often results in residual blurring and loss of detail, limiting its applicability for quantitative sodium imaging. Lachner et al. and Gnahm et al.

[15] have used a 4-D MRI to quantify the Na MRI image. 1 Anatomical priors from H MRI have been incorporated into the CS framework to address this issue, which improves fidelity but requires additional multimodal data acquisition and image registration. Benkhedah et al. utilized an adaptive combined reconstruction method to mitigate inter-coil channel noise correlation, but this requires additional noisy scans and increases total scan time. More recently, Christensen et al. applied global-local high-order singular value decomposition (GL-HOSVD) and tensor Marchenko-Pastur principal component analysis (tMPPCA) to denoise various x-core human data. While effective, both methods introduced subtle artifacts and varying degrees of image smoothing.

[0006] tMPPCA (tensor Marchenko-Pastur principal component analysis) is an improved denoising method based on high-order singular value decomposition (HOSVD). By introducing the Marchenko-Pastur distribution to automatically estimate signal rank, it reduces the need for user-defined parameters and improves the objectivity and robustness of denoising. This method effectively exploits data redundancy by recursively decomposing the tensor structure of multidimensional data, making it particularly suitable for small data blocks and high-dimensional data (such as multi-echo diffusion MRI). However, tMPPCA still has limitations in MRI denoising: although it can significantly reduce background noise, residual analysis shows that brain signal distribution may be biased (such as abnormally enhanced signal intensity in ventricular regions) and artifacts are introduced in some slices. Furthermore, to address the spatially varying noise caused by parallel imaging reconstruction in clinical data, tMPPCA requires additional noise level maps or local estimation methods; otherwise, it may not fully adapt to complex noise distributions.

[0007] GL-HOSVD (Global-Local High-Order Singular Value Decomposition) is a hybrid denoising algorithm that combines global pre-filtering with local block processing. It uses the global HOSVD stage to guide the local HOSVD stage and reduce the streak artifacts produced by local methods under low signal-to-noise ratio. This method performs well in the case of diffusion MRI denoising, but it is not suitable for 23 There are obvious defects in Na MRI denoising: experimental results show that the distribution of brain signals may change after denoising (such as abnormal signals in the ventricular region), and brain contour deviations still exist in the residual image, indicating that the denoising process introduces systematic errors. In addition, GL-HOSVD relies on user-defined threshold parameters (such as k_global and k_local), which need to be adjusted empirically. In clinical data, the noise characteristics may be complicated by preprocessing steps (such as parallel imaging or geometric correction, etc.), resulting in limited parameter generalization. At the same time, 23 Artifacts have been observed in Na MRI denoising, further limiting its reliability.

[0008] In summary, the existing tMPPCA and GL-HOSVD techniques have excellent performance for conventional MRI denoising, but in multi-core MRI denoising (especially 23 In the process of MRI denoising, different degrees of image blurring will occur, which means that although the noise can be removed, the denoising ability for the signal area is limited. Therefore, it is necessary to improve the existing MRI denoising algorithm so that it can not only remove the noise of conventional MRI (such as 1 H MRI) can be effectively denoised, and it can also effectively denoise higher and more complex multi-core MRI. Summary of the Invention

[0009] The purpose of the present invention is to provide a medical MRI image denoising method, apparatus, system, computer equipment, computer-readable storage medium and computer program product to solve the problems of blurred denoised images and limited denoising capabilities for signal areas when existing tMPPCA and GL-HOSVD technologies are applied to MRI denoising.

[0010] In order to achieve the above object, the present invention adopts the following technical solutions:

[0011] In a first aspect, a medical MRI image denoising method is provided, comprising:

[0012] Receiving medical MRI image signals;

[0013] Converting the medical MRI image signal into an image domain to obtain three-dimensional image data, wherein the three-dimensional image data contains R layers of two-dimensional images, each layer of the two-dimensional image contains P×Q pixels, and R, P, and Q represent positive integers respectively;

[0014] A 3D image sub-block is selected from the 3D image data, wherein the 3D image sub-block includes R layers of 2D sub-images, and each layer of 2D sub-image includes pixels, represents a positive integer less than or equal to P, represents a positive integer less than or equal to Q;

[0015] Searching for M three-dimensional image similar sub-blocks from the three-dimensional image data that are located within a predefined neighborhood of the three-dimensional image sub-block and are most similar to the three-dimensional image sub-block, where M represents a positive integer;

[0016] The three-dimensional image sub-block and the M three-dimensional image similar sub-blocks are respectively subjected to vector straightening processing, and the M+1 one-dimensional vectors obtained by superposition processing are obtained to obtain a size of The original matrix Y;

[0017] Assume that the original matrix Y = X + N + S, and solve the following minimization objective function to obtain the denoised matrix X:

[0018]

[0019] Where N represents Gaussian noise, S represents sparse noise, ||S||1 represents the L1 norm used to constrain the sparse noise S, and ||N|| F represents the F norm used to constrain the Gaussian noise N, represents the Schattenp-norm used to constrain the denoising matrix X, λ1, λ2 and λ3 represent preset regularization coefficients, W represents a multi-channel weighted matrix in the form of a diagonal matrix, r represents a positive integer less than or equal to R, δ r represents the noise standard deviation of the rth channel, I represents the unit matrix, min() represents the minimum function, i represents a positive integer, w i Indicates that σ is assigned i The weight of σ i represents the i-th singular value in the denoising matrix X, p represents the parameter value used to determine the norm type and takes values ​​in the interval (0,1], and w represents the weight of the nuclear norm;

[0020] Reconstructing a new three-dimensional image sub-block corresponding to the three-dimensional image sub-block and M new three-dimensional image similar sub-blocks corresponding one-to-one to the M three-dimensional image similar sub-blocks according to the denoising matrix X;

[0021] In the three-dimensional image data, the three-dimensional image sub-block is replaced with a new three-dimensional image sub-block, and the M three-dimensional image similar sub-blocks are replaced one-to-one with the M new three-dimensional image similar sub-blocks to obtain new three-dimensional image data that has undergone image denoising.

[0022] Based on the above invention, a new MRI image denoising scheme based on multi-channel weighted robust principal component analysis is provided. Specifically, the medical MRI image signal is first converted to the image domain to obtain three-dimensional image data. Then, three-dimensional image sub-blocks are selected and the most similar M three-dimensional image similar sub-blocks are searched within a predefined neighborhood. These sub-blocks are then vectorized, straightened, and superimposed to obtain the original matrix Y. Then, the original matrix Y is assumed to be X+N+S and the minimization objective function is solved to obtain the denoised matrix X. Finally, new sub-blocks corresponding to the aforementioned sub-blocks are reconstructed based on the denoised matrix X. New three-dimensional image data that has undergone image denoising is obtained by replacing the old and new sub-blocks. This scheme can solve the problems of blurred denoised images and limited denoising capabilities in signal areas that exist in existing tMPPCA and GL-HOSVD technologies when applied to MRI denoising, thereby improving the quality of medical MRI images, facilitating subsequent image analysis, and facilitating practical application and promotion.

[0023] In one possible design, converting the medical MRI image signal into an image domain to obtain three-dimensional image data includes:

[0024] When the medical MRI image signal is obtained by scanning a subject using a three-dimensional density-adaptive radial sequence, the medical MRI image signal is converted into an image domain using a non-uniform fast Fourier transform to obtain the three-dimensional image data.

[0025] In one possible design, searching the three-dimensional image data for M three-dimensional image similar sub-blocks that are located within a predefined neighborhood of the three-dimensional image sub-block and are most similar to the three-dimensional image sub-block includes:

[0026] For each other 3D image sub-block in the 3D image data that is located within a predefined neighborhood of the 3D image sub-block and has the same size as the 3D image sub-block, calculating an Euler distance between the 3D image sub-block and a corresponding sub-block;

[0027] Arranging the other three-dimensional image sub-blocks in order from near to far according to the Euler distance to obtain a three-dimensional image sub-block sequence;

[0028] The first M three-dimensional image sub-blocks are selected from the three-dimensional image sub-block sequence as the M three-dimensional image similar sub-blocks that are most similar to the three-dimensional image sub-block, where M represents a positive integer.

[0029] In a possible design, the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks are respectively subjected to vector straightening processing, and the M+1 one-dimensional vectors obtained by the processing are superimposed to obtain a vector of size The original matrix Y includes:

[0030] For each sub-block in the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks, the pixel value of the pixel located in the p'th row and q'th column in the corresponding r'th layer two-dimensional sub-image is used as the h'th element in the corresponding one-dimensional vector to obtain the corresponding one-dimensional vector, where r' represents a positive integer less than or equal to R, and p' represents a positive integer less than or equal to A positive integer, q′ represents less than or equal to A positive integer,

[0031] The mth one-dimensional vector among the M+1 one-dimensional vectors corresponding to the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks is used as the mth one-dimensional vector in the size of The m-th row element in the original matrix Y of is obtained, where m represents a positive integer less than or equal to M+1;

[0032] Reconstructing a new three-dimensional image sub-block corresponding to the three-dimensional image sub-block and M new three-dimensional image similar sub-blocks corresponding one-to-one to the M three-dimensional image similar sub-blocks according to the denoising matrix X, including:

[0033] If the element in the mth row of the original matrix Y is a one-dimensional vector of the three-dimensional image sub-block, then the element located in the mth row and the h′th column of the denoising matrix X is used as the pixel value of the pixel located in the p′th row and the q′th column of the r′th layer two-dimensional sub-image in the new three-dimensional image sub-block corresponding to the three-dimensional image sub-block;

[0034] If the element in the mth row of the original matrix Y is a one-dimensional vector of the m′th three-dimensional image similar sub-block in the M three-dimensional image similar sub-blocks, then the element located in the mth row and h′th column of the denoising matrix X is used as the pixel value of the pixel point located in the p′th row and q′th column of the r′th layer two-dimensional sub-image in the new three-dimensional image similar sub-block corresponding to the m′th three-dimensional image similar sub-block.

[0035] In one possible design, the method further includes:

[0036] Different three-dimensional image sub-blocks in the three-dimensional image data are selected in a traversal manner / and a loop manner, and after each three-dimensional image sub-block is selected, the three-dimensional image data is iteratively updated by sequentially passing through the similar sub-block search step, the one-dimensional vector superposition step, the denoising matrix solution step, the new sub-block reconstruction step, and the image sub-block replacement step in the medical MRI image denoising method according to claim 1 until a preset convergence condition is met.

[0037] In a second aspect, a medical MRI image denoising device is provided, comprising an image signal receiving unit, an image data conversion unit, an image sub-block selection unit, a similar sub-block search unit, a straightening and superposition processing unit, a denoising matrix solving unit, a new sub-block reconstruction unit, and an image data updating unit, which are sequentially communicatively connected;

[0038] The image signal receiving unit is used to receive medical MRI image signals;

[0039] The image data conversion unit is configured to convert the medical MRI image signal into an image domain to obtain three-dimensional image data, wherein the three-dimensional image data contains R layers of two-dimensional images, each layer of the two-dimensional image contains P×Q pixels, and R, P, and Q represent positive integers respectively;

[0040] The image sub-block selection unit is used to select a 3D image sub-block from the 3D image data, wherein the 3D image sub-block includes R layers of 2D sub-images, and each layer of 2D sub-image includes pixels, represents a positive integer less than or equal to P, represents a positive integer less than or equal to Q;

[0041] The similar sub-block searching unit is configured to search the 3D image data for M 3D image similar sub-blocks that are located within a predefined neighborhood of the 3D image sub-block and are most similar to the 3D image sub-block, where M represents a positive integer;

[0042] The straightening and superposition processing unit is used to perform vector straightening processing on the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks respectively, and superpose the M+1 one-dimensional vectors obtained by the processing to obtain a size of The original matrix Y;

[0043] The denoising matrix solving unit is configured to assume that the original matrix Y=X+N+S and solve the following minimization objective function to obtain the denoising matrix X:

[0044]

[0045] Where N represents Gaussian noise, S represents sparse noise, ||S||1 represents the L1 norm used to constrain the sparse noise S, and ||N|| F represents the F norm used to constrain the Gaussian noise N, represents the Schattenp-norm used to constrain the denoising matrix X, λ1, λ2 and λ3 represent preset regularization coefficients, W represents a multi-channel weighted matrix in the form of a diagonal matrix, r represents a positive integer less than or equal to R, δ r represents the noise standard deviation of the rth channel, I represents the unit matrix, min() represents the minimum function, i represents a positive integer, w i Indicates that σ is assigned i The weight of σ i represents the i-th singular value in the denoising matrix X, p represents the parameter value used to determine the norm type and takes values ​​in the interval (0,1], and w represents the weight of the nuclear norm;

[0046] The new sub-block reconstruction unit is configured to reconstruct, according to the denoising matrix X, a new 3D image sub-block corresponding to the 3D image sub-block and M new 3D image similar sub-blocks corresponding one-to-one to the M 3D image similar sub-blocks;

[0047] The image data updating unit is configured to replace the three-dimensional image sub-block with a new three-dimensional image sub-block in the three-dimensional image data, and replace the M three-dimensional image similar sub-blocks with the M new three-dimensional image similar sub-blocks in a one-to-one correspondence, to obtain new three-dimensional image data that has undergone image denoising.

[0048] In a third aspect, the present invention provides a medical MRI image denoising system, comprising a magnetic resonance imager and a host computer in communication with each other, wherein the magnetic resonance imager comprises a scanning module and a magnetic resonance receiving coil;

[0049] The scanning module is used to scan the subject by transmitting a sequence suitable for MRI imaging through a radio frequency transmitting coil, so as to excite hydrogen atoms in the subject to generate resonance;

[0050] The magnetic resonance receiving coil is used to receive and transmit medical MRI image signals to the host computer during scanning of the subject;

[0051] The host computer is used to execute the medical MRI image denoising method as described in the first aspect or any possible design of the first aspect.

[0052] In a fourth aspect, the present invention provides a computer device comprising a storage module, a processing module, and a transceiver module that are communicatively connected in sequence, wherein the storage module is used to store a computer program, the transceiver module is used to send and receive messages, and the processing module is used to read the computer program and execute the medical MRI image denoising method as described in the first aspect or any possible design of the first aspect.

[0053] In a fifth aspect, the present invention provides a computer-readable storage medium having instructions stored thereon. When the instructions are run on a computer, the medical MRI image denoising method as described in the first aspect or any possible design of the first aspect is executed.

[0054] In a sixth aspect, the present invention provides a computer program product comprising a computer program or instructions, which, when executed by a computer, implement the medical MRI image denoising method as described in the first aspect or any possible design of the first aspect.

[0055] Beneficial effects of the above scheme:

[0056] (1) The present invention creatively provides a new MRI image denoising scheme based on multi-channel weighted robust principal component analysis, namely, first converting the medical MRI image signal into the image domain to obtain three-dimensional image data, then selecting three-dimensional image sub-blocks and searching for the most similar M three-dimensional image similar sub-blocks in a predefined neighborhood, then vectorizing and straightening these sub-blocks and superimposing them to obtain the original matrix Y, then assuming that the original matrix Y = X + N + S and solving the minimization objective function to obtain the denoising matrix X, finally reconstructing new sub-blocks corresponding to the aforementioned sub-blocks according to the denoising matrix X, and obtaining new three-dimensional image data that has undergone image denoising by replacing the old and new sub-blocks, thereby solving the problems of blurred denoised images and limited denoising capabilities for signal areas when the existing tMPPCA and GL-HOSVD technologies are applied to MRI denoising, thereby improving the quality of medical MRI images and facilitating subsequent image analysis;

[0057] (2) This scheme extends the Schatten p-norm into a multi-channel weighted form: on the one hand, by assigning different importance to the singular value components, a more accurate low-rank approximation can be achieved; on the other hand, the channel-specific weighting matrix adaptively balances the noise suppression of each coil and effectively balances the inter-channel heterogeneity and the global low-rank structure by enforcing the low-rank constraint on all channels;

[0058] (3) This scheme also extends robust PCA by combining GRPCA with regularized noise modeling, which can separate low-rank components from noise, thereby improving the fidelity of anatomical structure recovery under complex noise conditions and facilitating practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0060] Figure 1 A flowchart of the medical MRI image denoising method provided in an embodiment of the present application.

[0061] Figure 2 This is an example diagram comparing the denoising effect of the medical MRI image denoising method provided in an embodiment of the present application with the tMMPCA method and the GL-HOSVD method.

[0062] Figure 3 A schematic diagram of the structure of a medical MRI image denoising device provided in an embodiment of the present application.

[0063] Figure 4 A schematic diagram of the structure of a medical MRI image denoising system provided in an embodiment of the present application.

[0064] Figure 5 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0065] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention will be briefly introduced below in conjunction with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structures of the drawings is only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these embodiments without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0066] It should be understood that although the terms first, second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are merely used to distinguish one object from another. For example, a first object can be referred to as a second object, and similarly, a second object can be referred to as a first object without departing from the scope of the exemplary embodiments of the present invention.

[0067] It should be understood that the term "and / or" that may appear in this document is merely a description of the association relationship between associated objects, indicating that there may be three relationships. For example, A and / or B can indicate three situations: A exists alone, B exists alone, or A and B exist at the same time. For another example, A, B and / or C can indicate the existence of any one of A, B and C or any combination of them. The term " / and" that may appear in this document describes another type of association object relationship, indicating that there may be two relationships. For example, A / and B can indicate two situations: A exists alone or A and B exist at the same time. In addition, the character " / " that may appear in this document generally indicates that the previous and next associated objects are in an "or" relationship.

[0068] Example

[0069] like Figure 1 As shown, the medical MRI image denoising method provided in the first aspect of this embodiment can be, but is not limited to, executed by a computer device having certain computing resources and being communicatively connected to a magnetic resonance imaging device, wherein the magnetic resonance imaging device includes, but is not limited to, a scanning module and a magnetic resonance receiving coil; the scanning module is used to transmit a sequence suitable for MRI imaging via a radio frequency transmitting coil to scan a subject (e.g., a patient) to stimulate hydrogen atoms in the subject to produce resonance; the magnetic resonance receiving coil is used to receive and transmit medical MRI image signals to the computer device during the scanning of the subject. The medical MRI image denoising method includes, but is not limited to, the following steps S1 to S8.

[0070] S1. Receive medical MRI image signals.

[0071] In step S1, the medical MRI image signal comes from the magnetic resonance receiving coil and can be received conventionally through existing wired communication technology. In addition, the medical MRI image signal is specifically exemplified by but not limited to medical 23 Na signal in MRI images.

[0072] S2. Convert the medical MRI image signal into an image domain to obtain three-dimensional image data, wherein the three-dimensional image data contains R layers of two-dimensional images, each layer of the two-dimensional image contains P×Q pixels, and R, P, and Q represent positive integers respectively.

[0073] In step S2, the size of the three-dimensional image data can be, for example, 50×80×80, that is, it contains 50 layers of two-dimensional images, and each layer of two-dimensional image contains 80×80 pixels. Specifically, the medical MRI image signal is converted into the image domain to obtain the three-dimensional image data, including but not limited to: when the medical MRI image signal is obtained by scanning the subject using a three-dimensional density-adapted radial sequence, the medical MRI image signal is converted into the image domain using a non-uniform fast Fourier transform to obtain the three-dimensional image data. The three-dimensional density-adapted radial sequence (DA-3DPR) is an existing scanning sequence, and the non-uniform fast Fourier transform (NUFFT) is also an existing technical means, which will not be described in detail here.

[0074] S3. Select a 3D image sub-block from the 3D image data, wherein the 3D image sub-block includes R layers of 2D sub-images, and each layer of 2D sub-images includes pixels, represents a positive integer less than or equal to P, Represents a positive integer less than or equal to Q.

[0075] In step S3, the 3D image sub-block is used as a local reference patch. Based on the example of step S2 above, the size of the 3D image sub-block can be, for example, 50×10×10, that is, it contains 50 layers of 2D sub-images, and each layer of 2D sub-image contains 10×10 pixels. The specific method for selecting the 3D image sub-block can be manual selection, random selection, or traversal selection using a sliding window method (for example, first select a 10×10 pixel 2D sub-image in the upper left corner of the 2D image, then the initial size of the 3D image sub-block is 50×10×10, then slide the window one or more pixels to the right on the 2D image, and continue to select the next 50×10×10 3D image sub-block).

[0076] S4. Searching for M three-dimensional image similar sub-blocks that are located in a predefined neighborhood of the three-dimensional image sub-block and are most similar to the three-dimensional image sub-block from the three-dimensional image data, where M represents a positive integer.

[0077] In step S4, based on the example of step S3 above, the size of the 3D image similar sub-block is also 50×10×10. The size of the predefined neighborhood can be specified by the user or conventionally determined based on the results of multiple limited experiments. Specifically, searching the 3D image data for M 3D image similar sub-blocks that are located within the predefined neighborhood of the 3D image sub-block and are most similar to the 3D image sub-block includes, but is not limited to, the following steps S41 to S43.

[0078] S41. For each other 3D image sub-block in the 3D image data that is located within a predefined neighborhood of the 3D image sub-block and has the same size as the 3D image sub-block, calculate the Euler distance between the 3D image sub-block and the corresponding sub-block.

[0079] In step S41 , since the 3D image sub-block and the other 3D image sub-blocks can both be regarded as 3D vectors, the Euler distance between the two can be calculated based on the existing Euler distance formula.

[0080] S42. Arrange the other three-dimensional image sub-blocks in order from near to far according to the Euler distance to obtain a three-dimensional image sub-block sequence.

[0081] S43. Select the first M three-dimensional image sub-blocks from the three-dimensional image sub-block sequence as the M three-dimensional image similar sub-blocks that are most similar to the three-dimensional image sub-block, where M represents a positive integer.

[0082] S5. Perform vector straightening processing on the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks respectively, and superimpose the M+1 one-dimensional vectors obtained by the processing to obtain a size of The original matrix Y.

[0083] In step S5, the vectorization straightening process is to process the current three-dimensional vector of the corresponding sub-block into a one-dimensional vector. Specifically, the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks are respectively subjected to vectorization straightening process, and the M+1 one-dimensional vectors obtained by the process are superimposed to obtain a vector of size The original matrix Y includes but is not limited to the following steps S51 to S52.

[0084] S51. For each sub-block in the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks, the pixel value of the pixel located in the p'th row and q'th column in the corresponding r'th layer two-dimensional sub-image is used as the h'th element in the corresponding one-dimensional vector to obtain the corresponding one-dimensional vector, where r' represents a positive integer less than or equal to R, and p' represents a positive integer less than or equal to A positive integer, q′ represents less than or equal to A positive integer,

[0085] S52. The mth one-dimensional vector among the M+1 one-dimensional vectors corresponding to the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks is used as the mth one-dimensional vector in the size of The m-th row element in the original matrix Y of is obtained to obtain the original matrix Y, where m represents a positive integer less than or equal to M+1.

[0086] S6. Assume that the original matrix Y = X + N + S, and solve the following minimization objective function to obtain the denoised matrix X:

[0087]

[0088] Where N represents Gaussian noise, S represents sparse noise, ||S||1 represents the L1 norm used to constrain the sparse noise S, and ||N|| F represents the F norm used to constrain the Gaussian noise N, represents the Schattenp-norm used to constrain the denoising matrix X, λ1, λ2 and λ3 represent preset regularization coefficients, W represents a multi-channel weighted matrix in the form of a diagonal matrix, r represents a positive integer less than or equal to R, δ r represents the noise standard deviation of the rth channel, I represents the unit matrix, min() represents the minimum function, i represents a positive integer, w i Indicates that σ is assigned i The weight of σ i represents the i-th singular value in the denoising matrix X, p represents the parameter value for determining the norm type and takes values ​​in the interval (0,1], and w represents the weight of the nuclear norm.

[0089] In step S6, the denoising matrix X represents a potential clean image (i.e., a clean, noise-free image that meets the requirements). The L1 norm, the F norm, and the Schatten p-norm are all existing norms. In addition, the specific solution process for minimizing the objective function can be implemented using, but is not limited to, the existing alternating direction method of multipliers (ADMM) algorithm.

[0090] S7. Reconstruct a new 3D image sub-block corresponding to the 3D image sub-block and M new 3D image similar sub-blocks corresponding one-to-one to the M 3D image similar sub-blocks according to the denoising matrix X.

[0091] In step S7, the reconstruction process of the new 3D image sub-block and the M new 3D image similar sub-blocks is the inverse process of the aforementioned step S5; specifically, based on the denoising matrix X, a new 3D image sub-block corresponding to the 3D image sub-block and M new 3D image similar sub-blocks corresponding one-to-one to the M 3D image similar sub-blocks are reconstructed, including but not limited to: if the element in the mth row of the original matrix Y is a one-dimensional vector of the 3D image sub-block, then the element in the mth row and h′th column of the denoising matrix X is used as The pixel value of the pixel point located in the p′th row and q′th column of the r′th layer two-dimensional sub-image in the new three-dimensional image sub-block corresponding to the three-dimensional image sub-block; if the element in the mth row of the original matrix Y is the one-dimensional vector of the m′th three-dimensional image similar sub-block in the M three-dimensional image similar sub-blocks, then the element located in the mth row and h′th column of the denoising matrix X is used as the pixel value of the pixel point located in the p′th row and q′th column of the r′th layer two-dimensional sub-image in the new three-dimensional image similar sub-block corresponding to the m′th three-dimensional image similar sub-block.

[0092] S8. In the three-dimensional image data, replace the three-dimensional image sub-block with a new three-dimensional image sub-block, and replace the M three-dimensional image similar sub-blocks with the M new three-dimensional image similar sub-blocks in a one-to-one correspondence to obtain new three-dimensional image data that has undergone image denoising.

[0093] Based on the above steps S1 to S8, this embodiment also conducted the following tests: 23 Na MRI image, and add the actual noise collected to the simulation data to obtain Figure 2 The denoising effect comparison chart shown below and the technical indicator comparison table shown in Table 1 below:

[0094] Table 1. Comparison of denoising indicators of this embodiment with those of the tMMPCA method and the GL-HOSVD method

[0095]

[0096] It can be seen from this that the denoising effect of the method described in this embodiment is better than that of the tMMPCA method and the GL-HOSVD method. From the PSNR (Peak Signal-to-Noise Ratio), SSIM (Structural Similarity Index) and MSE (Mean Squared Error) indicators and the difference graph, it can be seen that the denoising quality of the method described in this embodiment is better and closer to the real data, which can be proved to be able to effectively remove medical 23 Noise in NaMRI images.

[0097] Therefore, based on the medical MRI image denoising method described in the aforementioned steps S1 to S8, a new MRI image denoising scheme based on multi-channel weighted robust principal component analysis is provided, namely, the medical MRI image signal is first converted into the image domain to obtain three-dimensional image data, and then a three-dimensional image sub-block is selected and the most similar M three-dimensional image similar sub-blocks are searched in a predefined neighborhood, and then these sub-blocks are vectorized, straightened and superimposed to obtain the original matrix Y, and then the original matrix Y = X + N + S is assumed and the minimization objective function is solved to obtain the denoising matrix X, and finally a new sub-block corresponding to the aforementioned sub-block is reconstructed according to the denoising matrix X, and new three-dimensional image data that has been denoised is obtained by replacing the old and new sub-blocks. In this way, the problems of blurred images after denoising and limited denoising capabilities for signal areas when the existing tMPPCA and GL-HOSVD technologies are applied to MRI denoising can be solved, thereby improving the quality of medical MRI images and 23 The method was verified on NaMRI images, which is conducive to subsequent image analysis and facilitates practical application and promotion.

[0098] Based on the technical solution of the first aspect, this embodiment further provides a possible design for iteratively denoising medical MRI images. That is, the method further includes, but is not limited to: selecting different three-dimensional image sub-blocks in the three-dimensional image data in a traversal manner / a cyclic manner, and after each selection of a three-dimensional image sub-block, sequentially performing the similar sub-block search step (i.e., step S4), the one-dimensional vector superposition step (i.e., step S5), the denoising matrix solution step (i.e., step S6), the new sub-block reconstruction step (i.e., step S7), and the image sub-block replacement step (i.e., step S8) in the medical MRI image denoising method as described in the first aspect, to iteratively update the three-dimensional image data until a preset convergence condition is met. The above-mentioned convergence conditions have the following two cases: (1) the maximum number of denoising cycles is set to K in advance (it needs to be set according to experience based on different data and noise intensity), and reference patches (i.e., the three-dimensional image sub-blocks) are selected in the entire image. Then, for each reference patch, similar blocks are selected for denoising in the predefined neighborhood. All similar blocks are denoised once to complete a denoising cycle. If the current number of denoising cycles exceeds K, the algorithm stops. (2) When || X k -Z k || F ≤Tol,||X k+1 -X k || F ≤Tol and ||Z k+1 -Z k || F ≤Tol>0, the convergence is determined, where Tol represents a preset and small tolerance number, and Tol>0, k represents a positive integer, Xk Denotes the denoised matrices X, Z for the kth alternation when solving using the alternating direction multiplier algorithm. k represents the auxiliary variable matrix for the kth alternation when solving using the alternating direction multiplier algorithm, || || F represents the F-norm.

[0099] like Figure 3 As shown, the second aspect of this embodiment provides a virtual device for implementing the medical MRI image denoising method described in the first aspect or possible design 1, comprising an image signal receiving unit, an image data conversion unit, an image sub-block selection unit, a similar sub-block search unit, a straightening and superposition processing unit, a denoising matrix solving unit, a new sub-block reconstruction unit, and an image data updating unit, which are sequentially communicatively connected;

[0100] The image signal receiving unit is used to receive medical MRI image signals;

[0101] The image data conversion unit is configured to convert the medical MRI image signal into an image domain to obtain three-dimensional image data, wherein the three-dimensional image data contains R layers of two-dimensional images, each layer of the two-dimensional image contains P×Q pixels, and R, P, and Q represent positive integers respectively;

[0102] The image sub-block selection unit is used to select a 3D image sub-block from the 3D image data, wherein the 3D image sub-block includes R layers of 2D sub-images, and each layer of 2D sub-image includes pixels, represents a positive integer less than or equal to P, represents a positive integer less than or equal to Q;

[0103] The similar sub-block searching unit is configured to search the 3D image data for M 3D image similar sub-blocks that are located within a predefined neighborhood of the 3D image sub-block and are most similar to the 3D image sub-block, where M represents a positive integer;

[0104] The straightening and superposition processing unit is used to perform vector straightening processing on the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks respectively, and superpose the M+1 one-dimensional vectors obtained by the processing to obtain a size of The original matrix Y;

[0105] The denoising matrix solving unit is configured to assume that the original matrix Y=X+N+S and solve the following minimization objective function to obtain the denoising matrix X:

[0106]

[0107] Where N represents Gaussian noise, S represents sparse noise, ||S||1 represents the L1 norm used to constrain the sparse noise S, and ||N|| F represents the F norm used to constrain the Gaussian noise N, represents the Schattenp-norm used to constrain the denoising matrix X, λ1, λ2 and λ3 represent preset regularization coefficients, W represents a multi-channel weighted matrix in the form of a diagonal matrix, r represents a positive integer less than or equal to R, δ r represents the noise standard deviation of the rth channel, I represents the unit matrix, min() represents the minimum function, i represents a positive integer, w i Indicates that σ is assigned i The weight of σ i represents the i-th singular value in the denoising matrix X, p represents the parameter value used to determine the norm type and takes values ​​in the interval (0,1], and w represents the weight of the nuclear norm;

[0108] The new sub-block reconstruction unit is configured to reconstruct, according to the denoising matrix X, a new 3D image sub-block corresponding to the 3D image sub-block and M new 3D image similar sub-blocks corresponding one-to-one to the M 3D image similar sub-blocks;

[0109] The image data updating unit is configured to replace the three-dimensional image sub-block with a new three-dimensional image sub-block in the three-dimensional image data, and replace the M three-dimensional image similar sub-blocks with the M new three-dimensional image similar sub-blocks in a one-to-one correspondence, to obtain new three-dimensional image data that has undergone image denoising.

[0110] The working process, working details and technical effects of the aforementioned device provided in the second aspect of this embodiment can be referred to the medical MRI image denoising method described in the first aspect or possible design one, and will not be repeated here.

[0111] like Figure 4 As shown, a third aspect of this embodiment provides a physical system for implementing the medical MRI image denoising method described in the first aspect or possible design one, comprising a magnetic resonance imaging device and a host computer that are communicatively connected, wherein the magnetic resonance imaging device includes a scanning module and a magnetic resonance receiving coil;

[0112] The scanning module is used to scan the subject by transmitting a sequence suitable for MRI imaging through a radio frequency transmitting coil, so as to excite hydrogen atoms in the subject to generate resonance;

[0113] The magnetic resonance receiving coil is used to receive and transmit medical MRI image signals to the host computer during scanning of the subject;

[0114] The host computer is used to execute the medical MRI image denoising method as described in the first aspect or possible design one.

[0115] The working process, working details and technical effects of the aforementioned system provided in the third aspect of this embodiment can be referred to the medical MRI image denoising method described in the first aspect or possible design one, and will not be repeated here.

[0116] like Figure 5 As shown, the fourth aspect of this embodiment provides a computer device for executing the medical MRI image denoising method as described in the first aspect or possible design one, comprising a storage module, a processing module and a transceiver module connected in sequence, wherein the storage module is used to store a computer program, the transceiver module is used to send and receive messages, and the processing module is used to read the computer program and execute the medical MRI image denoising method as described in the first aspect or possible design one. For example, the storage module may include, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a flash memory, a first-input first-output memory (FIFO) and / or a first-input last-output memory (FILO), etc.; the processing module may include, but is not limited to, a microprocessor of the STM32F105 series. In addition, the computer device may also include, but is not limited to, a power module, a display screen and other necessary components.

[0117] The working process, working details and technical effects of the aforementioned computer device provided in the fourth aspect of this embodiment can be referred to the medical MRI image denoising method described in the first aspect or possible design one, and will not be repeated here.

[0118] A fifth aspect of this embodiment provides a computer-readable storage medium storing instructions for the medical MRI image denoising method as described in the first aspect or possible design 1. Specifically, the computer-readable storage medium stores instructions that, when executed on a computer, execute the medical MRI image denoising method as described in the first aspect or possible design 1. The computer-readable storage medium refers to a data storage medium and may include, but is not limited to, a floppy disk, a CD, a hard disk, a flash memory, a USB flash drive, and / or a memory stick. The computer may be a general-purpose computer, a dedicated computer, a computer network, or other programmable device.

[0119] The working process, working details and technical effects of the aforementioned computer-readable storage medium provided in the fifth aspect of this embodiment can be referred to the medical MRI image denoising method described in the first aspect or possible design one, and will not be repeated here.

[0120] A sixth aspect of the present embodiment provides a computer program product, including a computer program or instructions, which, when executed by a computer, implements the medical MRI image denoising method according to the first aspect or possible design 1. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.

[0121] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A medical MRI image denoising method, characterized in that: include: Receiving medical MRI image signals; Converting the medical MRI image signal into an image domain to obtain three-dimensional image data, wherein the three-dimensional image data contains R layers of two-dimensional images, each layer of the two-dimensional image contains P×Q pixels, and R, P, and Q represent positive integers respectively; A 3D image sub-block is selected from the 3D image data, wherein the 3D image sub-block includes R layers of 2D sub-images, and each layer of 2D sub-image includes pixels, represents a positive integer less than or equal to P, represents a positive integer less than or equal to Q; Searching for M three-dimensional image similar sub-blocks from the three-dimensional image data that are located within a predefined neighborhood of the three-dimensional image sub-block and are most similar to the three-dimensional image sub-block, where M represents a positive integer; The three-dimensional image sub-block and the M three-dimensional image similar sub-blocks are respectively subjected to vector straightening processing, and the M+1 one-dimensional vectors obtained by superposition processing are obtained to obtain a size of The original matrix Y; Assume that the original matrix Y = X + N + S, and solve the following minimization objective function to obtain the denoised matrix X: Where N represents Gaussian noise, S represents sparse noise, ||S||1 represents the L1 norm used to constrain the sparse noise S, and ||N|| F represents the F norm used to constrain the Gaussian noise N, represents the Schatten p-norm used to constrain the denoising matrix X, λ1, λ2 and λ3 respectively represent the preset regularization coefficients, W represents a multi-channel weighted matrix in the form of a diagonal matrix, r represents a positive integer less than or equal to R, δ r represents the noise standard deviation of the rth channel, I represents the unit matrix, min() represents the minimum function, i represents a positive integer, w i Indicates that σ is assigned i The weight of σ i represents the i-th singular value in the denoising matrix X, p represents the parameter value used to determine the norm type and takes values ​​in the interval (0,1], and w represents the weight of the nuclear norm; Reconstructing a new three-dimensional image sub-block corresponding to the three-dimensional image sub-block and M new three-dimensional image similar sub-blocks corresponding one-to-one to the M three-dimensional image similar sub-blocks according to the denoising matrix X; In the three-dimensional image data, the three-dimensional image sub-block is replaced with a new three-dimensional image sub-block, and the M three-dimensional image similar sub-blocks are replaced one-to-one with the M new three-dimensional image similar sub-blocks to obtain new three-dimensional image data that has undergone image denoising.

2. The medical MRI image denoising method according to claim 1, wherein: Converting the medical MRI image signal into an image domain to obtain three-dimensional image data includes: When the medical MRI image signal is obtained by scanning a subject using a three-dimensional density-adaptive radial sequence, the medical MRI image signal is converted into an image domain using a non-uniform fast Fourier transform to obtain the three-dimensional image data.

3. The medical MRI image denoising method according to claim 1, wherein: Searching for M three-dimensional image similar sub-blocks that are located in a predefined neighborhood of the three-dimensional image sub-block and are most similar to the three-dimensional image sub-block from the three-dimensional image data, comprising: For each other 3D image sub-block in the 3D image data that is located within a predefined neighborhood of the 3D image sub-block and has the same size as the 3D image sub-block, calculating an Euler distance between the 3D image sub-block and a corresponding sub-block; Arranging the other three-dimensional image sub-blocks in order from near to far according to the Euler distance to obtain a three-dimensional image sub-block sequence; The first M three-dimensional image sub-blocks are selected from the three-dimensional image sub-block sequence as the M three-dimensional image similar sub-blocks that are most similar to the three-dimensional image sub-block, where M represents a positive integer.

4. The medical MRI image denoising method according to claim 1, wherein: The three-dimensional image sub-block and the M three-dimensional image similar sub-blocks are respectively subjected to vector straightening processing, and the M+1 one-dimensional vectors obtained by superposition processing are obtained to obtain a size of The original matrix Y includes: For each sub-block in the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks, the pixel value of the pixel located in the p'th row and q'th column in the corresponding r'th layer two-dimensional sub-image is used as the h'th element in the corresponding one-dimensional vector to obtain the corresponding one-dimensional vector, where r' represents a positive integer less than or equal to R, and p' represents a positive integer less than or equal to A positive integer, q′ represents less than or equal to A positive integer, The mth one-dimensional vector among the M+1 one-dimensional vectors corresponding to the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks is used as the mth one-dimensional vector in the size of The m-th row element in the original matrix Y of is obtained, where m represents a positive integer less than or equal to M+1; Reconstructing a new three-dimensional image sub-block corresponding to the three-dimensional image sub-block and M new three-dimensional image similar sub-blocks corresponding one-to-one to the M three-dimensional image similar sub-blocks according to the denoising matrix X, including: If the element in the mth row of the original matrix Y is a one-dimensional vector of the three-dimensional image sub-block, then the element located in the mth row and the h′th column of the denoising matrix X is used as the pixel value of the pixel located in the p′th row and the q′th column of the r′th layer two-dimensional sub-image in the new three-dimensional image sub-block corresponding to the three-dimensional image sub-block; If the element in the mth row of the original matrix Y is a one-dimensional vector of the m′th three-dimensional image similar sub-block in the M three-dimensional image similar sub-blocks, then the element located in the mth row and h′th column of the denoising matrix X is used as the pixel value of the pixel point located in the p′th row and q′th column of the r′th layer two-dimensional sub-image in the new three-dimensional image similar sub-block corresponding to the m′th three-dimensional image similar sub-block.

5. The medical MRI image denoising method according to claim 1, wherein: The method further comprises: Different three-dimensional image sub-blocks in the three-dimensional image data are selected in a traversal manner / and a loop manner, and after each three-dimensional image sub-block is selected, the three-dimensional image data is iteratively updated by sequentially passing through the similar sub-block search step, the one-dimensional vector superposition step, the denoising matrix solution step, the new sub-block reconstruction step, and the image sub-block replacement step in the medical MRI image denoising method according to claim 1 until a preset convergence condition is met.

6. A medical MRI image denoising device, characterized in that: It includes an image signal receiving unit, an image data conversion unit, an image sub-block selection unit, a similar sub-block search unit, a straightening and superposition processing unit, a denoising matrix solving unit, a new sub-block reconstruction unit and an image data updating unit which are sequentially connected in communication; The image signal receiving unit is used to receive medical MRI image signals; The image data conversion unit is configured to convert the medical MRI image signal into an image domain to obtain three-dimensional image data, wherein the three-dimensional image data contains R layers of two-dimensional images, each layer of the two-dimensional image contains P×Q pixels, and R, P, and Q represent positive integers respectively; The image sub-block selection unit is used to select a 3D image sub-block from the 3D image data, wherein the 3D image sub-block includes R layers of 2D sub-images, and each layer of 2D sub-image includes pixels, represents a positive integer less than or equal to P, represents a positive integer less than or equal to Q; The similar sub-block searching unit is configured to search the 3D image data for M 3D image similar sub-blocks that are located within a predefined neighborhood of the 3D image sub-block and are most similar to the 3D image sub-block, where M represents a positive integer; The straightening and superposition processing unit is used to perform vector straightening processing on the three-dimensional image sub-block and the M three-dimensional image similar sub-blocks respectively, and superpose the M+1 one-dimensional vectors obtained by the processing to obtain a size of The original matrix Y; The denoising matrix solving unit is configured to assume that the original matrix Y=X+N+S and solve the following minimization objective function to obtain the denoising matrix X: Where N represents Gaussian noise, S represents sparse noise, ||S||1 represents the L1 norm used to constrain the sparse noise S, and ||N|| F represents the F norm used to constrain the Gaussian noise N, represents the Schatten p-norm used to constrain the denoising matrix X, λ1, λ2 and λ3 respectively represent the preset regularization coefficients, W represents a multi-channel weighted matrix in the form of a diagonal matrix, r represents a positive integer less than or equal to R, δ r represents the noise standard deviation of the rth channel, I represents the unit matrix, min() represents the minimum function, i represents a positive integer, w i Indicates that σ is assigned i The weight of σ i represents the i-th singular value in the denoising matrix X, p represents the parameter value used to determine the norm type and takes values ​​in the interval (0,1], and w represents the weight of the nuclear norm; The new sub-block reconstruction unit is configured to reconstruct, according to the denoising matrix X, a new 3D image sub-block corresponding to the 3D image sub-block and M new 3D image similar sub-blocks corresponding one-to-one to the M 3D image similar sub-blocks; The image data updating unit is configured to replace the three-dimensional image sub-block with a new three-dimensional image sub-block in the three-dimensional image data, and replace the M three-dimensional image similar sub-blocks with the M new three-dimensional image similar sub-blocks in a one-to-one correspondence, to obtain new three-dimensional image data that has undergone image denoising.

7. A medical MRI image denoising system, characterized in that: It includes a magnetic resonance imaging device and a host computer that are communicatively connected, wherein the magnetic resonance imaging device includes a scanning module and a magnetic resonance receiving coil; The scanning module is used to scan the subject by transmitting a sequence suitable for MRI imaging through a radio frequency transmitting coil, so as to excite hydrogen atoms in the subject to generate resonance; The magnetic resonance receiving coil is used to receive and transmit medical MRI image signals to the host computer during scanning of the subject; The host computer is used to execute the medical MRI image denoising method according to any one of claims 1 to 5.

8. A computer device, characterized in that: The method comprises a storage module, a processing module and a transceiver module which are communicatively connected in sequence, wherein the storage module is used to store a computer program, the transceiver module is used to send and receive messages, and the processing module is used to read the computer program and execute the medical MRI image denoising method according to any one of claims 1 to 5.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed on a computer, the medical MRI image denoising method according to any one of claims 1 to 5 is executed.

10. A computer program product comprising a computer program or instructions, characterized in that When the computer program or the instruction is executed by a computer, the medical MRI image denoising method according to any one of claims 1 to 5 is implemented.

Images