Quantitative magnetic resonance imaging method based on image structure and physical relaxation priors

Abstract

The present application discloses a magnetic resonance quantitative imaging method, apparatus, device and storage medium thereof based on image structure and physical relaxation priors. The method comprises: extracting image blocks with structural similarity based on a block matching method, and constructing a low-rank structure tensor based on the image blocks; constructing a low-rank parameter tensor for a weighted image using a signal physical relaxation prior; and establishing an image reconstruction model and solution method based on a low-rank tensor using the low-rank structure tensor and the low-rank parameter tensor. The above-mentioned scheme provided by the present application greatly accelerates the data scanning speed and reduces the quantitative imaging time. During image reconstruction, the reconstruction method proposed by the present invention can still accurately reconstruct a parameter-weighted image from highly undersampled data at a high acceleration factor.

Description

Technical Field

[0001] The present invention relates to magnetic resonance parameter quantitative imaging, and in particular to a magnetic resonance quantitative imaging method, device, equipment and storage medium thereof based on image structure and physical relaxation prior. Background Art

[0002] Magnetic resonance parameter quantitative imaging is a technique for quantifying certain specific physiological parameters of tissues. Compared with conventional magnetic resonance imaging, this technique can provide a large amount of tissue-specific information in addition to morphological assessment. It is an important imaging method for disease diagnosis and assessment that has emerged in recent years. Depending on the quantitative technique, when quantitatively measuring tissues, multiple data sets with different quantitative control parameters are usually acquired first, from which parameter-weighted images are reconstructed, and finally a parameter map is fitted from the image using a suitable parameter model. However, the acquisition of multiple data sets will increase the imaging time exponentially, resulting in increased motion artifacts, high radiofrequency power deposition, and patient discomfort. Sensitivity to motion during scanning and displacement during scanning will make quantification difficult, greatly reducing the clinical application value of this method. Therefore, accelerating imaging speed, improving imaging efficiency, and realizing clinically applicable magnetic resonance parameter quantitative imaging methods have important theoretical research and clinical application value.

[0003] In order to shorten the scanning time, the current technology mainly focuses on the following three directions: First, reduce the number of TSL. This method reduces the number of TSLs collected. 1ρ The number of weighted images is also reduced, thus reducing quantitative accuracy. Secondly, rapid imaging sequences are used, but due to hardware limitations, scanning speed does not increase significantly. Finally, rapid imaging techniques are currently commercially available, primarily parallel imaging techniques (such as sensitivity encoding (SENSE) and generalized autocalibrated partially parallel acquisition (GRAPPA)). However, due to the limitations of parallel imaging array coils, the higher the acceleration, the lower the signal-to-noise ratio of the image obtained. Therefore, scanning speeds using this method are typically only 2-3 times faster. In recent years, compressed sensing technology based on sparse sampling theory has received widespread attention and application in rapid magnetic resonance imaging. According to the theory of compressed sensing, as long as the signal is sparse or compressible, an incoherent measurement can be performed and optimization methods can be used to solve a minimization problem to accurately reconstruct the original signal from highly undersampled data.

[0004] Compressed sensing theory has been widely applied in magnetic resonance parameter quantitative imaging, improving scanning efficiency while ensuring accurate parameter weighted images and parameter values. For the space-parameter matrix composed of weighted image sequences, existing literature has proposed reconstruction methods based on low rank and sparse constraints; or using sparse regularization to effectively improve the imaging speed of bone and joint quantitative imaging.[5] ; or impose local low-rank constraints on the space-parameter matrix to further improve the reconstruction performance; but the vectorization or matrix processing of the data in the reconstruction process of the above methods destroys the original spatial structure of the data. On this basis, some technologies use high-order tensors to better explore the correlation of high-dimensional image data and maximize the retention of the original structure of the data. For example, the low-rank tensor reconstruction method based on image domain structural similarity and K-space domain structural similarity proposed in the existing literature, but it only studies the low-rank driven by structural similarity, and ignores the low-rank related to the physical relaxation of the signal. [9,10] .

[0005] Existing fast imaging methods based on compressed sensing theory mostly study the correlation of image data within a two-dimensional matrix framework. During the image reconstruction process, high-dimensional data needs to be vectorized or matrixed. For high-dimensional image data of three dimensions or above, this operation will destroy the inherent structure of the original image signal and conceal the redundant information originally existing in the high-dimensional data. Summary of the Invention

[0006] In view of the above-mentioned defects or deficiencies in the prior art, it is desirable to provide a magnetic resonance quantitative imaging method, apparatus, device and storage medium thereof based on image structure and physical relaxation priors.

[0007] In a first aspect, an embodiment of the present application provides a method for magnetic resonance quantitative imaging based on image structure and physical relaxation priors, the method comprising:

[0008] Image blocks with structural similarity are extracted based on the block matching method, and a low-rank structure tensor is constructed based on the image blocks;

[0009] Using the signal physics relaxation prior, a low-rank parameter tensor is constructed for the weighted image;

[0010] An image reconstruction model and solution method based on low-rank tensor are established through low-rank structure tensor and low-rank parameter tensor.

[0011] In one embodiment, the extracting image blocks with structural similarity based on a block matching method includes:

[0012] definition T 1ρ Weighted image, where N x 、N y 、N TSL are the number of pixels along the frequency encoding and phase encoding directions, and the number of TSLs respectively;

[0013] Search all image patches within the r×r neighborhood in

[0014] The similarity between image patches is measured by the normalized L2 distance, which is defined as:

[0015] when The distance is less than the threshold λ m When , the image block is defined and B i are image patches with structural similarity.

[0016] In one embodiment, the The distance is less than the threshold λ m Afterwards, the method further comprises:

[0017] Limit the number of similar image blocks to be searched to N patch , all similar image blocks are grouped into a block of size b 2 ×N patch where b is the size of the image block (b×b).

[0018] In one embodiment, constructing a low-rank structure tensor based on the image block includes:

[0019] Extract all T 1ρ Similar image blocks at a given pixel i in the weighted image;

[0020] Construct a third-order tensor with low-rank properties based on all similar image patches Set P i is the similar image block extraction operator, then

[0021] In one embodiment, constructing a low-rank parameter tensor for a weighted image using a signal physical relaxation prior includes:

[0022] According to T 1ρ The principle of quantitative imaging, based on the formula Perform point-by-point fitting on each pixel in the weighted image; where TSL represents the spin lock time, M represents the signal after TSL time, and M0 is a constant representing the reference signal when no preparation pulse is applied;

[0023] According to the estimated organization T 1ρ The quantitative values ​​are used to cluster the pixels in the image so that the pixels in each cluster group belong to the same type of tissue;

[0024] For the clustered pixels, different T 1ρ The pixels at the same position in the weighted image form a low-rank Hankel matrix;

[0025] The Hankel matrices are stacked to form a third-order parameter tensor with low-rank properties.

[0026] In one embodiment, the method of establishing a low-rank tensor-based image reconstruction model and solution method by using a low-rank structure tensor and a low-rank parameter tensor includes:

[0027] An image reconstruction model based on a low-rank tensor is established according to the low-rank structure tensor and the low-rank parameter tensor.

[0028]

[0029]

[0030] Where, E is the multi-channel encoding operation operator, which is equal to the product of the coil sensitivity matrix and the undersampling Fourier transform operator, X is the image to be reconstructed, Y is the acquired K-space data, ||·|| F represents the Frobenius norm, ||·|| * represents the nuclear norm, λ 1,i and λ 2,j are the regularization parameters, and are the structure tensor and parameter tensor respectively;

[0031] The optimization problem of the image reconstruction model is iteratively solved through low-rank tensor decomposition and alternating direction multiplier method.

[0032] In a second aspect, an embodiment of the present application further provides a magnetic resonance quantitative imaging device based on image structure similarity and physical relaxation prior, the device comprising:

[0033] A low-rank structure tensor unit is used to extract image blocks with structural similarity based on a block matching method and construct a low-rank structure tensor based on the image blocks;

[0034] Low-rank parameter tensor unit, used to construct low-rank parameter tensor for weighted image using signal physical relaxation prior;

[0035] A solution unit is established for establishing a low-rank tensor-based image reconstruction model and solution method through a low-rank structure tensor and a low-rank parameter tensor.

[0036] In a third aspect, an embodiment of the present application further provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, it implements any of the methods described in the embodiments of the present application.

[0037] In a fourth aspect, an embodiment of the present application further provides a computer device and a computer-readable storage medium on which a computer program is stored, wherein the computer program is used to implement any method described in the embodiments of the present application when the computer program is executed by a processor.

[0038] Beneficial effects of the present invention:

[0039] The magnetic resonance quantitative imaging method based on image structural similarity and physical relaxation priors provided by the present invention greatly accelerates data scanning speed and reduces quantitative imaging time. During image reconstruction, the reconstruction method proposed by the present invention can still accurately reconstruct parameter-weighted images from highly undersampled data at a high acceleration factor. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Other features, objects and advantages of the present application will become more apparent upon reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0041] Figure 1 A schematic diagram of a process for a magnetic resonance quantitative imaging method based on image structure similarity and physical relaxation prior provided in an embodiment of the present application is shown;

[0042] Figure 2 FIG2 shows an exemplary structural block diagram of a magnetic resonance quantitative imaging apparatus 200 based on image structure similarity and physical relaxation prior according to an embodiment of the present application;

[0043] Figure 3 A schematic diagram of the structure of a computer system of a terminal device suitable for implementing an embodiment of the present application is shown. DETAILED DESCRIPTION

[0044] To make the above-mentioned objects, features, and advantages of the present invention more readily apparent, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings. The following description sets forth numerous specific details to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways than those described herein, and those skilled in the art may make similar modifications without departing from the scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0045] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0046] Furthermore, the terms "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 the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0047] In the present invention, unless otherwise specified or limited, the terms "installed," "connected," "connect," "fixed," etc. should be understood in a broad sense. For example, they can refer to fixed connection, detachable connection, or integration; mechanical connection, electrical connection; direct connection, or indirect connection through an intermediate medium; internal communication between two components, or interaction between two components, unless otherwise specified. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0048] In the present invention, unless otherwise expressly specified or limited, when a first feature is "above" or "below" a second feature, it may mean that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediary. Furthermore, when a first feature is "above," "above," or "above" a second feature, it may mean that the first feature is directly above or diagonally above the second feature, or simply means that the first feature is at a higher level than the second feature. When a first feature is "below," "below," or "below" a second feature, it may mean that the first feature is directly below or diagonally below the second feature, or simply means that the first feature is at a lower level than the second feature.

[0049] It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it may be directly on the other element or there may be an intermediate element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be an intermediate element. The terms "vertical," "horizontal," "upper," "lower," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only implementation methods.

[0050] Please refer to Figure 1 , Figure 1 A flow chart of a magnetic resonance quantitative imaging method based on image structure similarity and physical relaxation prior provided in an embodiment of the present application is shown.

[0051] like Figure 1 As shown, the method includes:

[0052] Step 110: extracting image blocks with structural similarity based on a block matching method, and constructing a low-rank structure tensor based on the image blocks;

[0053] Step 120, constructing a low-rank parameter tensor for the weighted image using the signal physical relaxation prior;

[0054] Step 130: Establish a low-rank tensor-based image reconstruction model and solution method through the low-rank structure tensor and the low-rank parameter tensor.

[0055] The above technical solution greatly accelerates the data scanning speed and reduces the quantitative imaging time. During image reconstruction, the reconstruction method proposed in the present invention can still accurately reconstruct parameter-weighted images from highly undersampled data at a higher acceleration factor.

[0056] In some embodiments, the block matching method of the present invention is used to extract image blocks with structural similarity, including: defining T 1ρ Weighted image, where N x 、N y 、N TSL are the number of pixels along the frequency encoding and phase encoding directions, and the number of TSLs respectively; all image blocks within the search neighborhood r×r in The similarity between image patches is measured by the normalized L2 distance, which is defined as: when The distance is less than the threshold λ m When , the image block is defined and B i is the image block with structural similarity. The number of similar image blocks to be searched is limited to N patch, all similar image blocks are grouped into a block of size b 2 ×N patch where b is the size of the image block (b×b).

[0057] Specifically, the present invention extracts image blocks with structural similarity based on the block matching method, and the specific method is: define T 1ρ Weighted image, where N x 、N y 、N TSL are the number of pixels along the frequency encoding and phase encoding directions, and the number of TSLs, respectively. For the reference block B of size b×b at pixel i, i , search for all image blocks within its neighborhood r×r in The similarity between image patches is measured by the normalized L2 distance, which is defined as: When the distance in the formula is less than the threshold λ m When the image block and B i are similar. The number of similar image blocks found is limited to N patch , then all similar image blocks can form a block of size b 2 ×N patch The matrix of .

[0058] In some embodiments, the present invention constructs a low-rank structure tensor based on an image block, including: extracting all T 1ρ Similar image blocks at a given pixel point i in the weighted image; construct a third-order tensor with low-rank tensor properties based on all similar image blocks Set P i is the similar image block extraction operator, then

[0059] Specifically, due to T 1ρ The weighted image decays exponentially in the TSL direction, and its relaxation law is as follows: Where TSL represents the spin lock time, M represents the signal after TSL time, and M0 is a constant representing the reference signal when no preparation pulse is applied. 1ρ The weighted image data has high redundancy in the TSL direction. Using this relaxation prior information, all T 1ρ The similar image blocks at a given pixel point i in the weighted image are extracted to form a third-order tensor with low rank characteristics Define P i is the similar image block extraction operator, then

[0060] In some embodiments, the present invention utilizes the signal physical relaxation prior to construct a low-rank parameter tensor for the weighted image, including: 1ρ The principle of quantitative imaging, based on the formula The model is used to fit each pixel in the weighted image point by point; TSL represents the spin lock time, M represents the signal after TSL time, and M0 is a constant representing the reference signal when no preparation pulse is applied; according to the estimated tissue T 1ρ The quantitative value is used to cluster the pixels in the image so that the pixels in each cluster group belong to the same type of tissue; for the clustered pixels, different T 1ρ Pixels at the same position in the weighted image form a low-rank Hankel matrix; the Hankel matrix is ​​stacked to form a third-order parameter tensor with low-rank characteristics.

[0061] Specifically, according to T 1ρ The principle of quantitative imaging, based on the formula The relaxation model in is fitted point by point to each pixel in the weighted image to obtain the estimated tissue T 1ρ Quantitative value. Due to the T of the same type of tissue 1ρ The quantitative values ​​are equal, so the estimated tissue T 1ρ The quantitative values ​​are used to perform cluster analysis on the pixels in the image. The present invention uses a histogram analysis method to cluster the image pixels so that the pixels in each cluster group belong to the same type of tissue.

[0062] The present invention utilizes the physical relaxation prior of the signal to transform different T 1ρ The pixels at the same position in the weighted image form a low-rank Hankel matrix. Let the matrix be H[I(r)], where r represents the position of the pixel point and I(r) represents the pixel point at position r in the weighted image. Then we have

[0063]

[0064] Where I(r,TSL K ) indicates the TSL K The pixel at position r in the image, K is defined as greater than or equal to N TSL For each pixel belonging to the same type of tissue, a low-rank Hankel matrix is ​​constructed along the TSL direction according to the above formula, and finally these Hankel matrices are stacked to form a third-order parameter tensor with low-rank characteristics. Let N g is the number of cluster groups, N tissue is the number of pixels in each group belonging to the same type of tissue, is the parameter tensor of the j-th group, Construct an operation operator for the parameter tensor, then we have

[0065] In some embodiments, the present invention establishes an image reconstruction model based on a low-rank tensor and a solution method by using a low-rank structure tensor and a low-rank parameter tensor, including: establishing an image reconstruction model based on a low-rank tensor according to the low-rank structure tensor and the low-rank parameter tensor

[0066]

[0067]

[0068] Where, E is the multi-channel encoding operation operator, which is equal to the product of the coil sensitivity matrix and the undersampling Fourier transform operator, X is the image to be reconstructed, Y is the acquired K-space data, ||·|| F represents the Frobenius norm, ||·|| * represents the nuclear norm, λ 1,i and λ 2,j are the regularization parameters, and They are the structure tensor and parameter tensor respectively; the optimization problem of the image reconstruction model is iteratively solved by low-rank tensor decomposition and alternating direction multiplier method.

[0069] Specifically, the present invention combines the low-rank structure tensor and the parameter tensor to establish an image reconstruction model based on the low-rank tensor, and the model is as follows:

[0070]

[0071]

[0072] Where E is the multi-channel encoding operation operator, which is equal to the product of the coil sensitivity matrix and the undersampling Fourier transform operator, X is the image to be reconstructed, Y is the acquired K-space data, ||·|| F represents the Frobenius norm, ||·|| * represents the nuclear norm, λ 1,i and λ 2,j are the regularization parameters, and are the structure tensor and parameter tensor respectively. After introducing the Lagrange multiplier, the above formula can be transformed into:

[0073]

[0074] Among them, μ1 and μ2 penalty factors, α 1,i and α 2,j is the Lagrange multiplier, the formula is equivalent to:

[0075]

[0076] The present invention uses an alternating direction method of multipliers (ADMM) based on low rank tensor decomposition (HOSVD decomposition) to iteratively solve the optimization problem of the formula.

[0077] In the present invention, since the image signal decays exponentially in the parameter direction, its weighted image sequence has high redundancy in the spin lock time (TSL) direction. By utilizing this relaxation prior information, the present invention extracts image blocks with structural similarity based on the block matching method, thereby constructing a structure tensor.

[0078] Based on the physical relaxation prior of the signal, a low-rank Hankel matrix can be constructed for each pixel in the weighted image along the TSL direction. 1ρ The quantitative values ​​are equal, and a T can be estimated from the weighted image based on the physical relaxation model. 1ρ Parameter diagram, according to T in the parameter diagram 1ρ The quantitative values ​​are used to perform cluster analysis on the image pixels, and the parameter tensor of the clustered groups is constructed using the Hankel matrix.

[0079] By combining the structure tensor and parameter tensor, an image reconstruction model based on low-rank tensor is established. Low-rank tensor decomposition (HOSVD decomposition) is used to design an image reconstruction solution method based on the low-rank tensor.

[0080] Further, refer to Figure 2 , Figure 2 FIG2 shows an exemplary structural block diagram of a magnetic resonance quantitative imaging apparatus 200 based on image structure similarity and physical relaxation prior according to an embodiment of the present application.

[0081] like Figure 2 As shown, the device includes:

[0082] A low-rank structure tensor unit 210 is configured to extract image blocks with structural similarity based on a block matching method and construct a low-rank structure tensor based on the image blocks;

[0083] A low-rank parameter tensor unit 220 is used to construct a low-rank parameter tensor for the weighted image using a signal physical relaxation prior;

[0084] A solution unit 230 is established to establish a low-rank tensor-based image reconstruction model and a solution method through a low-rank structure tensor and a low-rank parameter tensor.

[0085] It should be understood that the units or modules described in the apparatus 200 are similar to those described in the reference Figure 1 The various steps in the described method correspond to each other. Therefore, the operations and features described above for the method are also applicable to the device 200 and the units contained therein, and will not be repeated here. The device 200 can be pre-implemented in the browser or other security application of the electronic device, or loaded into the browser or its security application of the electronic device by downloading or other means. The corresponding units in the device 200 can cooperate with the units in the electronic device to implement the solution of the embodiment of the present application.

[0086] Reference below Figure 3 , which shows a structural diagram of a computer system 300 suitable for implementing a terminal device or server of an embodiment of the present application.

[0087] like Figure 3 As shown, the computer system 300 includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage unit 308 into a random access memory (RAM) 303. Various programs and data required for the operation of the system 300 are also stored in the RAM 303. The CPU 301, ROM 302, and RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0088] The following components are connected to the I / O interface 305: an input section 306 including a keyboard, a mouse, and the like; an output section 307 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and a speaker; a storage section 308 including a hard disk; and a communication section 309 including a network interface card such as a LAN card or a modem. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. Removable media 311, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 310 as needed, so that computer programs read therefrom can be installed into the storage section 308 as needed.

[0089] In particular, according to the embodiments of the present disclosure, the above reference Figure 1 The described process can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a method for magnetic resonance quantitative imaging based on image structure similarity and physical relaxation priors, which includes a computer program tangibly embodied on a machine-readable medium, the computer program including a computer program for executing Figure 1In such an embodiment, the computer program may be downloaded and installed from a network via the communication section 309 and / or installed from the removable medium 311 .

[0090] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the aforementioned module, program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0091] The units or modules involved in the embodiments described in the present application can be implemented by software or by hardware. The described units or modules can also be set in a processor. For example, they can be described as: a processor includes a first sub-area generation unit, a second sub-area generation unit, and a display area generation unit. Among them, the names of these units or modules do not constitute a limitation of the unit or module itself under certain circumstances. For example, the display area generation unit can also be described as "a unit for generating a display area for text based on the first sub-area and the second sub-area".

[0092] As another aspect, the present application further provides a computer-readable storage medium, which may be the computer-readable storage medium included in the aforementioned apparatus in the above-mentioned embodiment, or may be a separate computer-readable storage medium not incorporated into the apparatus. The computer-readable storage medium stores one or more programs, which are used by one or more processors to execute the text generation method for a transparent window envelope described in the present application.

[0093] The above description is merely a preferred embodiment of the present application and an illustration of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to the technical solutions formed by the specific combination of the above-mentioned technical features, but also encompasses other technical solutions formed by any combination of the above-mentioned technical features or their equivalents without departing from the aforementioned inventive concept. For example, a technical solution formed by replacing the above-mentioned features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A magnetic resonance quantitative imaging method based on image structure and physical relaxation priors, characterized in that: The method includes: Image blocks with structural similarity are extracted based on the block matching method, and a low-rank structure tensor is constructed based on the image blocks; Using the signal physics relaxation prior, a low-rank parameter tensor is constructed for the weighted image; Establish a low-rank tensor-based image reconstruction model and solution method through low-rank structure tensor and low-rank parameter tensor; The method of constructing a low-rank parameter tensor for a weighted image by utilizing a signal physical relaxation prior includes: According to T 1ρ The principle of quantitative imaging, based on the formula M=M0e -TSLk / T1ρ (k=1,2...,N TSL ) performs point-by-point fitting on each pixel in the weighted image; wherein TSL represents the spin lock time, M represents the signal after TSL time, and M0 is a constant, indicating that T is not applied. 1ρ Reference signal when preparing pulse; According to the estimated organization T 1ρ The quantitative values ​​are used to cluster the pixels in the image so that the pixels in each cluster group belong to the same type of tissue; For the clustered pixels, different T 1ρ The pixels at the same position in the weighted image form a low-rank Hankel matrix; The Hankel matrix is ​​stacked to form a third-order parameter tensor with low rank characteristics; The image reconstruction model and solution method based on the low-rank tensor are established by using the low-rank structure tensor and the low-rank parameter tensor, including: An image reconstruction model based on a low-rank tensor is established according to the low-rank structure tensor and the low-rank parameter tensor. Where E is the multi-channel encoding operation operator, which is equal to the product of the coil sensitivity matrix and the undersampling Fourier transform operator, X is the image to be reconstructed, Y is the acquired K-space data, ||·|| represents the Frobenius norm, ||·|| * represents the nuclear norm, λ 1,i and λ 2,j are the regularization parameters, and are the structure tensor and parameter tensor respectively; The optimization problem of the image reconstruction model is iteratively solved through low-rank tensor decomposition and alternating direction multiplier method.

2. The magnetic resonance quantitative imaging method based on image structure and physical relaxation prior according to claim 1, characterized in that: The method of extracting image blocks with structural similarity based on a block matching method includes: definition T 1ρ Weighted image, where N x 、N y 、N TSL are the number of pixels along the frequency encoding and phase encoding directions, and the number of TSLs respectively; Search all image patches within the r×r neighborhood in The similarity between image patches is measured by the normalized L2 distance, which is defined as: when The distance is less than the threshold λ m When , the image block is defined and B i are image patches with structural similarity.

3. The magnetic resonance quantitative imaging method based on image structure and physical relaxation prior according to claim 2, characterized in that: When The distance is less than the threshold λ m Afterwards, the method further comprises: Limit the number of similar image blocks to be searched to N patch , all similar image blocks are grouped into a block of size b 2 ×N patch The matrix of b 2 is the size of the image block (b×b).

4. The magnetic resonance quantitative imaging method based on image structure and physical relaxation prior according to claim 3, characterized in that: The constructing of a low-rank structure tensor according to the image block includes: Extract all T 1ρ Similar image blocks at a given pixel i in the weighted image; Construct a third-order tensor with low-rank properties based on all similar image patches Set P i is the similar image block extraction operator, then 5. A magnetic resonance quantitative imaging device based on image structure and physical relaxation priors, characterized in that: The device includes: A low-rank structure tensor unit is used to extract image blocks with structural similarity based on a block matching method and construct a low-rank structure tensor based on the image blocks; Low-rank parameter tensor unit, used to construct low-rank parameter tensor for weighted image using signal physical relaxation prior; Establishing a solution unit for establishing a low-rank tensor-based image reconstruction model and a solution method through a low-rank structure tensor and a low-rank parameter tensor; The method of constructing a low-rank parameter tensor for a weighted image by utilizing a signal physical relaxation prior includes: According to T 1ρ The principle of quantitative imaging, based on the formula Fit each pixel in the weighted image point by point; where TSL represents the spin lock time, M represents the signal after TSL time, and M0 is a constant, indicating that T is not applied. 1ρ Reference signal when preparing pulse; According to the estimated organization T 1ρ The quantitative values ​​are used to cluster the pixels in the image so that the pixels in each cluster group belong to the same type of tissue; For the clustered pixels, different T 1ρ The pixels at the same position in the weighted image form a low-rank Hankel matrix; The Hankel matrix is ​​stacked to form a third-order parameter tensor with low rank characteristics; The image reconstruction model and solution method based on the low-rank tensor are established by using the low-rank structure tensor and the low-rank parameter tensor, including: An image reconstruction model based on a low-rank tensor is established according to the low-rank structure tensor and the low-rank parameter tensor. Where E is the multi-channel encoding operation operator, which is equal to the product of the coil sensitivity matrix and the undersampling Fourier transform operator, X is the image to be reconstructed, Y is the acquired K-space data, ||·|| represents the Frobenius norm, ||·|| * represents the nuclear norm, λ 1,i and λ 2,j are the regularization parameters, and are the structure tensor and parameter tensor respectively; The optimization problem of the image reconstruction model is iteratively solved through low-rank tensor decomposition and alternating direction multiplier method.

6. The magnetic resonance quantitative imaging device based on image structure and physical relaxation prior according to claim 5, characterized in that: The method of extracting image blocks with structural similarity based on a block matching method includes: definition T 1ρ Weighted image, where N x 、N y 、N TSL are the number of pixels along the frequency encoding and phase encoding directions, and the number of TSLs respectively; Search all image patches within the r×r neighborhood in The similarity between image patches is measured by the normalized L2 distance, which is defined as: when The distance is less than the threshold λ m When , the image block is defined and B i are image patches with structural similarity.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method according to any one of claims 1 to 4 is implemented.

8. A computer-readable storage medium having stored thereon a computer program for: When the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images