Motion parameter determination method and device, computer equipment and storage medium

Abstract

The invention relates to a motion parameter determination method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring first K space data obtained by sampling a target part, performing image reconstruction on first preset K space data in the first K space data to obtain a first reconstructed image, performing motion parameter estimation according to the first K space data and the first reconstructed image, and determining a target motion parameter of the target part. The sampling frequency of the first preset K space data is smaller than the preset sampling frequency, the first K space data comprises the sampling data passing through the K space center, and the sampling density of the K space center is larger than the preset sampling density, so that the first reconstructed image obtained based on the first preset K space data is a low-resolution reconstructed image; and a real motion-free image can be more easily converged. Therefore, the target motion parameters of the target part can be obtained with less iteration operation times, and the time consumption of iteration operation is reduced.

Description

Technical Field

[0001] The present application relates to the field of medical technology, and in particular to a method, apparatus, computer device, and storage medium for determining motion parameters. Background Art

[0002] During MRI, patient rigid motion can cause various artifacts in the images, such as ghosting, blurring, geometric distortion, or reduced signal-to-noise ratio. This can result in up to 20% of the image data being non-diagnostic or poorly diagnostic. In current clinical practice, to ensure successful MRI, patients are often sedated or anestheticed, but this approach can have adverse effects on patients. Therefore, correction for rigid motion in MRI is necessary.

[0003] At present, by incorporating the rigid motion transformation matrix into the forward signal model, a separable nonlinear least squares problem can be constructed. On this basis, the magnetic resonance image and motion parameters are iteratively solved based on the patient's K-space data to correct the rigid body motion of magnetic resonance imaging.

[0004] However, the method of obtaining motion parameters based on the above method has the problems of complex iteration and time-consuming operation. Summary of the Invention

[0005] Based on this, it is necessary to provide a motion parameter determination method, device, computer equipment and storage medium that can reduce the number of iterative operations and shorten the operation time in order to address the above technical problems.

[0006] In a first aspect, the present application provides a method for determining motion parameters, comprising:

[0007] Acquire first K-space data obtained by sampling the target part; the first K-space data includes sampling data passing through the center of the K-space; the sampling density of the center of the K-space is greater than a preset sampling density;

[0008] performing image reconstruction on first preset K-space data in the first K-space data to obtain a first reconstructed image; the first preset K-space data includes first K-space data whose distance from the center of the K-space is less than or equal to a preset distance, and whose sampling frequency is less than a preset sampling frequency;

[0009] Motion parameter estimation is performed based on the first K-space data and the first reconstructed image to determine target motion parameters of the target part.

[0010] In one embodiment, the performing motion parameter estimation based on the first K-space data and the first reconstructed image to determine the target motion parameter of the target part includes:

[0011] Dividing the first K-space data based on a first time resolution to obtain second K-space data;

[0012] For each piece of the second K-space data, motion parameter estimation is performed according to the second K-space data and the first reconstructed image to determine the target motion parameter of the target part.

[0013] In one embodiment, the performing motion parameter estimation based on the second K-space data and the first reconstructed image to determine the target motion parameter of the target part includes:

[0014] performing motion parameter estimation based on the second K-space data and the first reconstructed image to obtain initial motion parameters corresponding to the second K-space data;

[0015] Correcting the first K-space data according to the initial motion parameters to obtain third K-space data;

[0016] Motion parameter estimation is performed based on the third K-space data to determine target motion parameters of the target part.

[0017] In one embodiment, the estimating motion parameters based on the third K-space data to determine target motion parameters of the target part includes:

[0018] performing image reconstruction on second preset K-space data in the third K-space data to obtain a second reconstructed image, wherein the second preset K-space data includes third K-space data whose distance from the K-space center is less than or equal to the preset distance;

[0019] Dividing the third K-space data based on a second time resolution to obtain fourth K-space data; wherein the second time resolution is higher than the first time resolution;

[0020] For each of the fourth K-space data, motion parameter estimation is performed according to the fourth K-space data and the second reconstructed image to determine the target motion parameter of the target part.

[0021] In one embodiment, the performing motion parameter estimation based on the fourth K-space data and the second reconstructed image to determine the target motion parameter of the target part includes:

[0022] Based on the second reconstructed image, the fourth K-space data, a preset first regularization strength parameter and a regularization operator along the time dimension, constrained optimization processing is performed on the motion parameters corresponding to the fourth K-space data to obtain the target motion parameters.

[0023] In one embodiment, the method further comprises:

[0024] Reconstructing the first K-space data to obtain an image to be corrected;

[0025] Based on the first K-space data, the target motion parameter, a preset second regularization strength parameter and a regularization operator along the spatial dimension, constrained optimization processing is performed on the image to be corrected to obtain a target corrected image.

[0026] In a second aspect, the present application further provides a motion parameter determination device, comprising:

[0027] An acquisition module is configured to acquire first K-space data obtained by sampling a target part; the first K-space data includes sampling data passing through a center of the K-space; and a sampling density of the center of the K-space is greater than a preset sampling density;

[0028] a first reconstruction module, configured to perform image reconstruction on first preset K-space data in the first K-space data to obtain a first reconstructed image; the first preset K-space data includes first K-space data whose distance from the center of the K-space is less than or equal to a preset distance, and whose sampling frequency is less than a preset sampling frequency;

[0029] The parameter estimation module is used to perform motion parameter estimation based on the first K-space data and the first reconstructed image to determine the target motion parameter of the target part.

[0030] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the method steps provided in the first aspect when executing the computer program.

[0031] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which implements the method steps provided in the first aspect when the computer program is executed by a processor.

[0032] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which implements the method steps provided in the first aspect when executed by a processor.

[0033] The above-described motion parameter determination method, apparatus, computer device, and storage medium obtain first K-space data obtained by sampling a target part, perform image reconstruction on first preset K-space data within the first K-space data to obtain a first reconstructed image, and perform motion parameter estimation based on the first K-space data and the first reconstructed image to determine target motion parameters of the target part. The sampling density at the center of the K-space is greater than the preset sampling density, the first preset K-space data includes first K-space data whose distance from the center of the K-space is less than or equal to the preset distance, and the sampling frequency of the first preset K-space data is less than the preset sampling frequency. Because the sampling frequency of the first preset K-space data is less than the preset sampling frequency, the first K-space data includes sampled data passing through the center of the K-space, and the sampling density at the center of the K-space is greater than the preset sampling density, the first reconstructed image obtained based on the first preset K-space data is a low-resolution reconstructed image. The reduced spatial resolution and oversampling caused by low-resolution reconstruction make the first reconstructed image closer to a true motion-free image and more easily converge to a true motion-free image. Therefore, the target motion parameters of the target part can be determined based on the first K-space data and the first reconstructed image with fewer iterations, reducing the time required for iterations. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0035] Figure 1 A diagram illustrating an application environment of a method for determining motion parameters in one embodiment;

[0036] Figure 2 1 is a flow chart of a method for determining motion parameters in one embodiment;

[0037] Figure 3 is a schematic diagram of a reconstructed image containing motion artifacts in one embodiment;

[0038] Figure 4 is a schematic diagram of a first reconstructed image in one embodiment;

[0039] Figure 5 Schematic diagram of motion parameter diagram under different translation and rotation estimation errors;

[0040] Figure 6 1 is a flow chart of a method for determining target motion parameters in one embodiment;

[0041] Figure 71 is a flow chart of a method for determining target motion parameters in another embodiment;

[0042] Figure 8 1 is a flow chart of a method for determining target motion parameters in another embodiment;

[0043] Figure 9 is a schematic flow chart of an image reconstruction method in one embodiment;

[0044] Figure 10 A schematic diagram of translation estimation error in one embodiment;

[0045] Figure 11 is a schematic diagram of rotation estimation error in one embodiment;

[0046] Figure 12 is a schematic diagram of motion correction comparison results in one embodiment;

[0047] Figure 13 A schematic diagram of correcting image and motion parameters in one embodiment;

[0048] Figure 14 FIG. 1 is a structural block diagram of a motion parameter determination device in one embodiment. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0050] MRI is highly sensitive to patient motion, presenting significant challenges in clinical brain MRI scanning. Motion typically produces a range of artifacts in the resulting images, including ghosting, blurring, geometric distortion, and reduced signal-to-noise ratio. The extent and characteristics of these artifacts depend on the specific k-space acquisition strategy employed and the type of motion. In MRI, subject motion can render up to 20% of imaging data suboptimal or non-diagnostic, leading to delayed treatment, increased costs, and even false-positive or false-negative results. Motion during MRI scanning is more prevalent in certain populations, who often require additional attention and care during imaging. In current clinical practice, sedatives or anesthesia are often used for young children to ensure successful imaging. However, this approach can have adverse effects on patients.

[0051] The motion parameter determination method provided in the embodiment of the present application can be applied to Figure 1 The application environment shown in FIG. The application environment includes a computer device, which may be a server, and its internal structure diagram may be as shown in FIG. Figure 1As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store relevant data for determining motion parameters. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for determining motion parameters is implemented. The server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud computing services.

[0052] Those skilled in the art will understand that Figure 1 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0053] In an exemplary embodiment, Figure 2 As shown, a motion parameter determination method is provided, which is applied to Figure 1 The computer device in the example is used to illustrate, including the following S201 to S203.

[0054] S201, acquiring first K-space data obtained by sampling a target part; the first K-space data includes sampling data passing through the center of the K-space; the sampling density of the center of the K-space is greater than a preset sampling density.

[0055] Optionally, target parts where rigid body motion may occur include but are not limited to the head, limbs, abdomen, etc.

[0056] The first k-space data includes sampled data passing through the center of k-space. Therefore, sampling can be performed using either a radial or spiral sampling trajectory. Compared to a Cartesian sampling pattern, the radial sampling trajectory significantly oversamples the k-space center, effectively averaging motion effects and mitigating motion artifacts. As a result, motion artifacts are less prominent, making motion-blurred images more similar to motion-free images.

[0057] The sampling density of the center of the K space is greater than the preset sampling density, that is, the center of the K space is densely sampled.

[0058] In an embodiment of the present application, the first K-space data obtained by sampling the target part can be received in real time, or the first K-space data obtained by sampling the target part can be obtained after receiving a user-triggered instruction; or the first K-space data of the target part can be obtained based on a preset period.

[0059] S202, performing image reconstruction on first preset K space data in the first K space data to obtain a first reconstructed image; the first preset K space data includes first K space data whose distance from the K space center is less than or equal to a preset distance, and the sampling frequency of the first preset K space data is less than a preset sampling frequency.

[0060] Optionally, the preset area may be a circle with the center of the K space as the center and the preset distance as the radius; the preset area may also be a rectangular frame with the center of the K space as the center point and the preset distance as the side length, etc. For example, the preset distance is 1 / 4 of the spatial resolution.

[0061] Motion parameters can be successfully estimated using highly undersampled, low-resolution k-space data. The radial sampling trajectory itself is well-suited for low-resolution reconstruction using only the central portion of the acquired k-space data. Therefore, in this embodiment of the present application, a non-uniform fast Fourier transform can be used to perform image reconstruction on the first preset k-space data to obtain a low-resolution first reconstructed image.

[0062] Optionally, an image reconstruction algorithm based on deep learning, a reconstruction algorithm based on parallel imaging technology, a reconstruction algorithm based on low-rank matrix recovery, etc. can also be used.

[0063] S203 , performing motion parameter estimation based on the first K-space data and the first reconstructed image to determine target motion parameters of the target part.

[0064] By incorporating the rigid body transformation matrix T into the forward signal model, the alignment reconstruction formula can be expressed as:

[0065] (1)

[0066] Among them, x is the image to be corrected, T is the rigid body transformation matrix, and the rigid body transformation matrix is ​​about the motion parameters (3 translations, 3 rotations), i represents the number of segments of K-space data divided based on time resolution, S is the coil sensitivity, F is the Fourier transform matrix, A iis the undersampling mask corresponding to the i-th segment of K-space data, and y is the acquired K-space data. Assume that the motion parameters remain unchanged within the same segment. The two variables in the above formula (1) can be solved alternately using the following two formulas:

[0067] (2)

[0068] (3)

[0069] Where n is the number of iterations, formula (2) is the image reconstruction problem for a given rigid body transformation matrix, and formula (3) is the motion update problem for a given reconstructed image. Formula (3) can be solved by making the objective function relative to By iteratively solving these two formulas, the image to be reconstructed and the motion parameters can be gradually improved.

[0070] Since the present application performs image reconstruction through the first K-space data in a preset area to obtain a low-resolution first reconstructed image, the determination of the target motion parameters based on the first reconstructed image can be converted into a convex motion parameter update problem, so that the motion parameters obtained based on the first reconstructed image are close to the global minimum of the real motion, that is, the target motion parameters.

[0071] To verify the convexity of the proposed motion update problem, simulations were conducted. The image matrix used for the simulations was 240x240x240 pixels, and 12,000 32-channel 3D radial k-space data were generated using a multidimensional golden section trajectory. Motions of ±5 mm and ±5° were introduced by translating and rotating the image accordingly. 100 radial sampling lines (only one segment) were used, and a range of different translation and rotation combinations along the motion axis were tested.

[0072] like Figure 3 As shown in Figure 1, for scans containing motion, the image reconstructed based on non-uniform fast Fourier transform has severe undersampling and motion artifacts. If only one-quarter of the K-space data in the center of the K-space is used for reconstruction, the motion artifacts will be greatly reduced due to the reduction in spatial resolution. In addition, due to the oversampling of the center of the K-space, the undersampling rate is reduced by a factor of 16, making the reconstructed image oversampled. Figure 4 As shown, Figure 4 This is the result of image reconstruction based on 1 / 4 spatial resolution, as only the K-space data corresponding to the central quarter of the K-space spatial resolution is used. Due to the reduced spatial resolution, motion artifacts are weaker, and low-resolution reconstruction also leads to oversampling, so no undersampling artifacts are observed. Figure 5Figure 2 shows the motion parameters for different translation and rotation estimation errors. For ease of visualization, only translation / rotation along the motion axis is simulated, but similar results can be observed in other directions. It can be seen that in this case, the motion update problem is convex, and the global minimum is close to the true motion value.

[0073] Therefore, in this case, The first reconstructed image, y i is the first K-space data (y i Alternatively, the K-space data of each segment obtained by dividing the first K-space data based on the time resolution is obtained. The motion parameter obtained based on formula (3), the first reconstructed image, and the first K-space data is the minimum value. In this case, the number of iterations n is 0, and only one calculation is required to obtain the target motion parameter.

[0074] In another possible implementation, in order to further improve the accuracy of the target motion parameters, the number of iterations n can be set to 1. That is, based on formula (3), the first reconstructed image, and the first K-space data, the first motion parameters are obtained, and the first K-space data is corrected based on the first motion parameters. Image reconstruction is performed based on the corrected first K-space data in the preset area to obtain a new first reconstructed image Based on formula (3), the new first reconstructed image, and the first K-space data, a second motion parameter is obtained, and the second motion parameter is a target motion parameter.

[0075] In the above-described motion parameter determination method, first K-space data obtained by sampling a target part is acquired, first preset K-space data within the first K-space data is reconstructed to obtain a first reconstructed image, and motion parameters are estimated based on the first K-space data and the first reconstructed image to determine target motion parameters of the target part. The sampling density at the center of the K-space is greater than the preset sampling density, the first preset K-space data includes first K-space data whose distance from the center of the K-space is less than or equal to the preset distance, and the sampling frequency of the first preset K-space data is less than the preset sampling frequency. Because the sampling frequency of the first preset K-space data is less than the preset sampling frequency, the first K-space data includes sampled data passing through the center of the K-space, and the sampling density at the center of the K-space is greater than the preset sampling density, the first reconstructed image obtained based on the first preset K-space data is a low-resolution reconstructed image. The reduced spatial resolution and oversampling caused by the low-resolution reconstruction make the first reconstructed image closer to a true motion-free image and more easily converge to a true motion-free image. Therefore, the target motion parameters of the target part can be determined based on the first K-space data and the first reconstructed image with fewer iterations, reducing the time required for iterations.

[0076] Figure 6 FIG. 1 is a flow chart of a method for determining target motion parameters in one embodiment. Figure 6 As shown, the embodiment of the present application relates to a possible implementation method of how to estimate motion parameters based on first K-space data and a first reconstructed image to determine target motion parameters of a target part, including the following steps:

[0077] S601 : Divide the first K-space data based on a first time resolution to obtain second K-space data.

[0078] Optionally, the first time resolution may be 500 sampling lines, 400 sampling lines, etc.

[0079] S602 : For each second K-space data, estimate motion parameters according to the second K-space data and the first reconstructed image to determine target motion parameters of the target part.

[0080] In the embodiment of the present application, for each second K-space data, the motion parameters can be estimated based on the second K-space data and the first reconstructed image on the basis of the above formula (3), and the initial motion parameters corresponding to the second K-space data can be obtained, and each initial motion parameter is used as the target motion parameter. In this iteration, y i is the second K-space data, is the first reconstructed image.

[0081] In one possible implementation, after motion parameters are estimated based on the second K-space data and the first reconstructed image to obtain initial motion parameters corresponding to the second K-space data, the initial motion parameters are used to correct the first K-space data for the corresponding segment to obtain third K-space data. Motion parameters are estimated based on the third K-space data to determine target motion parameters for the target area.

[0082] In the embodiment of the present application, the first K-space data is divided based on the first temporal resolution to obtain respective second K-space data. For each second K-space data, motion parameters are estimated based on the second K-space data and the first reconstructed image to determine the target motion parameters of the target part. In the embodiment of the present application, by dividing the first K-space data to obtain respective second K-space data, and estimating the motion parameters for each second K-space data, the target motion parameters are determined more efficiently.

[0083] Figure 7 FIG. 1 is a flow chart of a method for determining target motion parameters in another embodiment. Figure 7 As shown, the embodiment of the present application relates to a possible implementation method of how to estimate motion parameters based on the second K-space data and the first reconstructed image to determine the target motion parameters of the target part, including the following steps:

[0084] S701 : Estimating motion parameters according to the second K-space data and the first reconstructed image to obtain initial motion parameters corresponding to the second K-space data.

[0085] In the embodiment of the present application, the second K-space data and the first reconstructed image are input into the above formula (3) to obtain the initial motion parameters corresponding to the second K-space data.

[0086] S702: Correct the first K-space data according to the initial motion parameters to obtain third K-space data.

[0087] In an embodiment of the present application, the initial motion parameters include rotation parameters and translation parameters. The first K-space data can be subjected to rotation correction based on the rotation parameters to obtain intermediate K-space data. The intermediate K-space data can then be subjected to translation correction based on the first K-space data and the translation parameters to obtain third K-space data.

[0088] S703: Estimating motion parameters based on the third K-space data to determine target motion parameters of the target part.

[0089] In the embodiment of the present application, the second preset K-space data in the third K-space data is used to perform image reconstruction to obtain a second reconstructed image, and motion parameters are estimated based on the second reconstructed image and the first K-space data to obtain target motion parameters. The area where the second preset K-space data is located is the same as the area where the first preset K-space data is located. In this iteration, y i is the first K-space data, is the second reconstructed image.

[0090] In one possible implementation, image reconstruction is performed on the third K-space data within a preset area to obtain a second reconstructed image, and the third K-space data is divided based on the second time resolution to obtain fourth K-space data. For each fourth K-space data, motion parameter estimation is performed based on the second reconstructed image and the fourth K-space data to obtain motion parameters corresponding to each fourth K-space data; wherein the target motion parameters include the motion parameters corresponding to each fourth K-space data. In this iteration, y i is the fourth K-space data, is a second reconstructed image. The second time resolution is greater than the first time resolution.

[0091] In another possible implementation, the third K-space data in the preset area can be reconstructed to obtain a second reconstructed image, and the third K-space data can be divided based on the second time resolution to obtain fourth K-space data. For each fourth K-space data, motion parameter estimation is performed based on the first reconstructed image and the fourth K-space data to obtain motion parameters corresponding to each fourth K-space data. In this iteration, yi is the fourth K-space data, Compared with the initial motion parameters, the temporal resolution of the target motion parameters is improved based on the fourth K-space data and the first reconstructed image. When performing rigid body motion correction, correction can be performed on K-space data with higher temporal resolution.

[0092] In this embodiment of the present application, motion parameters are estimated based on the second K-space data and the first reconstructed image to obtain initial motion parameters corresponding to the second K-space data. The first K-space data is corrected based on the initial motion parameters to obtain third K-space data. Motion parameters are estimated based on the third K-space data to determine target motion parameters for the target part. In this embodiment of the present application, the first K-space data is corrected based on the initial motion parameters to obtain third K-space data, and the target motion parameters are determined using the third K-space data, thereby improving the accuracy of the target motion parameter determination.

[0093] Figure 8 FIG. 1 is a flow chart of a method for determining target motion parameters in another embodiment. Figure 8 As shown, the embodiment of the present application relates to a possible implementation method of how to estimate motion parameters based on third K-space data to determine target motion parameters of a target part, including the following steps:

[0094] S801 , performing image reconstruction on second preset K space data in the third K space data to obtain a second reconstructed image; the second preset K space data includes third K space data whose distance from the K space center is less than or equal to a preset distance.

[0095] In the embodiment of the present application, the specific implementation of the reconstruction method used to perform image reconstruction on the second preset K-space data can be found in the above S202.

[0096] S802 , dividing the third K-space data based on a second time resolution to obtain fourth K-space data; the second time resolution is higher than the first time resolution.

[0097] The second time resolution is higher than the first time resolution. For example, if the first time resolution is 500 sampling lines, the second time resolution may be 100 sampling lines, 200 sampling lines, or the like.

[0098] S803 , for each fourth K-space data, perform motion parameter estimation according to the fourth K-space data and the second reconstructed image to determine a target motion parameter of the target part.

[0099] In the embodiment of the present application, the fourth K-space data and the second reconstructed image can be input into the above formula (3) to obtain the motion parameters corresponding to each fourth K-space data. In this iteration, yi is the fourth K-space data, is the second reconstructed image.

[0100] In this embodiment of the present application, image reconstruction is performed on the second preset K-space data within the third K-space data to obtain a second reconstructed image. The third K-space data is then divided based on the second temporal resolution to obtain fourth K-space data. For each fourth K-space data, motion parameter estimation is performed based on the fourth K-space data and the second reconstructed image to determine target motion parameters for the target part. Because the second reconstructed image has already undergone preliminary motion correction compared to the first reconstructed image, the motion parameters obtained using the second reconstructed image and the fourth K-space data are more accurate.

[0101] In one embodiment, motion parameters are estimated based on the fourth K-space data and the second reconstructed image to determine the target motion parameters of the target part, including: based on the second reconstructed image, the fourth K-space data, a preset first regularization strength parameter and a regularization operator along the time dimension, constrained optimization processing is performed on the motion parameters corresponding to the fourth K-space data to obtain the target motion parameters.

[0102] On the basis of the above formula (3), time continuity can be further introduced to constrain motion parameters. The formula for obtaining motion parameters from the above formula (3) can be further described as follows:

[0103] (4)

[0104] in, is the preset first regularization strength parameter, is the regularization operator along the time dimension.

[0105] In the embodiment of the present application, assuming that the second temporal resolution is 100 sampling lines, the time dimension of the fourth K-space data is 100, and the motion parameters corresponding to the fourth K-space data are 6*100. Constrained optimization processing is performed on the motion parameters corresponding to the fourth K-space data based on a regularization operator along the time dimension. This can be the motion parameters at time t2 minus the motion parameters at time t1, the motion parameters at time t3 minus the motion parameters at time t2, and so on, to obtain a 6*99 motion parameter matrix. The same motion parameters in the 6*99 motion parameter matrix are then added together.

[0106] It should be noted that any of the above implementations of determining the target motion parameters or initial motion parameters based on formula (3) can also use formula (4) to obtain the target motion parameters or initial motion parameters.

[0107] In an embodiment of the present application, based on the second reconstructed image, the fourth K-space data, the preset first regularization strength parameter and the regularization operator along the time dimension, the motion parameters corresponding to the fourth K-space data are constrained optimization processing to obtain the target motion parameters. Because the motion pattern is generally smooth and continuous, the first regularization strength parameter and the regularization operator along the time dimension make the determined target motion parameters more consistent with the smooth and continuous motion pattern.

[0108] In one embodiment, the method further includes: reconstructing the first K-space data to obtain an image to be corrected; performing constrained optimization processing on the image to be corrected based on the first K-space data, target motion parameters, a preset second regularization strength parameter and a regularization operator along the spatial dimension to obtain a target corrected image.

[0109] Based on the above formula (2), updating formula (2) can obtain:

[0110] (5)

[0111] in, is the preset first regularization strength parameter, is the regularization operator along the spatial dimension.

[0112] In this embodiment of the present application, assuming that the dimensions of the image to be corrected are 240*240*240, constrained optimization processing of the image to be corrected based on the regularization operator along the spatial dimension can be performed by subtracting the pixel value of the current coordinate from the pixel value of the next coordinate in one spatial direction, and so on, to obtain an intermediate image of 240*240*239. In the other two spatial directions, intermediate images of 240*239*240 and 239*240*240 are obtained respectively, and then the intermediate images are added together.

[0113] In an embodiment of the present application, a constrained optimization process is performed on the image to be corrected based on the first K-space data, the target motion parameter, a preset second regularization strength parameter, and a regularization operator along the spatial dimension to obtain a target corrected image. In an embodiment of the present application, the constrained optimization process is performed on the image to be corrected using the second regularization strength parameter and a regularization operator along the spatial dimension to make the corrected target corrected image smoother.

[0114] In an exemplary embodiment, Figure 9As shown, image reconstruction is performed on the first K-space data (i.e., the K-space data containing motion in the figure) using a non-uniform fast Fourier transform, resulting in an image containing motion and undersampling artifacts. Image reconstruction is performed using the first K-space data corresponding to 1 / 4 spatial resolution to obtain a first reconstructed image. The first K-space data is divided into 500 sampling lines at a temporal resolution to obtain second K-space data. Motion parameters are estimated based on the second K-space data and the first reconstructed image to obtain initial motion parameters. The first K-space data is corrected using the initial motion parameters to obtain third K-space data. Image reconstruction is performed using the third K-space data corresponding to 1 / 4 spatial resolution to obtain a second reconstructed image. The third K-space data is divided into 100 sampling lines at a temporal resolution to obtain fourth K-space data. Motion parameters are estimated based on the fourth K-space data and the second reconstructed image to determine the target motion parameters of the target site. Image reconstruction is performed based on the target motion parameters and the first K-space data to obtain a corrected target image (i.e., an image free of motion and undersampling artifacts). It can be seen that the motion and undersampling artifacts are eliminated while significantly reducing the number of iterations.

[0115] In an exemplary embodiment, to verify the accuracy and precision of the motion parameter estimation method, a series of simulations were performed. Within the generated k-space data, a range of motion amplitudes were simulated across 2,000 k-space data segments, with a maximum value of 10 mm / 10°. This means that the 20 segments of k-space data contained motion.

[0116] The method provided in this application is used to estimate motion parameters, and the translation estimation error and rotation estimation error of the estimated target motion parameters and the simulated added motion parameters are calculated.

[0117] In order to verify the accuracy of the motion parameter optimization problem, a non-motion image was obtained based on 1 / 4 spatial resolution, and the target motion parameter 1 was obtained based on the non-motion image and the K-space data with motion. The error between the target motion parameter 1 and the simulated motion parameter was determined.

[0118] In order to jointly test the accuracy of the motion parameter optimization problem and the image reconstruction optimization problem, a motion image is obtained based on 1 / 4 spatial resolution, and the target motion parameter 2 is obtained based on the motion image and the k-space data with motion. The error between the target motion parameter 2 and the simulated motion parameter is determined.

[0119] The accuracy and precision of motion parameter estimation are as follows Figure 10 and Figure 11The figure shows the accuracy and precision of translational and rotational motion estimation using either moving images (circles) or non-moving images (diamonds) as input. It can be seen that when the input image is non-moving, accurate estimation is achieved, with errors of 0.181±0.003 mm and 0.277±0.008° for a motion amplitude (translation amplitude / rotation amplitude) of 10 mm and 10°, respectively. However, the error increases slightly when the input image is moving, reaching 0.319±0.004 mm and 0.401±0.026° for a motion amplitude (translation amplitude / rotation amplitude) of 10 mm and 10°, respectively.

[0120] In an exemplary embodiment, Figure 12 As shown, in order to evaluate the effectiveness of the method of the present application in eliminating motion artifacts and accelerating image reconstruction, a multi-dimensional golden section trajectory is used to generate 32-channel three-dimensional radial K-space data with a matrix of 240x240x240. Figure 12 The first row shows the reference image obtained by simulating 58,000 sampling lines, the undersampled image without motion obtained by simulating 12,000 sampling lines in a traditional way (motion-free image, R=7.5), the undersampled image with motion obtained by introducing motion with amplitudes of ±10 mm and ±10˚ (motion-blurred image, R=7.5), and the motion-corrected image obtained by correcting the motion-blurred image. Figure 12 The second row in is the error image between the motion-free image and the reference image, the error image between the motion-blurred image and the reference image, and the error image between the motion-corrected image and the reference image, respectively, obtained based on the traditional method. Figure 13 The third line in FIG is a motion-free image, a motion-blurred image, and a motion-corrected image obtained by the method proposed in this application. Figure 12 The fourth row in the figure shows the error images between the motion-free image and the reference image, the error images between the motion-blurred image and the reference image, and the error images between the motion-corrected image and the reference image, obtained using the method proposed in this application. The normalized root mean square error value is displayed in the lower right corner of each image.

[0121] As can be seen, due to the unique point spread function of the 3D radial sampling trajectory, undersampling results in only minor artifacts, allowing for easy restoration of image quality. However, in the presence of motion, traditional reconstruction methods can lead to severe blurring. Using this reconstruction method, motion artifacts are successfully eliminated, effectively restoring image quality.

[0122] In an exemplary embodiment, Figure 13 As shown, Figure 13This study presents reconstruction results and estimated motion parameters from three motion scans using this method. Three healthy subjects (male, aged 27.2 ± 2.4 years) were scanned using a 32-channel head coil. A spoiled gradient echo sequence with three-dimensional radial full echo acquisition was used, using a multidimensional golden section sampling pattern. Scan parameters were: field of view (FOV) = 220 × 220 × 220 mm³, matrix size = 240 × 240 × 240, flip angle = 17°, repetition time (TR) = 7 ms, echo time (TE) = 2.8 ms, bandwidth = 250 Hz / pixel, number of spokes = 12,000, and total scan time = 84 seconds. Subjects were instructed to perform intermittent or continuous motions commonly seen in clinical scanning (regular-scale intermittent motion, large-scale intermittent motion, and regular-scale continuous motion).

[0123] It can be seen that the intermittent or continuous motion artifacts greatly affect the quality of the reconstructed image. By using this method to reconstruct, the motion and undersampling artifacts are greatly reduced, and the quality of all images is significantly improved.

[0124] It should be understood that, although the steps in the flowcharts of the above embodiments are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts of the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily to be performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0125] Based on the same inventive concept, embodiments of the present application further provide a motion parameter determination device for implementing the aforementioned motion parameter determination method. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more of the following embodiments of the motion parameter determination device can be found in the above-described limitations of the motion parameter determination method and are not further elaborated here.

[0126] In an exemplary embodiment, Figure 14 As shown, a motion parameter determination device is provided, comprising: an acquisition module 11, a first reconstruction module 12 and a parameter estimation module 13, wherein:

[0127] An acquisition module 11 is configured to acquire first K-space data obtained by sampling a target area; the first K-space data includes sampling data passing through the center of the K-space;

[0128] a first reconstruction module 12 configured to reconstruct first preset K-space data to obtain a first reconstructed image; the first preset K-space data comprising first K-space data whose distance from the K-space center is less than or equal to a preset distance, and whose sampling frequency is less than a preset sampling frequency;

[0129] The parameter estimation module 13 is configured to perform motion parameter estimation based on the first K-space data and the first reconstructed image to determine target motion parameters of the target part.

[0130] In one embodiment, the parameter estimation module 13 is specifically used to divide the first K space data based on the first time resolution to obtain each second K space data; for each second K space data, perform motion parameter estimation based on the second K space data and the first reconstructed image to determine the target motion parameter of the target part.

[0131] In one embodiment, the parameter estimation module 13 is specifically used to perform motion parameter estimation based on the second K-space data and the first reconstructed image to obtain initial motion parameters corresponding to the second K-space data; correct the first K-space data according to each initial motion parameter to obtain third K-space data; and perform motion parameter estimation based on the third K-space data to determine the target motion parameters of the target part.

[0132] In one embodiment, the parameter estimation module 13 is specifically used to perform image reconstruction on the second preset K space data in the third K space data to obtain a second reconstructed image; the second preset K space data includes third K space data whose distance from the K space center is less than or equal to the preset distance; the third K space data is divided based on the second time resolution to obtain fourth K space data; the second time resolution is higher than the first time resolution; for each fourth K space data, motion parameter estimation is performed based on the fourth K space data and the second reconstructed image to determine the target motion parameter of the target part.

[0133] In one embodiment, the parameter estimation module 13 is specifically used to perform constrained optimization processing on the motion parameters corresponding to the fourth K-space data based on the second reconstructed image, the fourth K-space data, a preset first regularization strength parameter and a regularization operator along the time dimension to obtain the target motion parameters.

[0134] In one embodiment, the motion parameter determination apparatus further comprises:

[0135] A second reconstruction module, configured to reconstruct the first K-space data to obtain an image to be corrected;

[0136] The constrained optimization module is used to perform constrained optimization processing on the image to be corrected based on the first K-space data, the target motion parameter, the preset second regularization strength parameter and the regularization operator along the spatial dimension to obtain the target corrected image.

[0137] Each module in the aforementioned motion parameter determination apparatus may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0138] In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps of any of the above method embodiments when executing the computer program.

[0139] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above method embodiments are implemented.

[0140] In one embodiment, a computer program product is provided, comprising a computer program, which implements the steps of any of the above method embodiments when executed by a processor.

[0141] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0142] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), quantum computing-based data processing logic devices, artificial intelligence (AI) processors, and the like.

[0143] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0144] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for determining motion parameters, characterized in that: The method comprises: Acquire first K-space data obtained by sampling the target part; the first K-space data includes sampling data passing through the center of the K-space; the sampling density of the center of the K-space is greater than a preset sampling density; performing image reconstruction on first preset K-space data in the first K-space data to obtain a first reconstructed image; the first preset K-space data includes first K-space data whose distance from the center of the K-space is less than or equal to a preset distance, and whose sampling frequency is less than a preset sampling frequency; Motion parameter estimation is performed based on the first K-space data and the first reconstructed image to determine target motion parameters of the target part.

2. The method according to claim 1, characterized in that The performing motion parameter estimation based on the first K-space data and the first reconstructed image to determine the target motion parameter of the target part includes: Dividing the first K-space data based on a first time resolution to obtain second K-space data; For each piece of the second K-space data, motion parameter estimation is performed according to the second K-space data and the first reconstructed image to determine the target motion parameter of the target part.

3. The method according to claim 2, characterized in that The performing motion parameter estimation based on the second K-space data and the first reconstructed image to determine the target motion parameter of the target part includes: performing motion parameter estimation based on the second K-space data and the first reconstructed image to obtain initial motion parameters corresponding to the second K-space data; Correcting the first K-space data according to the initial motion parameters to obtain third K-space data; Motion parameter estimation is performed based on the third K-space data to determine target motion parameters of the target part.

4. The method according to claim 3, characterized in that The performing motion parameter estimation based on the third K-space data to determine the target motion parameter of the target part includes: performing image reconstruction on second preset K-space data in the third K-space data to obtain a second reconstructed image, wherein the second preset K-space data includes third K-space data whose distance from the K-space center is less than or equal to the preset distance; Dividing the third K-space data based on a second time resolution to obtain fourth K-space data; wherein the second time resolution is higher than the first time resolution; For each of the fourth K-space data, motion parameter estimation is performed according to the fourth K-space data and the second reconstructed image to determine the target motion parameter of the target part.

5. The method according to claim 4, characterized in that The performing motion parameter estimation based on the fourth K-space data and the second reconstructed image to determine the target motion parameter of the target part includes: Based on the second reconstructed image, the fourth K-space data, a preset first regularization strength parameter and a regularization operator along the time dimension, constrained optimization processing is performed on the motion parameters corresponding to the fourth K-space data to obtain the target motion parameters.

6. The method according to any one of claims 1 to 5, characterized in that The method further comprises: Reconstructing the first K-space data to obtain an image to be corrected; Based on the first K-space data, the target motion parameter, a preset second regularization strength parameter and a regularization operator along the spatial dimension, constrained optimization processing is performed on the image to be corrected to obtain a target corrected image.

7. A motion parameter determination device, characterized in that: The device comprises: An acquisition module is configured to acquire first K-space data obtained by sampling a target part; the first K-space data includes sampling data passing through a center of the K-space; and a sampling density of the center of the K-space is greater than a preset sampling density; a first reconstruction module, configured to perform image reconstruction on first preset K-space data in the first K-space data to obtain a first reconstructed image; the first preset K-space data includes first K-space data whose distance from the center of the K-space is less than or equal to a preset distance, and whose sampling frequency is less than a preset sampling frequency; The parameter estimation module is used to perform motion parameter estimation based on the first K-space data and the first reconstructed image to determine the target motion parameter of the target part.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images