CN112489155A - Image reconstruction method and apparatus, electronic device, and machine-readable storage medium - Google Patents
Image reconstruction method and apparatus, electronic device, and machine-readable storage medium Download PDFInfo
- Publication number
- CN112489155A CN112489155A CN202011440541.8A CN202011440541A CN112489155A CN 112489155 A CN112489155 A CN 112489155A CN 202011440541 A CN202011440541 A CN 202011440541A CN 112489155 A CN112489155 A CN 112489155A
- Authority
- CN
- China
- Prior art keywords
- data
- image
- image data
- target image
- residual error
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 172
- 230000008569 process Effects 0.000 claims abstract description 126
- 238000012937 correction Methods 0.000 claims abstract description 48
- 238000001914 filtration Methods 0.000 claims abstract description 45
- 238000012545 processing Methods 0.000 claims abstract description 14
- 239000011159 matrix material Substances 0.000 claims description 12
- 239000000126 substance Substances 0.000 claims description 12
- 238000005259 measurement Methods 0.000 claims description 2
- 241001643392 Cyclea Species 0.000 claims 1
- 238000003384 imaging method Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000002595 magnetic resonance imaging Methods 0.000 description 6
- 238000005457 optimization Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 230000001629 suppression Effects 0.000 description 5
- 239000002131 composite material Substances 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 239000004973 liquid crystal related substance Substances 0.000 description 3
- 230000001133 acceleration Effects 0.000 description 2
- 238000009795 derivation Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000011084 recovery Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 210000003484 anatomy Anatomy 0.000 description 1
- 230000000747 cardiac effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an image reconstruction method, which comprises the following steps: executing the following loop process until the loop ending condition is met: based on a target image structure filter, carrying out filtering processing on image data and data residual of a target image in the previous cycle process to obtain first image data in the current cycle process; based on the guide image structure filter, carrying out filtering processing on target image data and data residual in the previous cycle process to obtain second image data in the current cycle process; obtaining image data of the target image in the circulation process according to the first image data and the second image data; obtaining a data residual error in the circulation process according to the undersampled data of the K space, the image data of the target image in the circulation process and the residual error correction factor; and when the loop ending condition is not met, the image data and the data residual error of the target image in the current loop process are filled as the image data and the data residual error in the next loop process.
Description
Technical Field
The present invention relates to the field of image processing technologies, and in particular, to an image reconstruction method, an image reconstruction device, an electronic apparatus, and a machine-readable storage medium.
Background
As an important diagnostic imaging technique, the major drawback of slow imaging speed of mri is that it limits further development in advanced clinical application fields including multi-contrast imaging, dynamic cardiac cine imaging, etc. Therefore, how to shorten the imaging time while maintaining high resolution of the image has been a difficult problem in the field of magnetic resonance.
Compressed sensing is used as a classic acceleration magnetic resonance imaging method based on a signal undersampling mechanism, under the condition that signal sparsity is met and a sampling matrix is irrelevant to a sparse transformation basis, data far below a Nyquist (Nyquist) sampling quantity is used, and good recovery of an original image is obtained through a nonlinear reconstruction algorithm. The standard compressed sensing imaging method is to solve an L1The method of norm regularization, also known as LASSO (LASSO) regression. The research on optimization algorithms for LASSO-like problems mainly focuses on first-order optimization algorithms with lower computation cost and faster convergence. For example, having o (1/k) after introducing the Nesterov acceleration strategy2) A fast iterative soft threshold algorithm of convergence rate; an alternating direction multiplier method in conjunction with variables to solve large scale problems can be separated. Such algorithms have good application to weak saliency, including even some non-saliency problems.
Recently, Plug-and-Play-Prior (PPP) -based iterative thresholding algorithms replace the original threshold shrinkage with an image denoising filter (filter Prior is Plug-and-Play Prior), i.e. couple image denoising to a forward model-based image restoration framework. The plug-and-play apriori based iterative threshold algorithm shows superior imaging quality compared with the traditional optimization algorithm.
However, the conventional plug-and-play-prior-based iterative threshold algorithm is only based on a single type of plug-and-play-prior iterative threshold algorithm, and does not consider more types of plug-and-play-prior, so that a reconstructed image which is well balanced in artifact suppression and structure protection cannot be obtained, and the image reconstruction performance cannot be improved.
Disclosure of Invention
In order to solve the technical problems in the prior art, the present invention provides an image reconstruction method and an image reconstruction apparatus that combine more types of plug-and-play priors.
An image reconstruction method provided according to an aspect of an embodiment of the present invention includes: executing the following loop process until the loop ending condition is met:
based on a target image structure filter, carrying out filtering processing on image data and data residual of a target image obtained in the previous circulation process to obtain first image data in the circulation process; based on the guide image structure filter, carrying out filtering processing on the target image data and the data residual error obtained in the previous cycle process to obtain second image data in the current cycle process; obtaining image data of a target image in the circulation process according to the first image data and the second image data in the circulation process; obtaining a data residual error in the circulation process according to the undersampled data of the K space, the image data of the target image in the circulation process, the target image data and the data residual error obtained in the previous circulation process;
when the cycle ending condition is not met, the image data and the data residual error of the target image obtained in the current cycle process are filled into the image data and the data residual error of the target image in the next cycle process.
In an example of the image reconstructing method provided in the above aspect, the obtaining a data residual in the current cycle according to the undersampled data of the K space, the image data of the target image in the current cycle, and the target image data and the data residual obtained in the previous cycle includes: obtaining a first residual error correction factor in the circulation process through divergence estimation of target image structure filtering according to image data and data residual errors of the target image obtained in the previous circulation process; obtaining a second residual error correction factor in the cycle process by guiding divergence estimation of image structure filtering according to the image data and the data residual error of the target image obtained in the previous cycle process; obtaining a residual error correction factor in the circulation process according to the first residual error correction factor and the second residual error correction factor in the circulation process; and obtaining a data residual error in the circulation process according to the undersampled data of the K space, the image data of the target image in the circulation process and the residual error correction factor.
In one example of the image reconstruction method provided in the above-described aspect, the image data and the data residual of the target image obtained in the previous cycle are subjected to a filtering process based on the target image structure filter according to the following equation 1 to obtain the first image data in the present cycle,
[1]
wherein the content of the first and second substances,
representing said first image data during the present cycle,
representing a filter operator, x, based on a target image structure filtert-1Image data representing the target image obtained during the previous cycle, zt-1Obtaining a data residual error in the last cycle process, wherein A represents an undersampled Fourier transform matrix;
and/or filtering the target image data and the data residual error obtained in the previous cycle process based on a guide image structure filter according to the following formula 2 to obtain second image data in the current cycle process;
[2]
wherein the content of the first and second substances,
representing said second image data during the present cycle,
representing a filter operator, x, based on a guided image structure filtert-1Image data representing the target image obtained during the previous cycle, zt-1And D, obtaining a data residual error in the last cycle, wherein A represents an undersampled Fourier transform matrix.
In one example of the image reconstruction method provided in the above-described aspect, the image data of the target image in the present cycle is calculated from the first image data and the second image data in the present cycle using the following equation 3,
[3]
wherein α represents a weight parameter, and 0<α<1，xtImage data representing the target image during the present cycle.
In one example of the image reconstruction method provided in the above-described aspect, the first residual correction factor in the present cycle is obtained from the image data and the data residual of the target image obtained in the previous cycle, and using equation 4 below,
[4]
wherein the content of the first and second substances,
a first residual correction factor representing the target image structure filter-based prior in the present loop, delta a metric constant representing the degree of problem underrun,
representing divergence operator, x, corresponding to a filter operator based on a target image structure filtert-1Image data representing the target image obtained during the previous cycle, zt-1Obtaining a data residual error in the last cycle process, wherein A represents an undersampled Fourier transform matrix;
and/or, obtaining a second residual error correction factor based on the guiding image structure filter prior in the cycle process by using the following formula 5 according to the image data and the data residual error of the target image obtained in the previous cycle process,
[5]
wherein the content of the first and second substances,
a second residual correction factor representing the guided image structure filter-based filtering prior in the present loop, delta a metric constant representing the degree of underdetermined problem,
representing divergence operators, x, corresponding to filtering operators based on guided image structure filterst-1Image data representing the target image obtained during the previous cycle, zt-1And D, obtaining a data residual error in the last cycle, wherein A represents an undersampled Fourier transform matrix.
In one example of the image reconstructing method provided in the above aspect, the residual error correction factor in the present cycle is calculated according to the first residual error correction factor and the second residual error correction factor in the present cycle by using the following equation 6,
[6]
wherein o istRepresents the residual error correction factor in the present cycle, alpha represents the weight parameter, and 0<α< 1。
In an example of the image reconstruction method provided in the above aspect, a data residual in the current cycle is calculated according to the undersampled data of the K space, the image data of the target image in the current cycle, and the residual correction factor, and using the following equation 7,
[7]zt＝b-Axt+ot
wherein z istAnd b represents undersampled data of a K space.
According to another aspect of embodiments of the present invention, there is provided an image reconstruction apparatus including: the system comprises a target image structure filter, a guide image structure filter, a target image data acquisition module and a data residual error acquisition module which are operated circularly until a circulation ending condition is met;
the target image structure filter is used for filtering the image data and the data residual of the target image obtained in the previous cycle process to obtain first image data in the current cycle process; the guide image structure filter is used for filtering the target image data and the data residual error obtained in the previous cycle process to obtain second image data in the current cycle process; the target image data acquisition module is used for acquiring the image data of the target image in the circulation process according to the first image data and the second image data in the circulation process; the data residual error acquisition module is used for acquiring a data residual error in the circulation process according to the undersampled data of the K space, the image data of the target image in the circulation process, the target image data and the data residual error acquired in the previous circulation process;
when the cycle ending condition is not met, the image data and the data residual error of the target image obtained in the current cycle process are filled into the image data and the data residual error of the target image in the next cycle process.
According to still another aspect of an embodiment of the present invention, there is provided an electronic apparatus including: at least one processor, and a memory coupled with the at least one processor, the memory storing instructions that, when executed by the at least one processor, cause the at least one processor to perform the image reconstruction method as described above.
According to a further aspect of embodiments of the present invention there is provided a machine-readable storage medium having stored thereon executable instructions that, when executed, cause the machine to perform the image reconstruction method as described above.
Has the advantages that: the invention adopts composite plug-and-play prior, can help to obtain images with good balance in the aspects of artifact suppression and structure protection, and can also improve the image reconstruction performance.
Drawings
The above and other aspects, features and advantages of embodiments of the present invention will become more apparent from the following description taken in conjunction with the accompanying drawings, in which:
FIG. 1 is a flow chart illustrating an image reconstruction method according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating an exemplary method for obtaining data residuals during the current cycle in the image reconstruction method according to the embodiment of the present invention;
FIG. 3 is a block diagram illustrating an image reconstruction apparatus according to an embodiment of the present invention;
fig. 4 is a block diagram illustrating an electronic device implementing an image reconstruction method according to an embodiment of the present invention.
Detailed Description
Hereinafter, specific embodiments of the present invention will be described in detail with reference to the accompanying drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the specific embodiments set forth herein. Rather, these embodiments are provided to explain the principles of the invention and its practical application to thereby enable others skilled in the art to understand the invention for various embodiments and with various modifications as are suited to the particular use contemplated.
As used herein, the term "include" and its variants mean open-ended terms in the sense of "including, but not limited to. The terms "based on," based on, "and the like mean" based at least in part on, "" based at least in part on. The terms "one embodiment" and "an embodiment" mean "at least one embodiment". The term "another embodiment" means "at least one other embodiment". The terms "first," "second," and the like may refer to different or the same object. Other definitions, whether explicit or implicit, may be included below. The definition of a term is consistent throughout the specification unless the context clearly dictates otherwise.
As described above, the plug-and-play apriori based iterative threshold algorithm is only an iterative threshold algorithm based on a single type of plug-and-play apriori, and does not consider more types of plug-and-play apriori. In this case, a reconstructed image having a good balance between artifact suppression and structural protection cannot be obtained, and thus image reconstruction performance cannot be improved.
In order to obtain a reconstructed image with a good balance between artifact suppression and structure protection, and thus improve image reconstruction performance, an embodiment of the present invention provides an image reconstruction method and an image reconstruction apparatus that combine more types of plug-and-play priors to perform image reconstruction. The image reconstruction method may be performed by an electronic device that performs the following loop process until a loop end condition is satisfied: based on a target image structure filter, carrying out filtering processing on image data and data residual of a target image obtained in the previous circulation process to obtain first image data in the circulation process; based on the guide image structure filter, carrying out filtering processing on the target image data and the data residual error obtained in the previous cycle process to obtain second image data in the current cycle process; obtaining image data of a target image in the circulation process according to the first image data and the second image data in the circulation process; obtaining a data residual error in the circulation process according to the undersampled data of the K space, the image data of the target image in the circulation process, the target image data and the data residual error obtained in the previous circulation process; when the cycle ending condition is not met, the image data and the data residual error of the target image obtained in the current cycle process are filled into the image data and the data residual error of the target image in the next cycle process.
Therefore, in the image reconstruction method, the composite plug-and-play prior of the target image structure filter and the guide image structure filter is adopted to filter the image, so that the image which is well balanced in the aspects of artifact suppression and structure protection can be obtained, and the image reconstruction performance can be improved.
In the description of the image reconstruction method according to the embodiment of the present invention, magnetic resonance imaging is exemplified. Next, some basic concepts and process derivations applied in the image reconstruction method according to the embodiment of the present invention will be explained by taking magnetic resonance imaging as an example.
The task of image reconstruction for magnetic resonance accelerated imaging can be considered as solving for an L1Norm canonical LASSO regression problem whose solving cost function includes a data fit term based on a noise Gaussian distribution hypothesis and L on image in transform domain coefficients1And (5) norm constraint.
Firstly, a PPP (plug and play prior) -AMP (Approximate Message transfer) algorithm framework for solving the optimization problem is constructed, and two types of plug and play prior are designed and coupled in an important way: one is a structural filter prior from the target image itself and the other is a guided image structural filter prior from other modalities or parameters. The two types of plug-and-play priors are weighted as a composite plug-and-play prior and coupled to a forward model framework of the AMP algorithm.
Specifically, the magnetic resonance imaging data acquisition process based on the K-space undersampling mechanism can be discretely expressed as the following equation 1.
[ equation 1] b ═ Ax + ξ
Here, the first and second liquid crystal display panels are,
for the image data of the target image to be reconstructed,
for the under-sampled data of the K-space,
in order to under-sample the fourier transform matrix,
assuming a gaussian-distributed noise.
The solution of the underdetermined problem is difficult, and under the framework of a compressed sensing theory, the L of the target image in a certain sparse transform domain coefficient is based on1Norm regularization may consider the near perfect recovery (or reconstruction) of the original target image x from the K-space undersampled data b. For the solution, an unconstrained optimization model is constructed, which is expressed as equation 2 below.
[ formula 2]
Here, λ is a positive constant of the balanced data fitting and the sparseness regularization. R (-) represents some sparse domain of variation.
A series of Iterative algorithms with lower computational cost are proposed for solving the convex optimization problem of equation 2, where a representative Algorithm is Iterative Soft-threshold Algorithm (ISTA), which can be expressed as equation 3 and equation 4 below.
[ formula 3 ]]xt＝Sτ(AHzt-1+xt-1)
[ formula 4]zt＝b-Axt
Here, a soft threshold algorithm S is iteratedτ(y)＝(|y|-τ)+sign(y),xtDenotes the t-th x-estimate (i.e. the image data of the reconstructed target image after the t-th iteration), and τ denotes the threshold parameter. Approximate message passing algorithm as a variation of ISTA, its core is to introduce Onsager correction term to correct residual ztAnd gaussianization is carried out, so that the algorithm performance is further improved. Therefore, the approximate message passing algorithm may be expressed as equation 5, equation 6, and equation 7 below.
[ formula 5]xt＝Sτ(AHzt-1+xt-1)
[ formula 6 ]]
[ formula 7]zt＝b-Axt+ot
Here, the first and second liquid crystal display panels are,
for Onsager correction term, SτIs' SτThe derivative of (d), delta is a metric constant to the degree of underdetermination of the problem,<·>representing a vector mean operation.
Then, a generalized denoising operator (i.e., a filter operator) is introduced to construct a PPP-AMP algorithm framework, and the constructed PPP-AMP algorithm can be expressed as the following equation 8, equation 9, and equation 10.
[ formula 8]
[ formula 9 ]]
[ formula 10 ]]zt＝b-Axt+ot
Here, the first and second liquid crystal display panels are,
in order to de-noise the operator,
and the divergence operator is corresponding to the denoising operator.
In the embodiment of the invention, two different types of denoising priors are coupled to construct an approximate message transfer algorithm with improved performance and based on a composite plug-and-play prior to reconstructing the image.
In one example, a first class of plug and play priors employs a Block Matching and 3D Filtering (BM 3D) based algorithm as the self-structure Filtering priors from the target image x, wherein the BM3D algorithm aims at constructing a three-dimensional matrix by metric Matching of euclidean distances with neighboring image blocks, Filtering in three-dimensional space as a whole, and then inverse transforming the Filtering result to a two-dimensional image.
In one example, a second class of plug-and-play priors employs a mutual Guided Image Filtering (muGIF) algorithm as the guide Image x from other modalities or parametersrThe muGIF aims to obtain the shared anatomical structure information between the target image and the guide image through the interactivity measurement of the guide image structure information similar to the target image, and strengthens the main structure characteristics of the target image while effectively inhibiting image artifacts. The interactive guide filtering realizes fully introducing guide image structure information on one hand, and avoids detail filtering deviation caused by image content with difference on the other hand.
The above is a detailed description of some basic concepts and process derivations applied to the image reconstruction method according to embodiments of the present invention, using magnetic resonance imaging as an example.
Next, an image reconstruction method and an image reconstruction apparatus that perform image reconstruction in combination with a hybrid plug-and-play prior according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.
The image reconstruction method according to an embodiment of the present invention may be performed by an electronic device, which may include a smartphone, a tablet computer, a personal computer, a cloud server, a server, and the like.
Fig. 1 is a flowchart illustrating an image reconstruction method according to an embodiment of the present invention.
Referring to fig. 1, in step S110, the image data and the data residual of the target image obtained in the previous cycle are filtered based on the target image structure filter to obtain the first image data in the present cycle.
In one example, the target image structure filter has the block matching based 3D filtering algorithm described above, which may enable filtering priors for the own structure from the target image.
In one example, the image data and the data residual of the target image obtained in the previous cycle are subjected to a filtering process based on the target image structure filter and according to equation 11 below to obtain the first image data in the present cycle.
[ formula 11]
Wherein the content of the first and second substances,
representing said first image data in the present loop process (the t-th iteration),
representing a filter operator, x, based on a target image structure filtert-1Image data representing the target image obtained in the previous cycle (t-1 st iteration)t-1And D, obtaining a data residual error in the last cycle, wherein A represents an undersampled Fourier transform matrix.
In step S120, the target image data and the data residual obtained in the previous cycle are filtered based on the guide image structure filter to obtain the second image data in the current cycle.
In one example, the guide image structure filter has the interactive guide image filtering algorithm described above, which can implement a structural filtering prior to guide images from other modalities or parameters. In one example, the guide image is similar to the target image.
In one example, the target image data and the data residual obtained during the previous cycle are subjected to a filtering process based on a guide image structure filter and according to equation 12 below to obtain second image data during the present cycle.
[ formula 12 ]]
Wherein the content of the first and second substances,
representing said second image data during the present cycle,
representing a filter operator based on a guided image structure filter.
In step S130, image data of the target image in the present loop is obtained from the first image data and the second image data in the present loop.
In one example, the image data of the target image in the present loop is calculated from the first image data and the second image data in the present loop by using the following equation 13.
[ formula 13 ]]
Wherein α represents a weight parameter, and 0<α<1，xtImage data representing the target image during the present cycle.
In step S140, a data residual in the current cycle is obtained according to the undersampled data of the K space, the image data of the target image in the current cycle, and the target image data and the data residual obtained in the previous cycle.
In one example, the undersampled data of K-space may be, for example, data acquired by undersampling of a magnetic resonance imaging apparatus.
Fig. 2 is a flowchart illustrating an exemplary method for acquiring data residuals during the current cycle in the image reconstruction method according to the embodiment of the present invention.
Referring to fig. 2, in step S141, a first residual error correction factor in the present loop is obtained from the image data and the data residual of the target image obtained in the previous loop and through divergence estimation of the target image structure filter.
In one example, the first residual correction factor in the present loop is obtained from the image data and data residual of the target image obtained in the previous loop, and using equation 14 below.
[ formula 14]
Wherein the content of the first and second substances,
a first residual correction factor representing the target image structure filter-based prior in the present loop, delta a metric constant representing the degree of problem underrun,
and representing divergence operators corresponding to the filtering operators of the filter based on the target image structure.
In step S142, a second residual error correction factor in the present loop is obtained by divergence estimation of the guided image structure filtering according to the image data and the data residual error of the target image obtained in the previous loop.
In one example, the second residual correction factor in the present loop is obtained from the image data and data residual of the target image obtained in the previous loop, and using equation 15 below.
[ formula 15]
Wherein the content of the first and second substances,
representing a second residual correction factor during this loop that is a priori based on the guided image structure filter,
representing divergence operators corresponding to the filter operators based on the guided image structure filter.
In step S143, a residual error correction factor in the present cycle is obtained according to the first residual error correction factor and the second residual error correction factor in the present cycle.
In one example, the residual error correction factor in the present cycle is calculated according to the first residual error correction factor and the second residual error correction factor in the present cycle by using the following equation 16.
[ formula 16]
Wherein o istRepresenting the residual correction factor during this cycle.
In step S144, a data residual in the current cycle is obtained according to the undersampled data of the K space, the image data of the target image in the current cycle, and the residual correction factor.
In one example, the data residual in the current cycle is calculated according to the undersampled data of the K space, the image data of the target image in the current cycle, and the residual correction factor, and by using the following equation 17.
[ formula 17]zt＝b-Axt+ot
Wherein z istAnd b represents the undersampled data of the K space.
With continued reference to fig. 1, in step S150, it is determined whether a loop end condition is satisfied. If yes, finishing reconstruction; if not, the image data and the data residual of the target image obtained in the current cycle process are filled as the image data and the data residual of the target image in the next cycle process, and the process goes to step S110.
Here, the loop end condition may be specified. In one example, the loop-ending condition may include reaching a predetermined number of loops (or iterations).
Fig. 3 is a block diagram illustrating an image reconstruction apparatus according to an embodiment of the present invention.
The image reconstruction apparatus 300 is applied to an electronic device to be executed by the electronic device. Referring to fig. 3, the image reconstruction apparatus 300 includes: a target image structure filter 310, a guide image structure filter 320, a target image data acquisition module 330, and a data residual acquisition module 340. The target image structure filter 310, the guide image structure filter 320, the target image data acquisition module 330, and the data residual acquisition module 340 operate in a loop until a loop end condition is satisfied. Wherein the loop-over condition may be specified. In one example, the loop-ending condition may include reaching a predetermined number of loops (or iterations).
The target image structure filter 310 is configured to perform filtering processing on the image data and the data residual of the target image obtained in the previous cycle to obtain the first image data in the present cycle. In one example, the target image structure filter 310 may be configured to perform a filtering process on the target image data and the data residual obtained in the previous cycle according to equation 11 above to obtain the first image data in the present cycle.
The guide image structure filter 320 is configured to perform filtering processing on the target image data and the data residual obtained in the previous cycle to obtain second image data in the present cycle. In one example, the guide image structure filter 320 may be configured to perform a filtering process on the target image data and the data residual obtained in the previous cycle according to equation 12 above to obtain the second image data in the present cycle.
The target image data acquisition module 330 is configured to obtain image data of a target image during the present cycle from the first image data and the second image data during the present cycle. In one example, the target image data acquisition module 330 may be configured to calculate the image data of the target image in the present cycle from the first image data and the second image data in the present cycle by using the above equation 13.
The data residual obtaining module 340 is configured to obtain a data residual in the current cycle process according to the undersampled data of the K space, the image data of the target image in the current cycle process, and the target image data and the data residual obtained in the previous cycle process. In one example, the data residual obtaining module 340 may be configured to calculate the data residual in the current cycle according to the undersampled data of the K space, the image data of the target image in the current cycle, and the target image data and the data residual obtained in the previous cycle, and by using the above equation 15, equation 16, and equation 17.
The image reconstruction method and the image reconstruction apparatus according to the embodiments of the present invention are described above with reference to fig. 1 to 3.
The image reconstruction apparatus according to the embodiment of the present invention may be implemented by hardware, or may be implemented by software, or a combination of hardware and software. The software implementation is taken as an example, and is formed by reading corresponding computer program instructions in the storage into the memory for operation through the processor of the device where the software implementation is located as a logical means. In an embodiment of the present invention, the use of the apparatus for image reconstruction may be implemented, for example, with an electronic device.
Fig. 4 is a block diagram illustrating an electronic device implementing an image reconstruction method according to an embodiment of the present invention.
Referring to fig. 4, the electronic device 400 may include at least one processor 410, a storage (e.g., a non-volatile storage) 420, a memory 430, and a communication interface 440, and the at least one processor 410, the storage 420, the memory 430, and the communication interface 440 are connected together via a bus 450. The at least one processor 410 executes at least one computer-readable instruction (i.e., the elements described above as being implemented in software) stored or encoded in memory.
In one example, computer-executable instructions are stored in the memory that, when executed, cause the at least one processor 410 to perform the following loop process until a loop-ending condition is satisfied: based on a target image structure filter, carrying out filtering processing on image data and data residual of a target image obtained in the previous circulation process to obtain first image data in the circulation process; based on the guide image structure filter, carrying out filtering processing on the target image data and the data residual error obtained in the previous cycle process to obtain second image data in the current cycle process; obtaining image data of a target image in the circulation process according to the first image data and the second image data in the circulation process; obtaining a data residual error in the circulation process according to the undersampled data of the K space, the image data of the target image in the circulation process, the target image data and the data residual error obtained in the previous circulation process; when the cycle ending condition is not met, the image data and the data residual error of the target image obtained in the current cycle process are filled into the image data and the data residual error of the target image in the next cycle process.
It should be appreciated that the computer-executable instructions stored in the memory, when executed, cause the at least one processor 410 to perform various operations and functions described in conjunction with fig. 1-3 above in various embodiments in accordance with the present invention.
According to one embodiment, a program product, such as a machine-readable medium, is provided. A machine-readable medium may have instructions (i.e., elements described above as being implemented in software) that, when executed by a machine, cause the machine to perform various operations and functions described in conjunction with fig. 1-3 above in various embodiments of the invention.
Specifically, a system or apparatus may be provided which is provided with a readable storage medium on which software program code implementing the functions of any of the above embodiments is stored, and causes a computer or processor of the system or apparatus to read out and execute instructions stored in the readable storage medium.
In this case, the program code itself read from the readable medium can realize the functions of any of the above-described embodiments, and thus the machine-readable code and the readable storage medium storing the machine-readable code constitute a part of the embodiments of the present invention.
Examples of the readable storage medium include floppy disks, hard disks, magneto-optical disks, optical disks (e.g., CD-ROMs, C D-R, CD-RWs, DVD-ROMs, DVD-RAMs, DVD-RWs), magnetic tapes, nonvolatile memory cards, and ROMs. Alternatively, the program code may be downloaded from a server computer or from the cloud via a communications network.
The foregoing description has described certain embodiments of this invention. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
Not all steps and elements in the above flows and system structure diagrams are necessary, and some steps or elements may be omitted according to actual needs. The execution order of the steps is not fixed, and can be determined as required. The apparatus structures described in the above embodiments may be physical structures or logical structures, that is, some units may be implemented by the same physical entity, or some units may be implemented by a plurality of physical entities, or some units may be implemented by some components in a plurality of independent devices.
The terms "exemplary," "example," and the like, as used throughout this specification, mean "serving as an example, instance, or illustration," and do not mean "preferred" or "advantageous" over other embodiments. The detailed description includes specific details for the purpose of providing an understanding of the described technology. However, the techniques may be practiced without these specific details. In some instances, well-known structures and devices are shown in block diagram form in order to avoid obscuring the concepts of the described embodiments.
Alternative embodiments of the present invention are described in detail with reference to the drawings, however, the embodiments of the present invention are not limited to the specific details in the above embodiments, and within the technical idea of the embodiments of the present invention, many simple modifications may be made to the technical solution of the embodiments of the present invention, and these simple modifications all belong to the protection scope of the embodiments of the present invention.
The previous description of the disclosure is provided to enable any person skilled in the art to make or use the disclosure. Various modifications to the disclosure will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other variations without departing from the scope of the disclosure. Thus, the description is not intended to be limited to the examples and designs described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. An image reconstruction method, characterized in that the image reconstruction method comprises:
executing the following loop process until the loop ending condition is met:
based on a target image structure filter, carrying out filtering processing on image data and data residual of a target image obtained in the previous circulation process to obtain first image data in the circulation process;
based on the guide image structure filter, carrying out filtering processing on the target image data and the data residual error obtained in the previous cycle process to obtain second image data in the current cycle process;
obtaining image data of a target image in the circulation process according to the first image data and the second image data in the circulation process;
obtaining a data residual error in the circulation process according to the undersampled data of the K space, the image data of the target image in the circulation process, the target image data and the data residual error obtained in the previous circulation process;
when the cycle ending condition is not met, the image data and the data residual error of the target image obtained in the current cycle process are filled into the image data and the data residual error of the target image in the next cycle process.
2. The image reconstruction method according to claim 1, wherein obtaining a data residual in the current cycle based on the undersampled data of the K space, the image data of the target image in the current cycle, and the target image data and the data residual obtained in the previous cycle comprises:
obtaining a first residual error correction factor in the circulation process through divergence estimation of target image structure filtering according to image data and data residual errors of the target image obtained in the previous circulation process;
obtaining a second residual error correction factor in the cycle process by guiding divergence estimation of image structure filtering according to the image data and the data residual error of the target image obtained in the previous cycle process;
obtaining a residual error correction factor in the circulation process according to the first residual error correction factor and the second residual error correction factor in the circulation process;
and obtaining a data residual error in the circulation process according to the undersampled data of the K space, the image data of the target image in the circulation process and the residual error correction factor.
3. The image reconstruction method according to claim 1 or 2, wherein the image data and the data residual of the target image obtained in the previous cycle are subjected to a filtering process based on the target image structure filter according to the following equation 1 to obtain the first image data in the present cycle,
[1]
wherein the content of the first and second substances,
representing said first image data during the present cycle,
representing a filter operator, x, based on a target image structure filtert-1Image data representing the target image obtained during the previous cycle, zt-1Obtaining a data residual error in the last cycle process, wherein A represents an undersampled Fourier transform matrix;
and/or filtering the target image data and the data residual error obtained in the previous cycle process based on a guide image structure filter according to the following formula 2 to obtain second image data in the current cycle process;
[2]
wherein the content of the first and second substances,
representing said second image data during the present cycle,
representing a filter operator, x, based on a guided image structure filtert-1Image data representing the target image obtained during the previous cycle, zt-1And D, obtaining a data residual error in the last cycle, wherein A represents an undersampled Fourier transform matrix.
4. The image reconstruction method according to claim 3, wherein the image data of the target image in the present cycle is calculated from the first image data and the second image data in the present cycle using the following equation 3,
[3]
wherein α represents a weight parameter, and 0<α<1，xtImage data representing the target image during the present cycle.
5. The image reconstructing method according to claim 2, wherein the first residual error correction factor in the present cycle is obtained from the image data and the data residual error of the target image obtained in the previous cycle and using the following equation 4,
[4]
wherein the content of the first and second substances,
a first residual correction factor representing the target image structure filter-based prior in the present loop, delta a metric constant representing the degree of problem underrun,
representing divergence operator, x, corresponding to a filter operator based on a target image structure filtert-1Image data representing the target image obtained during the previous cycle, zt-1Obtaining a data residual error in the last cycle process, wherein A represents an undersampled Fourier transform matrix;
and/or, obtaining a second residual error correction factor in the cycle process by using the following formula 5 according to the image data and the data residual error of the target image obtained in the previous cycle process,
[5]
wherein the content of the first and second substances,
represents the present cycleA second residual correction factor in the process, a priori based on the guided image structure filter filtering, delta represents a measurement constant of the underdetermined degree of the problem,
representing divergence operators, x, corresponding to filtering operators based on guided image structure filterst-1Image data representing the target image obtained during the previous cycle, zt-1And D, obtaining a data residual error in the last cycle, wherein A represents an undersampled Fourier transform matrix.
6. The image reconstructing method according to claim 5, wherein the residual error correction factor in the current cycle is calculated from the first residual error correction factor and the second residual error correction factor in the current cycle by using the following equation 6,
[6]
wherein o istRepresents the residual error correction factor in the present cycle, alpha represents the weight parameter, and 0<α<1。
7. The image reconstruction method according to claim 6, wherein the data residual in the current cycle is calculated from the undersampled data of K space, the image data of the target image in the current cycle, and the residual correction factor using the following equation 7,
[7] zt＝b-Axt+ot
wherein z istAnd b represents undersampled data of a K space.
8. The image reconstruction device is characterized by comprising a target image structure filter, a guide image structure filter, a target image data acquisition module and a data residual error acquisition module which are operated circularly until a cycle end condition is met;
the target image structure filter is used for filtering the image data and the data residual of the target image obtained in the previous cycle process to obtain first image data in the current cycle process;
the guide image structure filter is used for filtering the target image data and the data residual error obtained in the previous cycle process to obtain second image data in the current cycle process;
the target image data acquisition module is used for acquiring the image data of the target image in the circulation process according to the first image data and the second image data in the circulation process;
the data residual error acquisition module is used for acquiring a data residual error in the circulation process according to the undersampled data of the K space, the image data of the target image in the circulation process, the target image data and the data residual error acquired in the previous circulation process;
when the cycle ending condition is not met, the image data and the data residual error of the target image obtained in the current cycle process are filled into the image data and the data residual error of the target image in the next cycle process.
9. An electronic device, comprising:
at least one processor, and
a memory coupled with the at least one processor, the memory storing instructions that, when executed by the at least one processor, cause the at least one processor to perform the image reconstruction method of any of claims 1 to 7.
10. A machine readable storage medium having stored thereon executable instructions, which when executed cause the machine to perform the image reconstruction method of any one of claims 1 to 7.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011440541.8A CN112489155B (en) | 2020-12-07 | 2020-12-07 | Image reconstruction method and device, electronic equipment and machine-readable storage medium |
| PCT/CN2020/136550 WO2022120899A1 (en) | 2020-12-07 | 2020-12-15 | Image reconstruction method and apparatus, electronic device and machine-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011440541.8A CN112489155B (en) | 2020-12-07 | 2020-12-07 | Image reconstruction method and device, electronic equipment and machine-readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112489155A true CN112489155A (en) | 2021-03-12 |
| CN112489155B CN112489155B (en) | 2024-01-26 |
Family
ID=74941504
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011440541.8A Active CN112489155B (en) | 2020-12-07 | 2020-12-07 | Image reconstruction method and device, electronic equipment and machine-readable storage medium |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN112489155B (en) |
| WO (1) | WO2022120899A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113192151A (en) * | 2021-04-08 | 2021-07-30 | 广东工业大学 | MRI image reconstruction method based on structural similarity |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150287223A1 (en) * | 2014-04-04 | 2015-10-08 | The Board Of Trustees Of The University Of Illinois | Highly accelerated imaging and image reconstruction using adaptive sparsifying transforms |
| CN106485764A (en) * | 2016-11-02 | 2017-03-08 | 中国科学技术大学 | The quick exact reconstruction methods of MRI image |
| CN109671129A (en) * | 2018-12-14 | 2019-04-23 | 深圳先进技术研究院 | A kind of the dynamic magnetic resonance image method for reconstructing and device of auto-adaptive parameter study |
| CN111798391A (en) * | 2020-06-29 | 2020-10-20 | 东软医疗系统股份有限公司 | Image processing method and device, medical imaging equipment and system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110148215B (en) * | 2019-05-22 | 2023-05-19 | 哈尔滨工业大学 | Four-dimensional magnetic resonance image reconstruction method based on smooth constraint and local low-rank constraint model |
| CN111192214B (en) * | 2019-12-27 | 2024-03-26 | 上海商汤智能科技有限公司 | Image processing method, device, electronic equipment and storage medium |
-
2020
- 2020-12-07 CN CN202011440541.8A patent/CN112489155B/en active Active
- 2020-12-15 WO PCT/CN2020/136550 patent/WO2022120899A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150287223A1 (en) * | 2014-04-04 | 2015-10-08 | The Board Of Trustees Of The University Of Illinois | Highly accelerated imaging and image reconstruction using adaptive sparsifying transforms |
| CN106485764A (en) * | 2016-11-02 | 2017-03-08 | 中国科学技术大学 | The quick exact reconstruction methods of MRI image |
| CN109671129A (en) * | 2018-12-14 | 2019-04-23 | 深圳先进技术研究院 | A kind of the dynamic magnetic resonance image method for reconstructing and device of auto-adaptive parameter study |
| CN111798391A (en) * | 2020-06-29 | 2020-10-20 | 东软医疗系统股份有限公司 | Image processing method and device, medical imaging equipment and system |
Non-Patent Citations (1)
| Title |
|---|
| 宗春梅 等: "基于深度先验及非局部相似性的压缩感知核磁共振成像", 计算机应用, vol. 40, no. 10, pages 3054 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113192151A (en) * | 2021-04-08 | 2021-07-30 | 广东工业大学 | MRI image reconstruction method based on structural similarity |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2022120899A1 (en) | 2022-06-16 |
| CN112489155B (en) | 2024-01-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tezcan et al. | MR image reconstruction using deep density priors | |
| Kamilov et al. | Plug-and-play methods for integrating physical and learned models in computational imaging: Theory, algorithms, and applications | |
| Chung et al. | Score-based diffusion models for accelerated MRI | |
| US10671939B2 (en) | System, method and computer-accessible medium for learning an optimized variational network for medical image reconstruction | |
| Ongie et al. | Off-the-grid recovery of piecewise constant images from few Fourier samples | |
| Ravishankar et al. | Efficient blind compressed sensing using sparsifying transforms with convergence guarantees and application to magnetic resonance imaging | |
| Maggioni et al. | Nonlocal transform-domain filter for volumetric data denoising and reconstruction | |
| Ma | Improved iterative curvelet thresholding for compressed sensing and measurement | |
| Kadkhodaie et al. | Solving linear inverse problems using the prior implicit in a denoiser | |
| Wang et al. | A new alternating minimization algorithm for total variation image reconstruction | |
| St-Jean et al. | Non Local Spatial and Angular Matching: Enabling higher spatial resolution diffusion MRI datasets through adaptive denoising | |
| Zhang et al. | Denoising of 3D magnetic resonance images by using higher-order singular value decomposition | |
| Eslahi et al. | Compressive sensing image restoration using adaptive curvelet thresholding and nonlocal sparse regularization | |
| Elad et al. | Image denoising: The deep learning revolution and beyond—a survey paper | |
| Loktyushin et al. | Blind multirigid retrospective motion correction of MR images | |
| WO2020114329A1 (en) | Fast magnetic resonance parametric imaging and device | |
| Hammernik et al. | Physics-driven deep learning for computational magnetic resonance imaging: Combining physics and machine learning for improved medical imaging | |
| Cole et al. | Fast unsupervised mri reconstruction without fully-sampled ground truth data using generative adversarial networks | |
| Tian et al. | Boosting magnetic resonance image denoising with generative adversarial networks | |
| CN112489155B (en) | Image reconstruction method and device, electronic equipment and machine-readable storage medium | |
| Yeganli et al. | Noise removal from MR images via iterative regularization based on higher-order singular value decomposition | |
| Liu et al. | High-Fidelity MRI Reconstruction Using Adaptive Spatial Attention Selection and Deep Data Consistency Prior | |
| Puy et al. | Robust image reconstruction from multiview measurements | |
| Peng et al. | Accurate T2 mapping with sparsity and linear predictability filtering | |
| Qin et al. | Ground-Truth Free Meta-Learning for Deep Compressive Sampling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |