CN109741411B - Low-dose PET image reconstruction method, device, equipment and medium based on gradient domain - Google Patents
Low-dose PET image reconstruction method, device, equipment and medium based on gradient domain Download PDFInfo
- Publication number
- CN109741411B CN109741411B CN201811525022.4A CN201811525022A CN109741411B CN 109741411 B CN109741411 B CN 109741411B CN 201811525022 A CN201811525022 A CN 201811525022A CN 109741411 B CN109741411 B CN 109741411B
- Authority
- CN
- China
- Prior art keywords
- image
- gradient
- pet image
- pet
- reconstructed
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Nuclear Medicine (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention is applicable to the technical field of medical PET imaging, and provides a low-dose PET image reconstruction method, a device, equipment and a medium based on a gradient domain, wherein the method comprises the following steps: according to projection data acquired by PET equipment and a system matrix of the PET equipment, carrying out image reconstruction on a pre-initialized PET image to be reconstructed through a PET image reconstruction algorithm to obtain an initial reconstructed PET image, carrying out joint optimization solution on a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation according to the initial reconstructed PET image by adopting Lagrange multiplication to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image, thereby improving the reconstruction speed of the low-dose PET image, reducing the artifact degree of the reconstructed image and further improving the image quality of the low-dose PET image reconstruction.
Description
Technical Field
The invention belongs to the technical field of medical PET imaging, and particularly relates to a low-dose PET image reconstruction method, device, equipment and medium based on a gradient domain.
Background
Positron emission tomography (Positron Emission Tomography, PET for short) is an emission imaging technique (Emission Tomography, ET for short) that shows the metabolism of different tissues by injecting a radiopharmaceutical into the body. The PET technology is a novel imaging technology applied to clinic after computed tomography (Computed Tomography, CT for short) and magnetic resonance imaging (Magnetic Resonance Imaging, MRI for short), and the PET technology shows excellent performance in the fields of oncology, cardiovascular disease, nervous system disease research, new drug development research, and the like.
In PET imaging, the radiopharmaceutical is actually a molecular carrier that adheres to specific physiological tissues or pathological processes. The radioactive material is purposefully distributed in the human body under the guidance of the drug. The purpose of PET imaging is in fact to obtain a profile of the radioactive substance inside the human body, the working principle of which is: some radionuclides such as O-15, C-11, N-13, F-18, etc. are labeled on compounds required for metabolism of the human body and then are administered into the subject by arm intravenous injection, etc. The labeled compound decays during participation in metabolism in vivo, releasing positrons (electrons with one positive charge) annihilating with their surrounding (negatively charged) electrons to produce two gamma photons with energy of 511 keV. The photons are emitted in opposite directions on a straight line, all photons emitted from a specific area can be detected by using an external gamma camera, and then a certain algorithm is designed, so that the distribution condition of radioactive substances in the human body can be approximately obtained.
Since the radiopharmaceuticals used in PET examinations radiate the person who is in close contact with the medicine, the probability of cancer of the person who is irradiated is much higher than that of normal persons, and the consumption of the radiopharmaceuticals takes a certain weight in the cost of PET examinations. Therefore, according to the principle of reasonably using low dose (As Low As Reasonably Achievable, ALARA) proposed by the International radioprotection Commission (International Commission on Radiological Protection, ICRP) at the time of PET clinical diagnosis, images meeting clinical requirements are obtained with the minimum dose, and the radiation dose to patients is reduced as much as possible.
However, when PET image reconstruction is performed on the measurement data obtained by low-dose sampling, the speed of reconstructing an image by the existing conventional PET image reconstruction algorithm is slow, so that motion artifacts are generated on the reconstructed image, and the motion artifacts directly affect the diagnosis behaviors of doctors.
Disclosure of Invention
The invention aims to provide a low-dose PET image reconstruction method, device, equipment and medium based on a gradient domain, and aims to solve the problems that the low-dose PET image reconstruction speed is low and the quality of a reconstructed image is poor because the prior art cannot provide an effective low-dose PET image reconstruction method.
In one aspect, the present invention provides a gradient domain based low dose PET image reconstruction method comprising the steps of:
when a reconstruction request of a low-dose PET image is received, acquiring projection data acquired by PET equipment, and acquiring a system matrix of the PET equipment;
according to the projection data and the system matrix, performing image reconstruction on a pre-initialized PET image to be reconstructed through a preset PET image reconstruction algorithm to obtain an initial reconstructed PET image;
and carrying out joint optimization solution on a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation by adopting Lagrange multiplication according to the initial reconstructed PET image to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image.
In another aspect, the present invention provides a gradient domain based low dose PET image reconstruction apparatus, the apparatus comprising:
the system comprises a parameter acquisition unit, a control unit and a control unit, wherein the parameter acquisition unit is used for acquiring projection data acquired by PET equipment and acquiring a system matrix of the PET equipment when a reconstruction request of a low-dose PET image is received;
the initial reconstruction unit is used for carrying out image reconstruction on a pre-initialized PET image to be reconstructed through a preset PET image reconstruction algorithm according to the projection data and the system matrix to obtain an initial reconstructed PET image; and
and the reconstructed image obtaining unit is used for carrying out joint optimization solution on a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation by adopting Lagrange multiplication according to the initial reconstructed PET image to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image.
In another aspect, the present invention also provides a computing device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps as described above for the gradient domain based low dose PET image reconstruction method when executing the computer program.
In another aspect, the invention also provides a computer readable storage medium storing a computer program which when executed by a processor implements the steps as described above for the gradient domain based low dose PET image reconstruction method.
According to the invention, according to projection data acquired by the PET equipment and a system matrix of the PET equipment, a pre-initialized PET image to be reconstructed is subjected to image reconstruction through a PET image reconstruction algorithm to obtain an initial reconstructed PET image, and according to the initial reconstructed PET image, a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation are subjected to joint optimization solution by adopting Lagrange multiplication to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image, so that the reconstruction speed of a low-dose PET image is improved, the artifact degree of the reconstructed image is reduced, and the image quality of the low-dose PET image reconstruction is improved.
Drawings
FIG. 1 is a flowchart of an implementation of a gradient domain-based low dose PET image reconstruction method according to an embodiment of the present invention;
FIG. 2 is a flow chart for implementing iterative solution of Lagrangian equations using the Bregman iterative method in accordance with the first embodiment of the present invention;
fig. 3 is a schematic structural diagram of a gradient domain-based low-dose PET image reconstruction device according to a second embodiment of the present invention;
FIG. 4 is a schematic diagram of a preferred structure of a gradient domain-based low-dose PET image reconstruction device according to a second embodiment of the present invention;
FIG. 5 is a schematic diagram of another preferred structure of a gradient domain-based low-dose PET image reconstruction device according to a second embodiment of the present invention; and
fig. 6 is a schematic structural diagram of a computing device according to a third embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
The following describes in detail the implementation of the present invention in connection with specific embodiments:
embodiment one:
fig. 1 shows a flow of implementation of the gradient domain-based low-dose PET image reconstruction method according to the first embodiment of the present invention, and for convenience of explanation, only the portions relevant to the embodiment of the present invention are shown, which is described in detail below:
in step S101, when a request for reconstructing a low-dose PET image is received, projection data acquired by the PET apparatus is acquired, and a system matrix of the PET apparatus is acquired.
The embodiment of the invention is applicable to medical image processing platforms, systems or devices, such as personal computers, servers and the like. When a request for reconstructing a low dose PET image is received, undersampled projection data acquired by the PET device under low dose conditions is acquired, and a system matrix of the PET device is acquired, the system matrix being calculated from geometry information of the PET device.
In step S102, according to the projection data and the system matrix, an image reconstruction is performed on a pre-initialized PET image to be reconstructed by a preset PET image reconstruction algorithm, so as to obtain an initial reconstructed PET image.
In the embodiment of the invention, according to projection data and a system matrix, a pre-initialized PET image to be reconstructed is subjected to iterative operation for a preset number of times through a pre-set PET image reconstruction algorithm so as to reconstruct the image of the PET image to be reconstructed, and an initial reconstructed PET image is obtained, wherein the PET image to be reconstructed is a two-dimensional image, and the pre-set PET image reconstruction algorithm is a maximum likelihood expectation maximization algorithm (Maximum Likelihood Expectation Maximized, abbreviated as MLEM) or an ordered subset expectation maximization algorithm (Ordered Subset Expectation Maximization, abbreviated as OSEM) or a maximum posterior probability algorithm (Maximum A Posterior, MAP).
In initializing the PET image to be reconstructed, the pixel values of the PET image to be reconstructed are initialized to zero, as an example.
In step S103, according to the initial reconstructed PET image, a lagrangian multiplication is adopted to perform joint optimization solution on a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation, so as to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image.
In the embodiment of the invention, the lagrangian multiplication is adopted to reconstruct the pre-constructed image into an equation y u ＝G u m and pre-constructed gradient domain image feature selection equation
Performing simultaneous operation to obtain corresponding Lagrangian equation +.>
Then, the Lagrangian equation is optimized and solved to finally obtain a target reconstructed PET image corresponding to the initial reconstructed PET image, wherein y u For projection data, G u For the system matrix, m is the PET image to be reconstructed (i.e. the target reconstructed PET image), v 1 R is a preset weight parameter l Extracting matrices for image blocks, i.e. based on R l Extracting l image blocks from the gradient image omega, wherein D is a feature matrix of the gradient image omega, and alpha is the feature matrix of the gradient image omega l For the feature vector corresponding to the first image block extracted from the gradient image omega (i) Representing a horizontal/vertical gradient image corresponding to the original reconstructed PET image, i epsilon {1,2} represents the direction (horizontal/vertical) of the gradient image omega, and L represents a control coefficient for the sparsity of the feature vector.
In the embodiment of the invention, the control coefficient L of the feature vector sparsity is preferably set to be 5, so that the noise of the PET image which is sparsely represented by the learned feature matrix and the feature vector is better reduced.
When the lagrangian multiplication is adopted to carry out joint optimization solution on the pre-constructed image reconstruction equation and the pre-constructed gradient domain image feature selection equation, preferably, the lagrangian iteration method is adopted to carry out iteration solution on the lagrangian equation which is formed by combining the image reconstruction equation and the gradient domain image feature selection equation, so that the reconstruction speed of the PET image is improved.
Further preferably, a Bregman iteration method is adopted to decompose a Lagrangian equation into a gradient image update function, an iteration error correction function, a PET image reconstruction function and a feature extraction function, so as to carry out iteration solution on the gradient image update function, the iteration error correction function, the PET image reconstruction function and the feature extraction function to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image, thereby further improving the reconstruction speed of the PET image and improving the image quality of the reconstructed target reconstructed PET image.
Preferably, the feature extraction function obtained by decomposing the Lagrangian equation is
The obtained feature matrix D and feature vector alpha l The method is used as an initial value for updating the gradient image omega in the next iteration, so that the initial reconstructed PET image is converted from an image domain to a gradient domain, and the gradient domain image is subjected to feature learning, so that the initial reconstructed PET image is sparsely represented by the learned feature matrix and the feature vector, thereby reducing noise in the initial reconstructed PET image, and further improving the reconstruction effect of the subsequent PET image. Where k is the current iteration number.
Preferably, the gradient image update function obtained by decomposing the Lagrangian equation is as follows
The obtained gradient image omega is used as the target reconstructed PET image m in the next iterationInitial values of the row reconstruction are used for reducing the artifact degree of the target reconstructed PET image of the subsequent reconstruction. Wherein v is 2 The weight is preset for controlling the iteration error in the Bregman iteration, and b is an error correction value of the Bregman iteration.
In an embodiment of the present invention, it is further preferable that v 2 Set to 1 to further reduce the artifact level of the subsequently reconstructed target reconstructed PET image.
Preferably, the iterative error correction function obtained by decomposing the Lagrangian equation
The obtained error correction value b is used as an initial value for reconstructing the target reconstructed PET image m in the next iteration, so that the image quality of the reconstructed PET image is improved.
Preferably, the PET image reconstruction function obtained by decomposing Lagrangian equation
ω k B for the gradient image of the kth iteration k The error correction value for the kth iteration, thereby improving the image quality of the reconstructed PET image.
As shown in fig. 2, the iterative solution of the lagrangian equation using the Bregman iterative method is preferably achieved by:
in step S201, a gradient image corresponding to the initially reconstructed PET image is updated using a gradient image update function according to a preset initial feature matrix and a preset initial feature vector.
In step S202, the gradient image is restored from the gradient domain to the image domain using the PET image reconstruction function according to the updated gradient image and the error correction value obtained by the iterative error correction function, to obtain the target reconstructed PET image.
In step S203, it is determined whether the current iteration number reaches a preset iteration threshold.
In the embodiment of the present invention, when the current iteration number reaches a preset iteration threshold (e.g., 50 times), step S204 is performed, otherwise, step S205 is skipped.
In step S204, a target reconstructed PET image is output.
In step S205, the target reconstructed PET image is set as an initial reconstructed PET image, and a corresponding number of image blocks are extracted from gradient images corresponding to the initial reconstructed PET image according to a preset image block extraction matrix, the gradient images including a horizontal gradient image and a vertical gradient image.
In the embodiment of the invention, an initial reconstructed PET image is firstly converted from an image domain to a gradient domain to obtain a horizontal gradient image and a vertical gradient image, and a corresponding number of horizontal image blocks and vertical image blocks are respectively extracted from the horizontal gradient image and the vertical gradient image according to a preset image block extraction matrix.
In step S206, feature learning is performed on the image block until the feature matrix corresponding to the gradient image obtained by learning and the feature vector corresponding to the image block satisfy the feature extraction function.
In the embodiment of the invention, feature learning is performed on the extracted horizontal image blocks and the extracted vertical image blocks respectively until a horizontal/vertical feature matrix corresponding to the horizontal/vertical gradient image obtained by learning and a horizontal/vertical feature vector corresponding to the horizontal/vertical image blocks meet a feature extraction function, wherein each column of the feature matrix corresponds to the feature vector corresponding to each image block one by one.
In step S207, the feature matrix and the feature vector are set as an initial feature matrix and an initial feature vector, respectively, the current iteration number is increased by 1 time, and the process goes to step S201, and the next iteration is continued to reconstruct the PET image.
The step S201-the step S207 are used for carrying out iterative solution on the Lagrangian equation so as to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image, thereby reducing the artifact degree of the reconstructed image and improving the image quality of the reconstructed low-dose PET image.
In the embodiment of the invention, according to the projection data acquired by the PET equipment and the system matrix of the PET equipment, the pre-initialized PET image to be reconstructed is subjected to image reconstruction through a PET image reconstruction algorithm to obtain an initial reconstructed PET image, and according to the initial reconstructed PET image, a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation are subjected to joint optimization solution by adopting Lagrange multiplication to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image, so that the reconstruction speed of the low-dose PET image is improved, the artifact degree of the reconstructed image is reduced, and the image quality of the low-dose PET image reconstruction is improved.
Embodiment two:
fig. 3 shows the structure of a gradient domain-based low-dose PET image reconstruction device according to the second embodiment of the present invention, and for convenience of explanation, only the portions related to the second embodiment of the present invention are shown, including:
a parameter acquisition unit 31 for acquiring projection data acquired by the PET apparatus and acquiring a system matrix of the PET apparatus when a request for reconstructing a low-dose PET image is received;
an initial reconstruction unit 32, configured to perform image reconstruction on a pre-initialized PET image to be reconstructed by a preset PET image reconstruction algorithm according to the projection data and the system matrix, so as to obtain an initial reconstructed PET image; and
the reconstructed image obtaining unit 33 is configured to perform joint optimization solution on a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation by using lagrangian multiplication according to the initial reconstructed PET image, so as to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image.
As shown in fig. 4, the reconstructed image obtaining unit 33 preferably includes:
and the iteration solving unit 331 is used for carrying out iteration solving on a Lagrangian equation which is formed by combining an image reconstruction equation and a gradient domain image characteristic selection equation by adopting a Bregman iteration method.
Further preferably, the iterative solving unit 331 includes:
the equation decomposition unit 3311 is configured to decompose the lagrangian equation into a gradient image update function, an iterative error correction function, a PET image reconstruction function, and a feature extraction function by using a Bregman iteration method, so as to perform iterative solution on the gradient image update function, the iterative error correction function, the PET image reconstruction function, and the feature extraction function, thereby obtaining a target reconstructed PET image corresponding to the initial reconstructed PET image.
Further preferably, as shown in fig. 5, the equation decomposing unit 3311 includes:
a gradient image updating unit 51, configured to update a gradient image corresponding to the initially reconstructed PET image by using a gradient image updating function according to a preset initial feature matrix and a preset initial feature vector;
a PET image reconstruction unit 52, configured to recover the gradient image from the gradient domain to the image domain using the PET image reconstruction function according to the updated gradient image and the error correction value obtained by the iterative error correction function, to obtain a target reconstructed PET image;
an iteration number judging unit 53, configured to judge whether the current iteration number reaches a preset iteration threshold;
a PET image output unit 54 for outputting a target reconstructed PET image;
the image block extracting unit 55 is configured to set the target reconstructed PET image as an initial reconstructed PET image, and extract a corresponding number of image blocks from gradient images corresponding to the initial reconstructed PET image according to a preset image block extracting matrix, where the gradient images include a horizontal gradient image and a vertical gradient image;
the feature learning unit 56 is configured to perform feature learning on the image block until the feature matrix corresponding to the gradient image obtained by learning and the feature vector corresponding to the image block satisfy the feature extraction function; and
a parameter setting unit 57 for setting the feature matrix and the feature vector as an initial feature matrix and an initial feature vector, respectively, and triggering the gradient image updating unit 51 to continue the next iteration to reconstruct the PET image.
In the embodiment of the present invention, each unit of the low dose PET image reconstruction device based on the gradient domain may be implemented by a corresponding hardware or software unit, and each unit may be an independent software or hardware unit, or may be integrated into one software or hardware unit, which is not used to limit the present invention. In particular, the implementation of each unit may refer to the description of the foregoing embodiment one, which is not repeated herein.
Embodiment III:
fig. 6 shows the structure of a computing device provided by the third embodiment of the present invention, and only the portions relevant to the embodiment of the present invention are shown for convenience of explanation.
The computing device 6 of an embodiment of the present invention includes a processor 60, a memory 61, and a computer program 62 stored in the memory 61 and executable on the processor 60. The processor 60, when executing the computer program 62, implements the steps of the embodiment of the gradient domain based low dose PET image reconstruction method described above, such as steps S101 to S103 shown in fig. 1. Alternatively, the processor 60, when executing the computer program 62, implements the functions of the units in the above-described embodiments of the apparatus, such as the functions of the units 31 to 33 shown in fig. 3.
In the embodiment of the invention, according to the projection data acquired by the PET equipment and the system matrix of the PET equipment, the pre-initialized PET image to be reconstructed is subjected to image reconstruction through a PET image reconstruction algorithm to obtain an initial reconstructed PET image, and according to the initial reconstructed PET image, a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation are subjected to joint optimization solution by adopting Lagrange multiplication to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image, so that the reconstruction speed of the low-dose PET image is improved, the artifact degree of the reconstructed image is reduced, and the image quality of the low-dose PET image reconstruction is improved.
The computing device of the embodiment of the invention can be a personal computer or a server. The steps implemented when the processor 60 executes the computer program 62 in the computing device 6 to implement the gradient domain-based low dose PET image reconstruction method may refer to the description of the foregoing method embodiments, and will not be repeated here.
Embodiment four:
in an embodiment of the present invention, a computer readable storage medium is provided, which stores a computer program which, when executed by a processor, implements the steps in the embodiment of the gradient domain based low dose PET image reconstruction method described above, for example, steps S101 to S103 shown in fig. 1. Alternatively, the computer program, when executed by a processor, implements the functions of the units in the above-described embodiments of the apparatus, such as the functions of the units 31 to 33 shown in fig. 3.
In the embodiment of the invention, according to the projection data acquired by the PET equipment and the system matrix of the PET equipment, the pre-initialized PET image to be reconstructed is subjected to image reconstruction through a PET image reconstruction algorithm to obtain an initial reconstructed PET image, and according to the initial reconstructed PET image, a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation are subjected to joint optimization solution by adopting Lagrange multiplication to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image, so that the reconstruction speed of the low-dose PET image is improved, the artifact degree of the reconstructed image is reduced, and the image quality of the low-dose PET image reconstruction is improved.
The computer readable storage medium of embodiments of the present invention may include any entity or device capable of carrying computer program code, recording medium, such as ROM/RAM, magnetic disk, optical disk, flash memory, and so on.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.
Claims (14)
1. A method for gradient domain-based low dose PET image reconstruction, the method comprising the steps of:
when a reconstruction request of a low-dose PET image is received, acquiring projection data acquired by PET equipment, and acquiring a system matrix of the PET equipment;
according to the projection data and the system matrix, performing image reconstruction on a pre-initialized PET image to be reconstructed through a preset PET image reconstruction algorithm to obtain an initial reconstructed PET image;
according to the initial reconstructed PET image, carrying out joint optimization solution on a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation by adopting Lagrangian multiplication to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image;
the method comprises the steps of carrying out joint optimization solution on a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation by using Lagrangian multiplication to obtain a target reconstruction PET image corresponding to the initial reconstruction PET image, wherein the method comprises the following steps: combining the pre-constructed image reconstruction equation and the pre-constructed gradient domain image feature selection equation by using Lagrange multiplication to obtain a corresponding Lagrange equation;
and carrying out optimization solving on the Lagrangian equation to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image;
the pre-constructed image reconstruction equation includes y u ＝G u m, the pre-constructed gradient domain image feature selection equation comprises
The obtaining the corresponding Lagrangian equation includes:
2. The method of claim 1, wherein the step of jointly optimizing the pre-constructed image reconstruction equation and the pre-constructed gradient domain image feature selection equation using lagrangian multiplication comprises:
and adopting a Bregman iteration method to carry out iteration solution on a Lagrangian equation which is formed by combining the image reconstruction equation and the gradient domain image characteristic selection equation.
3. The method of claim 2, wherein the step of iteratively solving the lagrangian equation using a Bregman iterative method comprises:
and decomposing the Lagrangian equation into a gradient image update function, an iteration error correction function, a PET image reconstruction function and a feature extraction function by adopting a Bregman iteration method so as to carry out iteration solution on the gradient image update function, the iteration error correction function, the PET image reconstruction function and the feature extraction function, thereby obtaining a target reconstructed PET image corresponding to the initial reconstructed PET image.
4. A method according to claim 3, wherein the step of iteratively solving the lagrangian equation using a Bregman iterative method comprises:
updating a gradient image corresponding to the initial reconstructed PET image by using the gradient image updating function according to a preset initial feature matrix and a preset initial feature vector;
recovering the gradient image from the gradient domain to the image domain by using the PET image reconstruction function according to the updated gradient image and an error correction value obtained by the iterative error correction function to obtain a target reconstructed PET image;
judging whether the current iteration number reaches a preset iteration threshold value or not;
if yes, outputting the target reconstructed PET image;
otherwise, setting the target reconstructed PET image as the initial reconstructed PET image, and extracting a corresponding number of image blocks from gradient images corresponding to the initial reconstructed PET image according to a preset image block extraction matrix, wherein the gradient images comprise horizontal gradient images and vertical gradient images;
performing feature learning on the image block until the feature matrix corresponding to the gradient image and the feature vector corresponding to the image block obtained through learning meet the feature extraction function;
and setting the feature matrix and the feature vector as the initial feature matrix and the initial feature vector respectively, and jumping to the step of updating the gradient image by using the gradient image updating function.
5. The method of claim 3, wherein the feature extraction function is
K is the current iteration number.
6. A method according to claim 3, wherein the gradient image update function
Wherein V is 2 The weights representing the control iteration errors in the Bregman iteration, b being the error correction value of the Bregman iteration.
7. A method according to claim 3, wherein the iterative error correction function
8. The method of claim 3, wherein the PET image reconstruction function
V 1 Is a preset weight parameter omega k B for the gradient image of the kth iteration k Is the error correction value for the kth iteration.
9. A gradient domain-based low dose PET image reconstruction device, the device comprising:
the system comprises a parameter acquisition unit, a control unit and a control unit, wherein the parameter acquisition unit is used for acquiring projection data acquired by PET equipment and acquiring a system matrix of the PET equipment when a reconstruction request of a low-dose PET image is received;
the initial reconstruction unit is used for carrying out image reconstruction on a pre-initialized PET image to be reconstructed through a preset PET image reconstruction algorithm according to the projection data and the system matrix to obtain an initial reconstructed PET image; and
and the reconstructed image obtaining unit is used for carrying out joint optimization solution on a pre-constructed image reconstruction equation and a pre-constructed gradient domain image feature selection equation by adopting Lagrange multiplication according to the initial reconstructed PET image to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image.
10. The apparatus according to claim 9, wherein the reconstructed image obtaining unit includes:
and the iteration solving unit is used for carrying out iteration solving on a Lagrangian equation which is formed by combining the image reconstruction equation and the gradient domain image characteristic selection equation by adopting a Bregman iteration method.
11. The apparatus of claim 10, wherein the iterative solution unit comprises:
and the equation decomposition unit is used for decomposing the Lagrangian equation into a gradient image update function, an iterative error correction function, a PET image reconstruction function and a feature extraction function by adopting a Bregman iteration method so as to carry out iterative solution on the gradient image update function, the iterative error correction function, the PET image reconstruction function and the feature extraction function to obtain a target reconstructed PET image corresponding to the initial reconstructed PET image.
12. The apparatus of claim 11, wherein the equation decomposition unit comprises: the gradient image updating unit is used for updating the gradient image corresponding to the initial reconstructed PET image by using the gradient image updating function according to a preset initial feature matrix and a preset initial feature vector;
the PET image reconstruction unit is used for recovering the gradient image from the gradient domain to the image domain by using the PET image reconstruction function according to the updated gradient image and the error correction value obtained by the iterative error correction function to obtain a target reconstructed PET image;
the iteration number judging unit is used for judging whether the current iteration number reaches a preset iteration threshold value or not;
a PET image output unit for outputting the target reconstructed PET image;
the image block extraction unit is used for setting the target reconstructed PET image as the initial reconstructed PET image, and extracting a corresponding number of image blocks from gradient images corresponding to the initial reconstructed PET image according to a preset image block extraction matrix, wherein the gradient images comprise horizontal gradient images and vertical gradient images;
the feature learning unit is used for carrying out feature learning on the image block until the feature matrix corresponding to the gradient image obtained by learning and the feature vector corresponding to the image block meet the feature extraction function; and
and the parameter setting unit is used for setting the feature matrix and the feature vector as the initial feature matrix and the initial feature vector respectively and triggering the gradient image updating unit to execute the step of updating the gradient image by using the gradient image updating function.
13. A computing device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method of any of claims 1 to 8 when the computer program is executed.
14. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of the method according to any one of claims 1 to 8.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811525022.4A CN109741411B (en) | 2018-12-13 | 2018-12-13 | Low-dose PET image reconstruction method, device, equipment and medium based on gradient domain |
| PCT/CN2019/071098 WO2020118828A1 (en) | 2018-12-13 | 2019-01-10 | Gradient domain-based low-dose pet image reconstruction method and apparatus, device, and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811525022.4A CN109741411B (en) | 2018-12-13 | 2018-12-13 | Low-dose PET image reconstruction method, device, equipment and medium based on gradient domain |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109741411A CN109741411A (en) | 2019-05-10 |
| CN109741411B true CN109741411B (en) | 2023-04-28 |
Family
ID=66359015
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811525022.4A Active CN109741411B (en) | 2018-12-13 | 2018-12-13 | Low-dose PET image reconstruction method, device, equipment and medium based on gradient domain |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109741411B (en) |
| WO (1) | WO2020118828A1 (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112365593B (en) * | 2020-11-12 | 2024-03-29 | 江苏赛诺格兰医疗科技有限公司 | PET image reconstruction method and system |
| CN112614204A (en) * | 2020-12-29 | 2021-04-06 | 哈尔滨理工大学 | Capacitance tomography image reconstruction method based on improved least square |
| CN112766395B (en) * | 2021-01-27 | 2023-11-28 | 中国地质大学（北京） | Image matching method and device, electronic equipment and readable storage medium |
| CN116503570B (en) * | 2023-06-29 | 2023-11-24 | 聚时科技(深圳)有限公司 | Three-dimensional reconstruction method and related device for image |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103247061A (en) * | 2013-02-05 | 2013-08-14 | 南方医科大学 | Augmented lagrangian iterative reconstruction method of X-ray image and CI image |
| CN103559728A (en) * | 2013-10-31 | 2014-02-05 | 南方医科大学 | PET image maximum posterior reconstruction method based on united prior model with dissection function |
| CN104408758A (en) * | 2014-11-12 | 2015-03-11 | 南方医科大学 | Low-dose processing method of energy spectrum CT image |
| CN104574416A (en) * | 2015-01-27 | 2015-04-29 | 南方医科大学 | Low-dose energy spectrum CT image denoising method |
| CN104657950A (en) * | 2015-02-16 | 2015-05-27 | 浙江大学 | Dynamic PET (positron emission tomography) image reconstruction method based on Poisson TV |
| CN104751429A (en) * | 2015-01-27 | 2015-07-01 | 南方医科大学 | Dictionary learning based low-dosage energy spectrum CT image processing method |
| CN106056554A (en) * | 2016-06-01 | 2016-10-26 | 南昌大学 | Magnetic-resonance fast imaging method based on gradient-domain convolution sparse coding |
| CN106056644A (en) * | 2016-05-24 | 2016-10-26 | 深圳先进技术研究院 | Data processing method and data processing device for CT scanning |
| CN106530258A (en) * | 2016-11-22 | 2017-03-22 | 哈尔滨工业大学 | Fast iterative magnetic resonance image reconstruction method based on high-order total variation regularization |
| CN106683144A (en) * | 2016-12-30 | 2017-05-17 | 上海联影医疗科技有限公司 | Image iteration reconstruction method and device |
| CN107358653A (en) * | 2017-08-15 | 2017-11-17 | 北京数字精准医疗科技有限公司 | Imaging reconstruction method and device |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2232444B1 (en) * | 2007-12-20 | 2011-10-12 | Wisconsin Alumni Research Foundation | Method for dynamic prior image constrained image reconstruction |
| US8879811B2 (en) * | 2012-03-28 | 2014-11-04 | Siemens Aktiengesellschaft | Alternating direction of multipliers method for parallel MRI reconstruction |
| US9600910B2 (en) * | 2014-01-08 | 2017-03-21 | Rensselaer Polytechnic Institute | Attenuation map reconstruction from TOF PET data |
| US9734601B2 (en) * | 2014-04-04 | 2017-08-15 | The Board Of Trustees Of The University Of Illinois | Highly accelerated imaging and image reconstruction using adaptive sparsifying transforms |
| CN107845120B (en) * | 2017-09-27 | 2020-11-27 | 深圳先进技术研究院 | PET image reconstruction method, system, terminal and readable storage medium |
-
2018
- 2018-12-13 CN CN201811525022.4A patent/CN109741411B/en active Active
-
2019
- 2019-01-10 WO PCT/CN2019/071098 patent/WO2020118828A1/en active Application Filing
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103247061A (en) * | 2013-02-05 | 2013-08-14 | 南方医科大学 | Augmented lagrangian iterative reconstruction method of X-ray image and CI image |
| CN103559728A (en) * | 2013-10-31 | 2014-02-05 | 南方医科大学 | PET image maximum posterior reconstruction method based on united prior model with dissection function |
| CN104408758A (en) * | 2014-11-12 | 2015-03-11 | 南方医科大学 | Low-dose processing method of energy spectrum CT image |
| CN104574416A (en) * | 2015-01-27 | 2015-04-29 | 南方医科大学 | Low-dose energy spectrum CT image denoising method |
| CN104751429A (en) * | 2015-01-27 | 2015-07-01 | 南方医科大学 | Dictionary learning based low-dosage energy spectrum CT image processing method |
| CN104657950A (en) * | 2015-02-16 | 2015-05-27 | 浙江大学 | Dynamic PET (positron emission tomography) image reconstruction method based on Poisson TV |
| CN106056644A (en) * | 2016-05-24 | 2016-10-26 | 深圳先进技术研究院 | Data processing method and data processing device for CT scanning |
| CN106056554A (en) * | 2016-06-01 | 2016-10-26 | 南昌大学 | Magnetic-resonance fast imaging method based on gradient-domain convolution sparse coding |
| CN106530258A (en) * | 2016-11-22 | 2017-03-22 | 哈尔滨工业大学 | Fast iterative magnetic resonance image reconstruction method based on high-order total variation regularization |
| CN106683144A (en) * | 2016-12-30 | 2017-05-17 | 上海联影医疗科技有限公司 | Image iteration reconstruction method and device |
| CN107358653A (en) * | 2017-08-15 | 2017-11-17 | 北京数字精准医疗科技有限公司 | Imaging reconstruction method and device |
Non-Patent Citations (2)
| Title |
|---|
| 低剂量口腔CT成像***关键技术与成像方法研究;胡战利 等;《集成技术》;20130915;全文 * |
| 快速磁共振成像;梁栋;《2018第二十届全国波谱学学术年会会议论文摘要集》;20181012;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109741411A (en) | 2019-05-10 |
| WO2020118828A1 (en) | 2020-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109741411B (en) | Low-dose PET image reconstruction method, device, equipment and medium based on gradient domain | |
| Gong et al. | Iterative PET image reconstruction using convolutional neural network representation | |
| US11576628B2 (en) | Full dose PET image estimation from low-dose PET imaging using deep learning | |
| US9155514B2 (en) | Reconstruction with partially known attenuation information in time of flight positron emission tomography | |
| CN109215093B (en) | Low-dose PET image reconstruction method, device, equipment and storage medium | |
| US20190304144A1 (en) | System and method for generating attenuation map | |
| CN110151210B (en) | Medical image processing method, system, device and computer readable medium | |
| US8421021B2 (en) | Motion correction of SPECT images | |
| CN107638188B (en) | Image attenuation correction method and device | |
| US11727610B2 (en) | System and method for image processing | |
| CN106999135B (en) | Radiation emission imaging system and method | |
| Cheng et al. | Applications of artificial intelligence in nuclear medicine image generation | |
| CN113226183A (en) | Determination of metabolic rate from static PET scan | |
| CN110415310B (en) | Medical scanning imaging method, device, storage medium and computer equipment | |
| CN108932741A (en) | Dynamic PET parameter imaging method, device, system and computer readable storage medium | |
| US11941805B2 (en) | Systems and methods for image processing | |
| US10217250B2 (en) | Multi-view tomographic reconstruction | |
| EP4148680A1 (en) | Attenuation correction-based weighting for tomographic inconsistency detection | |
| Karakatsanis et al. | Generalized 3D and 4D motion compensated whole-body PET image reconstruction employing nested EM deconvolution | |
| Tsui et al. | Advances in 4D gated cardiac PET imaging for image quality improvement and cardiac motion and contractility estimation | |
| CN112634147B (en) | PET image noise reduction method, system, device and medium for self-supervision learning | |
| US20230069017A1 (en) | Systems and methods for image reconstruction | |
| US8437525B2 (en) | Method and system for using a modified ordered subsets scheme for attenuation weighted reconstruction | |
| US20220215599A1 (en) | Systems and methods for motion estimation in pet imaging using ai image reconstructions | |
| US20240193828A1 (en) | Systems and methods of list-mode image reconstruction in positron emission tomography (pet) systems |
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 |