Image Reconstruction Method and Device for Positron Emission Tomography Data

Abstract

The present application discloses an image reconstruction method and apparatus for positron emission tomography data, relating to the field of image data processing, aiming to improve the problem that the existing positron emission tomography images have low diagnostic accuracy due to the influence of respiratory motion. The method includes: acquiring single-phase CT image data and positron emission tomography data of multiple phases after phase division; selecting a reference phase reconstruction image from multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and acquiring a motion field transformation matrix matching the reference phase reconstruction image; performing image reconstruction processing based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases to obtain a reconstructed image of the positron emission tomography data.

Description

Technical Field

[0001] This application relates to the technical field of image data processing, and in particular, to an image reconstruction method and device for positron emission tomography data. Background Art

[0002] Positron Emission Tomography (PET) is a relatively advanced clinical imaging technology in the field of nuclear medicine. When using a positron emission tomograph to perform a routine clinical scan on a patient's upper abdomen or chest area, due to respiratory movement, the lungs undergo periodic expansion and contraction movements, and the scanning time for each bed position of PET is relatively long compared to the respiratory cycle (about 1 - 2 minutes), resulting in the measured data being the superposition result after multiple respiratory cycle movements of the lungs. As a result, the boundaries between the lungs and the liver become blurred in the PET reconstructed image, that is, image artifacts are generated. When there are lesion points on tissue organs such as the lungs or the apex of the liver, the periodic movement will cause the size of the lesion points to expand in the reconstructed image, the contrast of the lesion points to decrease, and they may even be submerged in the background, affecting the diagnostic accuracy.

[0003] Currently, to solve the above problems, the existing technologies mainly include obtaining multi-phase PET images by using respiratory gating scanning, performing image reconstruction in an alternating iteration manner to obtain single-phase PET images, and segmenting the image to only consider the region of interest and performing motion correction on it. In addition, 4D MR or CT technology will be combined to determine the region of interest. However, using the respiratory gating scanning method will increase the total scanning time, thus affecting the user experience and scanning efficiency; when performing image reconstruction in an alternating iteration manner, if PET data with a conventional scanning duration is used, it is necessary to estimate the radioactive activity distribution map and the linear attenuation coefficient map simultaneously, which will affect the accuracy of both sets of results; combining 4D MR or CT technology will cause the patient to receive additional radiation dose; when only considering the region of interest, it is easy to ignore the motion deformation of other regions, and finally, errors such as image artifacts will occur in the reconstructed image. Summary of the Invention

[0004] In view of this, this application provides an image reconstruction method and device for positron emission tomography data, mainly aiming to improve the technical problem that the existing positron emission tomography images have low diagnostic accuracy due to the influence of respiratory movement.

[0005] According to one aspect of this application, an image reconstruction method for positron emission tomography data is provided, including:

[0006] Obtaining single-phase CT image data and positron emission tomography data of multiple phases after phase division;

[0007] Select a reference phase reconstruction image from multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and obtain a motion field transformation matrix matching the reference phase reconstruction image;

[0008] Perform image reconstruction processing based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases to obtain a reconstruction image of the positron emission tomography data.

[0009] Preferably, the obtaining of the motion field transformation matrix matching the reference phase reconstruction image specifically includes:

[0010] Construct a motion field transformation matrix based on the motion field information from the reference phase reconstruction image to the basic reconstruction images of each phase;

[0011] Update the motion field transformation matrix through image transformation and projection processing to perform motion correction based on the updated motion field transformation matrix.

[0012] Preferably, the updating of the motion field transformation matrix through projection processing specifically includes:

[0013] Based on the motion field transformation matrix, orthogonally project the basic reconstruction images of each phase into the reference phase reconstruction image to obtain the orthogonal projection data of the basic reconstruction images;

[0014] Accumulate the orthogonal projection data of each phase to obtain the calculated orthogonal projection value of the reference phase reconstruction image;

[0015] If the difference between the calculated orthogonal projection value and the measured orthogonal projection value of the reference phase reconstruction image is less than a preset difference threshold, determine the motion field transformation matrix as the updated motion field transformation matrix;

[0016] If the difference between the calculated orthogonal projection value and the measured orthogonal projection value of the reference phase reconstruction image is greater than or equal to the preset difference threshold, re-execute the update of the motion field transformation matrix.

[0017] Preferably, the performing of image reconstruction processing on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases specifically includes:

[0018] Based on the updated motion field transformation matrix, transform the reference phase reconstruction image into the basic reconstruction images of multiple phases to obtain projection data of multiple phases;

[0019] Performing image reconstruction processing on the projection data of multiple phases, the measured projection data of multiple phases, and the single-phase CT image data to obtain a reconstructed image of the positron emission tomography data.

[0020] Preferably, before obtaining the positron emission tomography data of multiple phases after phase division, the method further includes:

[0021] Obtaining positron emission tomography data;

[0022] Performing phase division on the positron emission tomography data according to the respiratory gating signal to obtain positron emission tomography data of multiple phases.

[0023] Preferably, the method further includes:

[0024] If the positron emission tomography data of multiple phases after phase division contains time-of-flight information, obtaining a distribution map of annihilation event positions for each phase based on pseudo-backprojection;

[0025] Based on the distribution map of annihilation event positions, obtaining a first motion field transformation matrix between the basic reconstruction images of each phase, and a second motion field transformation matrix from the basic reconstruction images of each phase to the reference phase reconstruction image, to obtain a reference motion field transformation matrix as prior information for updating the motion field transformation matrix.

[0026] Preferably, after obtaining the positron emission tomography data of multiple phases after phase division, the method further includes:

[0027] Performing data optimization on the positron emission tomography data of multiple phases based on a deep neural network model that has completed model training, where the deep neural network model is trained based on a model training set including positron emission tomography data of a first duration and positron emission tomography data of a second duration, and the first duration is at least twice greater than the second duration.

[0028] According to another aspect of the present application, there is provided an image reconstruction device for positron emission tomography data, including:

[0029] A first acquisition module for acquiring single-phase CT image data and positron emission tomography data of multiple phases after phase division;

[0030] A second acquisition module for selecting a reference phase reconstruction image from the multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and acquiring a motion field transformation matrix matching the reference phase reconstruction image;

[0031] A reconstruction module, configured to perform image reconstruction processing based on the sports field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases, so as to obtain a reconstructed image of the positron emission tomography data.

[0032] Preferably, the second acquisition module specifically includes:

[0033] A construction unit, configured to construct a sports field transformation matrix based on the sports field information between the reference-phase reconstructed image and the basic reconstructed images of each phase;

[0034] An update unit, configured to update the sports field transformation matrix through image transformation and projection processing, so as to perform motion correction based on the updated sports field transformation matrix.

[0035] Preferably, the update unit specifically includes:

[0036] A forward projection subunit, configured to forward project the basic reconstructed images of each phase into the reference-phase reconstructed image based on the sports field transformation matrix, so as to obtain forward projection data of the basic reconstructed images;

[0037] An accumulation subunit, configured to perform accumulation processing on the forward projection data of each phase, so as to obtain a calculated forward projection value of the reference-phase reconstructed image;

[0038] A determination subunit, configured to determine that the sports field transformation matrix is the updated sports field transformation matrix if the difference between the calculated forward projection value and the measured forward projection value of the reference-phase reconstructed image is less than a preset difference threshold;

[0039] An update subunit, configured to re-execute the update of the sports field transformation matrix if the difference between the calculated forward projection value and the measured forward projection value of the reference-phase reconstructed image is greater than or equal to the preset difference threshold.

[0040] Preferably, the reconstruction module specifically includes:

[0041] A forward projection unit, configured to transform the reference-phase reconstructed image into the basic reconstructed images of multiple phases based on the updated sports field transformation matrix, so as to obtain projection data of multiple phases;

[0042] A reconstruction unit, configured to perform image reconstruction processing based on the projection data of multiple phases, the measured projection data of multiple phases, and the single-phase CT image data, so as to obtain a reconstructed image of the positron emission tomography data.

[0043] Preferably, before the first acquisition module, the device further includes:

[0044] A third acquisition module, configured to acquire positron emission tomography (PET) data;

[0045] A division module, configured to perform phase division on the positron emission tomography data according to a respiratory gating signal, so as to obtain positron emission tomography data of multiple phases.

[0046] Preferably, the apparatus further includes:

[0047] A fourth acquisition module, configured to, if the positron emission tomography data of multiple phases after the phase division includes time-of-flight information, obtain a distribution map of annihilation event positions of each phase based on pseudo back-projection;

[0048] The fourth acquisition module is further configured to obtain a first motion field transformation matrix between the basic reconstruction images of each phase and a second motion field transformation matrix from the basic reconstruction images of each phase to the reference phase reconstruction image based on the distribution map of annihilation event positions, so as to obtain a reference motion field transformation matrix as prior information for updating the motion field transformation matrix.

[0049] Preferably, after the first acquisition module, the apparatus further includes:

[0050] An optimization module, configured to perform data optimization on the positron emission tomography data of multiple phases based on a deep neural network model that has completed model training. The deep neural network model is obtained by performing model training based on a model training set including positron emission tomography data of a first duration and positron emission tomography data of a second duration, where the first duration is at least twice greater than the second duration.

[0051] According to another aspect of the present application, there is provided a storage medium storing at least one executable instruction, and the executable instruction causes a processor to perform operations corresponding to the above-mentioned image reconstruction method for positron emission tomography data.

[0052] According to still another aspect of the present application, there is provided a terminal including a processor, a memory, a communication interface, and a communication bus, and the processor, the memory, and the communication interface complete communication with each other through the communication bus;

[0053] The memory is configured to store at least one executable instruction, and the executable instruction causes the processor to perform operations corresponding to the above-mentioned image reconstruction method for positron emission tomography data.

[0054] By means of the above technical solutions, the technical solutions provided by the embodiments of the present application at least have the following advantages:

[0055] The present application provides an image reconstruction method and apparatus for positron emission tomography data. First, single-phase CT image data and positron emission tomography data of multiple phases after phase division are obtained. Secondly, a reference phase reconstruction image is selected from multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and a motion field transformation matrix matching the reference phase reconstruction image is obtained. Finally, image reconstruction processing is performed based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases to obtain a reconstructed image of the positron emission tomography data. Compared with the prior art, in the embodiments of the present application, a motion field transformation matrix matching a multi-phase basic reconstruction image and a reference phase reconstruction image is obtained, and the positron emission tomography data is reconstructed based on the motion field transformation matrix and the single-phase CT image data, avoiding the situation where positron emission tomography images are affected by respiratory motion and resulting in artifacts, thereby effectively improving the diagnostic accuracy and protecting patients from additional radiation doses at the same time.

[0056] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of the present application more obvious and understandable, the following specifically describes the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present application. And throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0058] Figure 1 A flowchart of an image reconstruction method for positron emission tomography data provided by an embodiment of the present application is shown;

[0059] Figure 2 A flowchart of another image reconstruction method for positron emission tomography data provided by an embodiment of the present application is shown;

[0060] Figure 3 A flowchart of obtaining a motion field transformation matrix based on a deep neural network model provided by an embodiment of the present application is shown;

[0061] Figure 4 A flowchart of an image reconstruction calculation for positron emission tomography data provided by an embodiment of the present application is shown;

[0062] Figure 5 A flowchart of calculating a reference motion field transformation matrix provided by an embodiment of the present application is shown;

[0063] Figure 6 Shows the flowchart of data optimization based on a deep neural network model provided by an embodiment of the present application;

[0064] Figure 7 Shows the block diagram of a positron emission tomography data image reconstruction device provided by an embodiment of the present application;

[0065] Figure 8 Shows the structural schematic diagram of a terminal provided by an embodiment of the present application. Detailed implementation manners

[0066] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be completely conveyed to those skilled in the art.

[0067] Meanwhile, it should be understood that, for the sake of description, the sizes of the various parts shown in the drawings are not drawn in actual proportional relationship.

[0068] The following description of at least one exemplary embodiment is merely illustrative and in no way limits the present application and its application or use.

[0069] Well-known technologies, methods, and devices for those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the said technologies, methods, and devices should be regarded as part of the specification.

[0070] It should be noted that: like reference numerals and letters denote like items in the following drawings, and thus, once an item is defined in one drawing, it does not need to be further discussed in subsequent drawings.

[0071] Embodiments of the present application can be applied to a computer system / server, which can operate together with many other general-purpose or special-purpose computing system environments or configurations. Examples of well-known computing systems, environments, and / or configurations suitable for use with a computer system / server include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, minicomputer systems, mainframe computer systems, and distributed cloud computing technology environments including any of the above systems, and so on.

[0072] A computer system / server may be described in the general context of computer system-executable instructions, such as program modules, executed by the computer system. Generally, program modules may include routines, programs, target programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. The computer system / server may be implemented in a distributed cloud computing environment where tasks are executed by remote processing devices linked through a communication network. In a distributed cloud computing environment, program modules may be located on local or remote computing system storage media including storage devices.

[0073] An embodiment of the present application provides an image reconstruction method for positron emission tomography data, as Figure 1 shown, the method includes:

[0074] 101. Obtain single-phase CT image data and positron emission tomography data of multiple phases after phase division.

[0075] Under normal circumstances, when performing positron emission tomography according to the conventional clinical scanning duration, due to the influence of respiratory movement, artifacts will occur, resulting in a decrease in diagnostic accuracy. In the embodiment of the present application, as the current execution end, the positron emission tomography data of multiple phases obtained by dividing the user's conventional positron emission tomography data, and the single-phase CT image data of the conventional spiral CT scan are used as input content for image reconstruction of positron emission tomography data.

[0076] 102. Select a reference phase reconstruction image from multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and obtain a motion field transformation matrix matching the reference phase reconstruction image.

[0077] In the embodiment of the present application, the basic reconstruction images are obtained by performing image reconstruction without attenuation correction (NAC, non-attenuation corrections) on the positron emission tomography data of multiple phases after phase division. The reference phase reconstruction image is selected from multiple basic reconstruction images. For example, the basic reconstruction image with the largest similarity value to the single-phase CT image data is selected as the reference phase reconstruction image, or other phases may be selected as the reference phase reconstruction image, which is not specifically limited in the embodiment of the present application. The motion field transformation matrix is used to characterize the transformation relationship between the reference phase reconstruction image and the basic reconstruction image.

[0078] It should be noted that when image reconstruction is performed under the conventional scanning duration, due to the low statistics, it affects the selection of the reconstructed image of the reference phase and the accuracy of the obtained motion field transformation matrix. The forward projection data can be obtained by transforming and re-projecting the basic reconstructed images of multiple phases, and further comparing the forward projection data of multiple phases to update the motion field transformation matrix, so as to obtain a more matching motion field transformation matrix.

[0079] 103. Image reconstruction processing is performed based on the motion field transformation matrix, single-phase CT image data, and positron emission tomography data of multiple phases to obtain a reconstructed image of the positron emission tomography data.

[0080] In the embodiment of the present application, using the motion field transformation matrix obtained in step 102, the positron emission tomography data obtained in step 101, and the single-phase CT image data, a reconstructed image of the positron emission tomography data is obtained.

[0081] It should be noted that when reconstructing the image, the reconstructed image of the reference phase can be transformed into the basic reconstructed images of multiple phases based on the motion field transformation matrix to obtain the projection data of multiple phases (i.e., calculate the projection data), and then image reconstruction of the single-phase positron emission tomography data is performed based on the calculated projection data, the measured projection data (the measured projection data can be directly obtained during positron emission tomography), and the single-phase CT image data; it can also be to perform image reconstruction on multiple phases by transforming the single-phase CT image data based on the motion field transformation matrix, and then transform the reconstructed images of multiple phases to the reference phase and merge them to obtain the reconstructed image of the positron emission tomography data of the final reference phase. The embodiment of the present application does not make specific limitations.

[0082] For further illustration and limitation, the embodiment of the present application provides another method for image reconstruction of positron emission tomography data, as Figure 2 shown, the method includes:

[0083] 201. Construct a motion field transformation matrix based on the motion field information from the reconstructed image of the reference phase to the basic reconstructed images of each phase.

[0084] In the embodiment of the present application, exemplarily, the basic reconstructed image with the largest similarity value to the single-phase CT image data is selected as the reconstructed image of the reference phase. Specifically, the similarity between the basic reconstructed images of each phase and the single-phase CT image data can be measured, and the basic reconstructed image with the largest similarity value is selected as the reconstructed image of the reference phase. Further, based on the reconstructed image of the reference phase and multiple basic reconstructed images, a motion field transformation matrix is calculated. Specifically, let the reconstructed image of the reference phase be X 0 , then the basic reconstructed images of other phases are Xn , the motion field transformation matrix from the reconstructed image of other phases to the reconstructed image of the reference phase can be expressed as W n→0 , then the image transformation will satisfy the following relational expression:

[0085] X 0 = W n→0 X n , n = 1,... N - 1

[0086] Through the above formula, the initial motion field information of N - 1 reconstructed images transformed from other phase - based reconstructed images to the reference - phase reconstructed image can be obtained. Further, the motion field transformation matrix W from the reference - phase reconstructed image to other phase - based reconstructed images can be obtained 0→i .

[0087] 202. Based on the motion field transformation matrix, project the base reconstruction images of each phase orthogonally into the reference - phase reconstruction image to obtain the orthogonal projection data of the base reconstruction images.

[0088] In the embodiments of the present application, in order to obtain a more accurate motion field transformation matrix, first, based on the motion field transformation matrix constructed in step 201, project the base reconstruction images of each phase orthogonally into the reference - phase reconstruction image.

[0089] Specifically, the calculation formula for the orthogonal projection data is as follows:

[0090]

[0091] Among them, X represents the iterative image, Y' represents the transformed orthogonal projection data, P represents the system matrix, N represents the normalization factor, A represents the attenuation factor, S and R respectively represent the scattering factor and the random factor. At this time, A is set as a matrix of all 1s.

[0092] 203. Accumulate the orthogonal projection data of each phase to obtain the calculated orthogonal projection value of the reference - phase reconstruction image.

[0093] In the embodiments of the present application, specifically, the calculated orthogonal projection value Y 0,calculate of the reference - phase reconstruction image is calculated as follows:

[0094]

[0095]

[0096]

[0097] Among them, ω i represents the cumulative weight value of the orthogonal projection data of the current phase, satisfying that the cumulative value is 1, and usually taking the value of 1 / (N - 1).

[0098] 204a. If the difference between the calculated orthographic projection value and the measured orthographic projection value of the reconstructed image in the reference period is less than the preset difference threshold, determine that the motion field transformation matrix is the updated motion field transformation matrix.

[0099] In the embodiments of the present application, the motion field transformation matrix can be updated based on the objective function to minimize the difference between the calculated orthographic projection value and the measured orthographic projection value of the reconstructed image in the reference period. Further, the update can be completed by setting the number of iterative updates or the preset difference threshold as the standard. Specifically, the objective function is as follows:

[0100] F(n) = Norm(Y 0 - Y' n→0 ) + α n Φ(X 0 - W n→0 X n ), n = 1,..., N - 1

[0101] F(n) = Norm(Y 0 - Y 0,calculate ) + α N Φ(X 0 - X 0,calculate )

[0102] Among them, α represents the weight, which is used to control the influence of the second term. Norm represents taking the norm, and the L2 norm is usually used for calculation, which is used to represent the difference between the measured orthographic projection value of the reconstructed image in the reference period and the calculated orthographic projection value; the second term represents the function of calculating the difference between the normal reconstruction result of the reconstructed image in the reference period and the transformed calculation value, that is, while updating the motion field transformation matrix, it is still necessary to ensure that the transformation results of the basic reconstruction images in other periods and the reconstructed image in the reference period have the smallest difference.

[0103] 204b. If the difference between the calculated orthographic projection value and the measured orthographic projection value of the reconstructed image in the reference period is greater than or equal to the preset difference threshold, re - execute the update of the motion field transformation matrix.

[0104] The embodiments of the present application are the corresponding steps to step 204a. If the difference between the calculated orthographic projection value and the measured orthographic projection value of the reconstructed image in the reference period is greater than or equal to the preset difference threshold or the number of iterative updates does not reach the preset standard, repeat the update process of the motion field transformation matrix until the preset standard is reached.

[0105] Optionally, in a specific application scenario, as another implementation method of the above steps, the motion field transformation matrix can also be directly obtained from the projection data or the corresponding basic reconstruction images of multiple phases based on a deep neural network model that has completed model training. Specifically, such as Figure 3As shown below. First, obtain positron emission tomography (PET) scan data that is at least twice the duration of a conventional clinical scan, and calculate the target motion field transformation matrix. Second, use the PET scan data of the conventional clinical scan duration or the corresponding basic reconstructed image as the input item of the deep neural network model, and train to obtain the predicted motion field transformation matrix. Third, compare the predicted motion field transformation matrix with the target motion field transformation matrix until a preset standard is met to complete the model training and obtain the deep neural network model with completed model training. Finally, directly obtain the motion field transformation matrix from the projection data of multiple phases based on the deep neural network model with completed model training.

[0106] 205a. Based on the updated motion field transformation matrix, transform the reference phase reconstructed image into the basic reconstructed images of multiple phases to obtain the projection data of multiple phases.

[0107] In the embodiment of the present application, based on the motion field transformation matrix obtained in step 204a, the motion field transformation matrix W from the reference phase reconstructed image to other phases can be obtained accordingly. i→0 , specifically, the corresponding relationship between the projection data of each phase and the reference phase reconstructed image is as follows:

[0108] Y n = PX n = PW 0→n X 0 , n = 1,..., N - 1

[0109]

[0110] Among them, T represents the transpose, indicating that the projection data of different phases can be used to reconstruct the reference phase reconstructed image.

[0111] 206a. Based on the projection data of multiple phases, the measured projection data of multiple phases, and the single-phase CT image data, perform image reconstruction processing to obtain the reconstructed image of the positron emission tomography data.

[0112] In the embodiment of the present application, by way of example, taking the reference phase as an example, based on the above steps, the updated standard MLEM calculation formula is as follows:

[0113]

[0114] Through the above formula, the reconstructed image of the positron emission tomography data can be obtained according to the projection data of multiple phases, the measured projection data of multiple phases, and the corresponding single-phase CT image data.

[0115] Optionally, as another implementation method of steps 205a to 206a, in the embodiments of the present application, based on the updated motion field transformation matrix, the single-phase CT image data is transformed into the basic reconstruction images of multiple phases. The multi-phase CT image sequence is as follows:

[0116] XCT n =W 0→n XCT 0 , n = 1,... N-1

[0117] Wherein, XCT 0 represents the reference-phase CT image obtained by acquisition, and XCT n represents the CT images of other phases obtained by the above formula calculation. The reconstruction images of the corresponding phases reconstructed from the positron emission tomography data of multiple phases and the calculated CT images of multiple phases are respectively based on the updated motion field transformation matrix, and the reconstruction images of the positron emission tomography data of multiple phases are transformed to the reference phase and merged to obtain the reconstruction image of the positron emission tomography data of the final reference phase. It can be understood that due to the limited statistics of the positron emission tomography data of each phase, the reconstructed image may have an increase in noise, thereby affecting the quality of the reconstructed image. In order to improve the quality of the reconstructed image, weighted superposition processing can be performed according to the regions with relatively small changes in the corresponding image regions of the motion field transformation matrix to improve the signal-to-noise ratio of the image. Exemplarily, the reconstruction image of the positron emission tomography data of any phase is calculated as follows:

[0118]

[0119]

[0120]

[0121] m, n = 0,... N-1

[0122] Wherein, i represents the pixel point in the image, and W i represents the motion adjustment amount of the current pixel point i, and ω represents the weight factor. It should be noted that when the motion adjustment amount is greater than the preset threshold, it is not introduced into the calculation, and at the same time, the weight factor is normalized to ensure that the sum of the weights is 1.

[0123] For further illustration and limitation, in the embodiments of the present application, before obtaining the positron emission tomography data of multiple phases after phase division, the embodiment method further includes: obtaining the positron emission tomography data; performing phase division on the positron emission tomography data according to the respiratory gating signal to obtain the positron emission tomography data of multiple phases.

[0124] It should be noted that, in order to perform phase division on positron emission tomography (PET) data, it can be divided according to the respiratory gating signal. The respiratory gating signal can be obtained through an external gating device or calculated based on the data change rule. The embodiments of the present application do not make specific limitations. Further, the image data is divided according to the respiratory gating signal.

[0125] In a specific application scenario, the reconstruction calculation process of positron emission tomography data is as Figure 4 shown. First, a patient is subjected to a conventional clinical PET scan to obtain PET data and a conventional spiral CT scan to obtain a CT single-phase image. Secondly, the PET data is divided into phases according to the respiratory gating signal, and NAC reconstruction is performed to obtain the NAC image of the PET phase. Further, the motion is estimated between the NAC image of the PET phase and the PET reference phase obtained by comparing with the CT single-phase image to obtain the motion field information. Thirdly, the motion field information is updated and iterated with the PET phase forward projection data as the constraint condition. When the deviation is less than the preset threshold or the iteration times are reached, the final motion field information is obtained. Finally, image reconstruction is performed based on the final motion field information, the PET phase data, and the CT single-phase image to obtain a reconstructed image without artifacts.

[0126] Preferably, in order to make the obtained motion field transformation matrix more accurate, in the embodiments of the present application, a reference motion field transformation matrix is introduced according to the time-of-flight information. The method of the embodiment further includes: if the positron emission tomography data of multiple phases after phase division contains time-of-flight information, then based on the pseudo-back projection, the annihilation event position distribution map of each phase is obtained; based on the annihilation event position distribution map, the first motion field transformation matrix between the basic reconstruction images of each phase and the second motion field transformation matrix from the basic reconstruction images of each phase to the reference phase reconstruction image are obtained to obtain the reference motion field transformation matrix as the prior information for updating the motion field transformation matrix.

[0127] Specifically, the calculation process of the reference motion field transformation matrix is as Figure 5As described above, according to the positron emission tomography data of multiple divided phases, each phase data is processed separately, that is, for the detected annihilation events, the most likely annihilation position in three-dimensional space is calculated using the detected time-of-flight difference information. Usually, the position probability of each annihilation event on the line of response (LOR) is distributed according to a Gaussian distribution model, that is, the position with the largest probability distribution value is the occurrence position of the annihilation event. Further, the annihilation event position distribution maps corresponding to each phase are obtained. According to the annihilation event position distribution maps, the first motion transformation matrix between the basic reconstruction images of each phase, and the second motion field transformation matrix from the basic reconstruction images of each phase to the reconstruction image of the reference phase are obtained. Furthermore, a reference motion field transformation matrix is obtained to be used as prior information to regularize and constrain the motion transformation matrix obtained in the above steps. Based on this, the objective function can be updated as follows:

[0128]

[0129] Among them, β represents the weight used to control the influence of the third term. The third term represents a function of calculating the difference between the motion transformation matrix obtained from the reconstruction image and the motion transformation matrix obtained from the annihilation event position distribution map, that is, while updating the motion field transformation matrix, it is still necessary to ensure that the difference from the reference motion transformation matrix result is minimized.

[0130] Since multi-phase gated scans are performed under the conventional clinical scan duration, the projection data statistics under each phase are significantly reduced compared to the phase data statistics under the conventional gated scan duration, resulting in a decrease in the subsequent image reconstruction accuracy. To avoid the error influence caused by low data statistics, data optimization processing of the low-statistics projection data can be performed using a deep neural network model to increase the statistics of the projection data, thereby improving the confirmation of the subsequent reference phase and the calculation accuracy of the motion field transformation matrix. In the embodiments of the present application, after obtaining the positron emission tomography data of multiple phases after phase division, the method of the embodiment further includes: performing data optimization on the positron emission tomography data of multiple phases based on a deep neural network model that has completed model training. The deep neural network model is obtained by training based on a model training set including positron emission tomography data of a first duration and positron emission tomography data of a second duration, where the first duration is at least twice greater than the second duration.

[0131] Specifically, the process of data optimization based on the deep neural network model is as follows Figure 6As shown. First, a deep neural network model is constructed to obtain positron emission tomography data of a first duration, wherein the first duration is at least twice the duration of a conventional clinical scan, for example, the duration of a conventional clinical scan is 5 minutes, and the first duration is at least 10 minutes. The positron emission tomography data of a second duration is intercepted from the positron emission tomography data of the first duration, wherein the second duration is the duration of a conventional clinical scan, for example, 5 minutes of positron emission tomography data is intercepted from 10 minutes of positron emission tomography data. Based on the positron emission tomography data of the first duration and the positron emission tomography data of the second duration, a model training set is generated to train the deep neural network model until the error between the prediction result obtained based on the data after data optimization by the model and the prediction result of the positron emission tomography data of the second duration reaches a stable state, and a deep neural network model that has completed model training is obtained.

[0132] The present application provides an image reconstruction method for positron emission tomography data, which firstly obtains single-phase CT image data and positron emission tomography data of multiple phases after phase division; secondly, based on the single-phase CT image data, a reference phase reconstruction image is selected from multiple basic reconstruction images of the positron emission tomography data, and a motion field transformation matrix matching the reference phase reconstruction image is obtained; finally, image reconstruction processing is performed based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases to obtain a reconstructed image of the positron emission tomography data. Compared with the prior art, the embodiment of the present application obtains a motion field transformation matrix matching the multi-phase basic reconstruction image with the reference phase reconstruction image, and reconstructs the positron emission tomography data based on the motion field transformation matrix and the single-phase CT image data, thereby avoiding the situation where the positron emission tomography image is affected by respiratory motion and causes artifacts, thereby effectively improving the diagnostic accuracy and protecting the patient from additional radiation dose.

[0133] Furthermore, as a response to the above Figure 1 The present invention provides an image reconstruction device for positron emission tomography data, such as Figure 7 As shown, the device comprises:

[0134] A first acquisition module 31 , a second acquisition module 32 , and a reconstruction module 33 .

[0135] The first acquisition module 31 is used to acquire single-phase CT image data and positron emission tomography data of multiple phases after phase division;

[0136] A second acquisition module 32, configured to select a reference phase reconstruction image from multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and acquire a motion field transformation matrix matching the reference phase reconstruction image;

[0137] A reconstruction module 33, configured to perform image reconstruction processing based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases, to obtain a reconstruction image of the positron emission tomography data.

[0138] In a specific application scenario, the second acquisition module 32 specifically includes:

[0139] A construction unit, configured to construct a motion field transformation matrix based on the motion field information from the reference phase reconstruction image to the basic reconstruction images of each phase;

[0140] An update unit, configured to update the motion field transformation matrix through image transformation and projection processing, so as to perform motion correction based on the updated motion field transformation matrix.

[0141] In a specific application scenario, the update unit specifically includes:

[0142] A forward projection subunit, configured to project the basic reconstruction images of each phase onto the reference phase reconstruction image based on the motion field transformation matrix, to obtain forward projection data of the basic reconstruction images;

[0143] An accumulation subunit, configured to perform accumulation processing on the forward projection data of each phase, to obtain a calculated forward projection value of the reference phase reconstruction image;

[0144] A determination subunit, configured to determine that the motion field transformation matrix is the updated motion field transformation matrix if the difference between the calculated forward projection value and the measured forward projection value of the reference phase reconstruction image is less than a preset difference threshold;

[0145] An update subunit, configured to re-execute the update of the motion field transformation matrix if the difference between the calculated forward projection value and the measured forward projection value of the reference phase reconstruction image is greater than or equal to the preset difference threshold.

[0146] In a specific application scenario, the reconstruction module 33 specifically includes:

[0147] A forward projection unit, configured to transform the reference phase reconstruction image into the basic reconstruction images of multiple phases based on the updated motion field transformation matrix, to obtain projection data of multiple phases;

[0148] A reconstruction unit for performing image reconstruction processing based on the projection data of multiple phases, the measured projection data of multiple phases, and the single-phase CT image data to obtain a reconstructed image of the positron emission tomography data.

[0149] In a specific application scenario, before the first acquisition module 31, the device further includes:

[0150] A third acquisition module for acquiring positron emission tomography data;

[0151] A partitioning module for partitioning the positron emission tomography data into multiple phases according to a respiratory gating signal to obtain positron emission tomography data of multiple phases.

[0152] In a specific application scenario, the device further includes:

[0153] A fourth acquisition module for, if the positron emission tomography data of multiple phases after phase partitioning contains time-of-flight information, obtaining a distribution map of annihilation event positions for each phase based on pseudo-backprojection;

[0154] The fourth acquisition module is further configured to obtain a first motion field transformation matrix between the basic reconstruction images of each phase and a second motion field transformation matrix from the basic reconstruction images of each phase to the reference phase reconstruction image based on the distribution map of annihilation event positions, so as to obtain a reference motion field transformation matrix as prior information for updating the motion field transformation matrix.

[0155] In a specific application scenario, after the first acquisition module 31, the device further includes:

[0156] An optimization module for optimizing the positron emission tomography data of multiple phases based on a deep neural network model that has completed model training. The deep neural network model is obtained by training based on a model training set including positron emission tomography data of a first duration and positron emission tomography data of a second duration, where the first duration is at least twice greater than the second duration.

[0157] The present application provides an image reconstruction device for positron emission tomography data. First, single-phase CT image data and positron emission tomography data of multiple phases after phase division are acquired. Secondly, a reference phase reconstruction image is selected from multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and a motion field transformation matrix matching the reference phase reconstruction image is obtained. Finally, image reconstruction processing is performed based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases to obtain a reconstructed image of the positron emission tomography data. Compared with the prior art, in the embodiment of the present application, a motion field transformation matrix matching the multi-phase basic reconstruction image and the reference phase reconstruction image is obtained, and the positron emission tomography image data is reconstructed based on the motion field transformation matrix and the single-phase CT image data, avoiding the situation of artifacts in the positron emission tomography image caused by the influence of respiratory motion, thereby effectively improving the diagnostic accuracy and at the same time protecting the patient from additional radiation dose.

[0158] According to an embodiment of the present application, there is provided a storage medium storing at least one executable instruction, and the computer executable instruction can execute the image reconstruction method of positron emission tomography data in any of the above method embodiments.

[0159] Based on such understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various implementation scenarios of the present application.

[0160] Figure 8 FIG. shows a schematic structural diagram of a terminal according to an embodiment of the present application. The specific implementation of the terminal is not limited in the specific embodiments of the present application.

[0161] As Figure 8 shown, the terminal may include: a processor 402, a communication interface 404, a memory 406, and a communication bus 408.

[0162] Among them: the processor 402, the communication interface 404, and the memory 406 communicate with each other through the communication bus 408.

[0163] The communication interface 404 is used to communicate with network elements of other devices such as clients or other servers.

[0164] A processor 402 is configured to execute a program 410, and specifically, can execute the relevant steps in the embodiments of the above image reconstruction method for positron emission tomography data.

[0165] Specifically, the program 410 may include program code, and the program code includes computer operation instructions.

[0166] The processor 402 may be a central processing unit (CPU), or a specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application. One or more processors included in the terminal may be of the same type, such as one or more CPUs; or may be of different types, such as one or more CPUs and one or more ASICs.

[0167] A memory 406 is configured to store the program 410. The memory 406 may include a high-speed RAM memory, and may also include non-volatile memory, such as at least one disk memory.

[0168] The program 410 is specifically configured to cause the processor 402 to perform the following operations:

[0169] Obtain single-phase CT image data and positron emission tomography data of multiple phases after phase division;

[0170] Select a reference phase reconstruction image from multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and obtain a motion field transformation matrix matching the reference phase reconstruction image;

[0171] Perform image reconstruction processing based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases to obtain a reconstructed image of the positron emission tomography data.

[0172] The storage medium may further include an operating system and a network communication module. The operating system is a program for managing the hardware and software resources of the physical device for reconstructing the above positron emission tomography image data, and supports the operation of information processing programs and other software and / or programs. The network communication module is used to implement communication between components inside the storage medium, and communication with other hardware and software in the information processing physical device.

[0173] In each of the embodiments described in this specification, a progressive approach is adopted. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the embodiments, reference can be made to each other. For system embodiments, since they basically correspond to method embodiments, the description is relatively simple. For the relevant parts, reference can be made to the corresponding description in the method embodiments.

[0174] The methods and systems of the present application can be implemented in many ways. For example, the methods and systems of the present application can be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of the steps for the method is only for illustration. The steps of the method of the present application are not limited to the specific order described above, unless otherwise specifically stated. In addition, in some embodiments, the present application can also be implemented as a program recorded in a recording medium, and these programs include machine-readable instructions for implementing the method according to the present application. Therefore, the present application also covers a recording medium storing a program for executing the method according to the present application.

[0175] Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present application can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. Optionally, they can be implemented by program codes executable by the computing device. Thus, they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described can be executed in a different order from here, or they can be separately made into individual integrated circuit modules, or multiple modules or steps among them can be made into a single integrated circuit module for implementation. In this way, the present application is not limited to any specific combination of hardware and software.

[0176] The above are only the preferred embodiments of the present application and are not used to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. An image reconstruction method for positron emission tomography data, characterized in that, it includes: Obtaining single-phase CT image data and positron emission tomography data of multiple phases after phase division; Selecting a reference phase reconstruction image from multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and obtaining a motion field transformation matrix matching the reference phase reconstruction image; Performing image reconstruction processing based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases to obtain a reconstructed image of the positron emission tomography data; Among them, the obtaining of the motion field transformation matrix matching the reference phase reconstruction image specifically includes: Constructing a motion field transformation matrix based on the motion field information between the reference phase reconstruction image and the basic reconstruction images of each phase; Based on the motion field transformation matrix, orthogonally projecting the basic reconstruction images of each phase into the reference phase reconstruction image to obtain the orthogonal projection data of the basic reconstruction images; Accumulating and processing the orthogonal projection data of each phase to obtain the calculated orthogonal projection value of the reference phase reconstruction image; If the difference between the calculated orthogonal projection value and the measured orthogonal projection value of the reference phase reconstruction image is less than a preset difference threshold, determine the motion field transformation matrix as the updated motion field transformation matrix; If the difference between the calculated orthogonal projection value and the measured orthogonal projection value of the reference phase reconstruction image is greater than or equal to the preset difference threshold, re-execute the update of the motion field transformation matrix to perform motion correction based on the updated motion field transformation matrix; The performing image reconstruction processing based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases specifically includes: Based on the updated motion field transformation matrix, transforming the reference phase reconstruction image into the basic reconstruction images of multiple phases to obtain projection data of multiple phases; Performing image reconstruction processing based on the projection data of multiple phases, the measured projection data of multiple phases, and the single-phase CT image data to obtain a reconstructed image of the positron emission tomography data.

2. The method according to claim 1, characterized in that, Before obtaining the positron emission tomography data of multiple phases after phase division, the method further includes: Obtaining positron emission tomography data; Performing phase division on the positron emission tomography data according to a respiratory gating signal to obtain positron emission tomography data of multiple phases.

3. The method according to claim 1, characterized in that, The method further includes: If the positron emission tomography data of multiple phases after phase division contains time-of-flight information, obtaining a distribution map of annihilation event positions of each phase based on pseudo-backprojection; Based on the distribution map of the annihilation event positions, obtain a first motion field transformation matrix between the basic reconstruction images of each phase, and a second motion field transformation matrix from the basic reconstruction images of each phase to the reference phase reconstruction image, to obtain a reference motion field transformation matrix as prior information for updating the motion field transformation matrix.

4. The method according to claim 1, wherein, after obtaining the positron emission tomography data of multiple phases after phase division, the method further includes: optimizing the positron emission tomography data of multiple phases based on a deep neural network model that has completed model training, the deep neural network model being obtained by training based on a model training set including positron emission tomography data of a first duration and positron emission tomography data of a second duration, wherein the first duration is at least twice greater than the second duration.

5. An image reconstruction device for positron emission tomography data, wherein, it includes: a first acquisition module, configured to acquire single-phase CT image data and positron emission tomography data of multiple phases after phase division; a second acquisition module, configured to select a reference phase reconstruction image from the multiple basic reconstruction images of the positron emission tomography data based on the single-phase CT image data, and acquire a motion field transformation matrix matching the reference phase reconstruction image; a reconstruction module, configured to perform image reconstruction processing based on the motion field transformation matrix, the single-phase CT image data, and the positron emission tomography data of multiple phases to obtain a reconstruction image of the positron emission tomography data; the second acquisition module specifically includes: a construction unit, configured to construct a motion field transformation matrix based on the motion field information between the reference phase reconstruction image and the basic reconstruction images of each phase; an update unit, configured to update the motion field transformation matrix through image transformation and projection processing to perform motion correction based on the updated motion field transformation matrix; the update unit specifically includes: a forward projection subunit, configured to forward project the basic reconstruction images of each phase into the reference phase reconstruction image based on the motion field transformation matrix to obtain the forward projection data of the basic reconstruction images; an accumulation subunit, configured to perform accumulation processing on the forward projection data of each phase to obtain a calculated forward projection value of the reference phase reconstruction image; a determination subunit, configured to determine that the motion field transformation matrix is the updated motion field transformation matrix if the difference between the calculated forward projection value and the measured forward projection value of the reference phase reconstruction image is less than a preset difference threshold; an update subunit, configured to re-execute the update of the motion field transformation matrix if the difference between the calculated forward projection value and the measured forward projection value of the reference phase reconstruction image is greater than or equal to the preset difference threshold; the reconstruction module specifically includes: A forward projection unit, configured to transform the reference phase reconstruction image into the base reconstruction images of multiple phases based on the updated motion field transformation matrix, so as to obtain projection data of multiple phases; A reconstruction unit, configured to perform image reconstruction processing based on the projection data of multiple phases, the measured projection data of multiple phases, and the single-phase CT image data, so as to obtain a reconstructed image of the positron emission tomography data.

6. A storage medium, in which at least one executable instruction is stored, and the executable instruction causes a processor to perform operations corresponding to the image reconstruction method of the positron emission tomography data according to any one of claims 1-4.

7. An electronic device, comprising: a processor, a memory, a communication interface, and a communication bus, where the processor, the memory, and the communication interface complete communication with each other through the communication bus; the memory is configured to store at least one executable instruction, and the executable instruction causes the processor to perform operations corresponding to the image reconstruction method of the positron emission tomography data according to any one of claims 1-4.

Images