Image processing method, device, medium and equipment

Abstract

The present invention provides an image processing method and apparatus, medium, and equipment, which obtains PET scanning data of multiple time frames for a scanned object, uses a static PET reconstruction method to reconstruct static PET images on the PET scanning data, obtains multiple static PET images, combines the static PET images into a dynamic PET image sequence, divides the image region according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, obtains multiple target image regions, obtains the mean value of the radioactivity corresponding to each target image region, and uses the radioactivity value to perform parameter reconstruction on the PET scanning data of the scanned object, and obtains a parameter image for characterizing the metabolic rate in the body of the scanned object. Through the present invention, secondary propagation of noise is avoided, a parameter image with higher quality and more in line with the actual metabolic law is obtained, and the imaging efficiency of the parameter image is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of positron emission tomography technology, in particular to an image processing method and device, medium and equipment. Background Art

[0002] PET / CT (Positron Emission Tomography / Computed Tomography) is positron emission tomography / computerized tomography, which is a medical imaging device that combines the two medical imaging technologies of PET and CT.

[0003] Parametric reconstruction is a new PET image reconstruction method. The generated image is a parametric image, which is an indicator that quantitatively represents the human metabolic rate (the traditional PET image is the SUV value, which refers to the ratio of the local tissue tracer radioactivity to the systemic average injected activity, and is a semi-quantitative indicator). Parameter reconstruction uses the concentration of radioactive activity in plasma as the input function and uses multiple scan data to calculate the dynamic parameter image associated with the drug. The existing parameter reconstruction methods are divided into two types: indirect method and direct method. The indirect method first reconstructs the activity SUV image of each bed and each scan, and then uses the activity image combined with the pharmacokinetic model to reconstruct the parameters; the direct method directly uses the scanned chord diagram data and the pharmacokinetic model for reconstruction.

[0004] In the process of PET / CT parameter reconstruction, the following problems may exist: improper method selection. The use of indirect method for parameter reconstruction is prone to secondary propagation of noise, thus affecting image quality; insufficient utilization of spatiotemporal information, for example, only considering spatial filtering without considering the influence of temporal information; insufficient utilization of computing resources. Parameter reconstruction is performed only after all scans are completed. The reconstruction machine is idle during the scanning process, wasting time and resources, etc. Summary of the invention

[0005] In view of the above problems, the present invention proposes an image processing method, device, medium, and equipment. The image processing method provided by the present invention performs parameter reconstruction by asynchronously performing static PET reconstruction and scanning, using a direct method, and introducing spatiotemporal prior information, which not only avoids the secondary propagation interference of noise in the projection data on the parameter image, but also adds spatiotemporal prior information to the iterative solution process of generating the parameter image, so that the reconstructed image quality is higher, and it is a high-speed, high-quality parameter image generation method.

[0006] According to a first aspect of the present invention, there is provided an image processing method, comprising:

[0007] Acquire PET scan data of multiple time frames for the scanned object, and reconstruct static PET images on the PET scan data using a static PET reconstruction method to obtain multiple static PET images, wherein each voxel point in each static PET image corresponds to a radioactivity;

[0008] Combining the plurality of static PET images into a dynamic PET image sequence, dividing the dynamic PET image sequence into image regions according to the temporal variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, to obtain a plurality of target image regions;

[0009] Obtaining the mean radioactivity value of the static PET image corresponding to each target image region as the radioactivity value corresponding to each voxel point in the target image region;

[0010] The radioactivity value is used to perform parameter reconstruction on the PET scan data of the scanned object to obtain a parameter image for characterizing the metabolic rate in the body of the scanned object.

[0011] Optionally, before performing a PET scan on the scanned object, the method further includes:

[0012] Performing a CT scan on the scanned object to obtain a CT image of a corresponding scanned part that is the same as the scanned part corresponding to the PET image sequence;

[0013] The static PET image is reconstructed using the CT image to perform attenuation correction.

[0014] Optionally, dividing the dynamic PET image sequence into image regions according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length to obtain the target image region includes:

[0015] Clustering the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within the preset time length to obtain the voxel point class of the scanned object;

[0016] Using the voxel point class to divide the dynamic PET image sequence into image regions; wherein the same voxel point class corresponds to the same image region;

[0017] The image region is divided and adjusted according to the connected domain information of the CT image to obtain the target image region.

[0018] Optionally, before performing parameter reconstruction on the PET scan data of the scanned object using the radioactivity value, the method further includes:

[0019] For the heart region of the scanned object, clustering is performed according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within the preset time length to obtain a plasma voxel point class for characterizing the plasma in the scanned object;

[0020] Function fitting is performed according to the time variation characteristics of the radioactivity corresponding to each voxel point in the plasma voxel point class within the preset time length to obtain the plasma input function.

[0021] Optionally, the performing parameter reconstruction on the PET scan data of the scanned object includes:

[0022] The PET scan data is parameter reconstructed using an iterative reconstruction method based on the radioactivity value and the plasma input function, specifically comprising:

[0023] The PET scan data is iteratively reconstructed using spatiotemporal prior parameters, which are calculated using the following formula:

[0024]

[0025] Wherein, R is a time-space prior parameter; β s is the spatial prior parameter; β t is the time prior parameter; γ is the empirical parameter of scanning time; N is the number of image voxel points; N i is the neighborhood of voxel point i; w ij represents the inverse of the distance between voxel point i and voxel point j; T represents the number of scanning passes; w t is the time prior weight of voxel point i; is the radioactivity value of voxel i at time point t; and

[0026] The K image and V image in the parameter image are updated using the following formula:

[0027]

[0028]

[0029] in, Used to characterize the FDG value captured or phosphorylated by cells in the human body; Used to characterize the free FDG value in the human body; N is the number of image voxels; C p (n) is the value of the plasma input function at n; S p (n) is the integral value of the plasma input function from 0 to n; This is a static PET image.

[0030] Optionally, the adopting a static PET reconstruction method to perform static PET image reconstruction on the PET scan data comprises:

[0031] Reconstructing static PET images of the PET scan data using an iterative reconstruction method, and saving a plurality of correction factors corresponding to the reconstruction of each static PET image; the correction factors are correction data for adjusting the deviation of the scan data caused by external factors during the scanning process;

[0032] When the radioactivity value is used to perform parameter reconstruction on the PET scan data of the scanned object, the correction factor is loaded, and the correction factor is used to perform parameter reconstruction on the PET scan data of the scanned object.

[0033] Optionally, acquiring PET scan data of a plurality of time frames of the scanned object and reconstructing a static PET image of the PET scan data using a static PET reconstruction method comprises:

[0034] Performing PET scans of multiple time frames on the scanned object to obtain multiple pieces of PET scan data; wherein the PET scan uses a step-by-step scan or a continuous bed-moving scan;

[0035] A static PET reconstruction method is used to perform static PET image reconstruction on each PET scan data. The static PET image reconstruction is performed asynchronously with the PET scan, including: each time a PET scan data is acquired, a static PET image reconstruction is performed on the PET scan data.

[0036] According to a second aspect of the present invention, an image processing device is provided, comprising:

[0037] A static image reconstruction module, used to obtain PET scan data of multiple time frames for the scanned object, and to perform static PET image reconstruction on the PET scan data using a static PET reconstruction method to obtain multiple static PET images, wherein each voxel point in each of the static PET images corresponds to a radioactivity;

[0038] A dynamic image synthesis module, used for combining the plurality of static PET images into a dynamic PET image sequence, dividing the dynamic PET image sequence into image regions according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, and obtaining a plurality of target image regions;

[0039] An image parameter acquisition module, used to acquire the mean value of the radioactivity of the static PET image corresponding to each target image region as the radioactivity value corresponding to each voxel point in the target image region;

[0040] The parameter image generation module is used to perform parameter reconstruction on the PET scan data of the scanned object using the radioactivity value to obtain a parameter image for characterizing the metabolic rate in the body of the scanned object.

[0041] According to a third aspect of the present invention, a computer-readable storage medium is provided, on which a computer program is stored, characterized in that when the computer program is executed by a processor, the steps of the image processing method as described in any one of the first aspect of the present invention are implemented.

[0042] According to a fourth aspect of the present invention, a computer device is proposed, 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 image processing method described in any one of the first aspects of the present invention when executing the computer program.

[0043] The present invention provides an image processing method, device, medium, and equipment, which obtains PET scanning data of multiple time frames for a scanned object, reconstructs static PET images of the PET scanning data using a static PET reconstruction method, obtains multiple static PET images, combines the static PET images into a dynamic PET image sequence, divides the dynamic PET image sequence into image regions according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, obtains multiple target image regions, obtains the mean radioactivity of the static PET image corresponding to each target image region, and uses the radioactivity value corresponding to each voxel point in the target image region to perform parameter reconstruction on the PET scanning data of the scanned object, and obtains a parameter image for characterizing the metabolic rate in the body of the scanned object. The image processing method provided by the present invention uses a direct method for parameter reconstruction, avoids secondary propagation of noise, and adds spatiotemporal prior information, so that a parameter image with higher quality and more in line with actual metabolic laws can be obtained, and PET scanning and static PET image reconstruction can also be set to be asynchronously executed, further improving the efficiency of parameter image imaging, and is an efficient, high-speed, and high-quality parameter image imaging method.

[0044] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below.

[0045] Based on the following detailed description of specific embodiments of the present invention in conjunction with the accompanying drawings, those skilled in the art will become more aware of the above and other objects, advantages and features of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0047] Figure 1 A schematic diagram showing a flow chart of an image processing method provided by an embodiment of the present invention;

[0048] Figure 2a A schematic diagram of a process of an iterative reconstruction method provided by an embodiment of the present invention is shown;

[0049] Figure 2b A schematic diagram showing a flow chart of an iterative reconstruction method provided by another embodiment of the present invention;

[0050] Figure 3 A schematic diagram showing the structure of an image processing device provided by an embodiment of the present invention is shown;

[0051] Figure 4 A schematic diagram showing the structure of an image processing device provided by another embodiment of the present invention is shown;

[0052] Figure 5 A schematic diagram of the physical structure of a computer device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0053] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present invention and to enable the scope of the present invention to be fully communicated to those skilled in the art.

[0054] PET / CT is an imaging device that combines PET (Positron Emission Computed Tomography) and CT (Computed Tomography). Among them, PET receives a pair of gamma photons in opposite directions generated by the positrons generated during the decay of radionuclides injected into the human body and annihilates them with free electrons in the human body. After a series of processing, it can infer the distribution of tracers in the human body and reflect the physiological, pathological, biochemical and metabolic changes of human tissues at the molecular level. CT uses the different absorption capabilities of various tissues in the human body to obtain an image matrix through the attenuation of X-rays in the human body and reconstruction calculations, and has a high resolution for tissue density. PET is used to reflect the physiological uptake of tissues, and CT is used to reflect the structure of organs.

[0055] However, each pixel of the PET image generated by PET / CT represents the Standard Uptake Value (SUV), which is the ratio of the local tissue tracer radioactivity to the whole body radioactivity. In order to intuitively view the human metabolic rate, it is usually necessary to perform parameter reconstruction on the PET image to obtain a parameter image used to represent the human metabolic rate index.

[0056] In view of this, an embodiment of the present invention provides an image processing method to achieve efficient and high-quality parameter reconstruction. Figure 1 As shown, the method may at least include the following steps S101 to S104.

[0057] Step S101, acquiring PET scan data of multiple time frames for the scanned object, and reconstructing static PET images on the PET scan data using a static PET reconstruction method to obtain multiple static PET images, wherein each voxel point in each static PET image corresponds to a radioactive activity.

[0058] In practical applications, scanning parameters for obtaining PET scan data can be set according to different scanning requirements, and a PET scan can be performed on a target scanning part of the scanned object to obtain initial PET scan data. For each time frame of PET scan data, a static PET image can be reconstructed using a static PET reconstruction method, such as an analytical method, an iterative method, etc., to obtain multiple static PET images corresponding to multiple time frames.

[0059] It should be understood that each voxel point in each static PET image corresponds to a radioactivity. Among them, the voxel point in the static PET image is equivalent to the voxel point in the two-dimensional image, which is the smallest representation unit of the static PET image. The radioactivity corresponding to the voxel point is used to characterize the radioactivity at the position of the voxel point. The static PET image may include plasma voxel points used to characterize the plasma in the scanned object and tissue voxel points used to characterize the tissue in the scanned object.

[0060] Optionally, PET scan data of multiple time frames for the scanned object are acquired, and a static PET image is reconstructed from the PET scan data using a static PET reconstruction method. Specifically, PET scans of multiple time frames may be performed on the scanned object to acquire multiple pieces of PET scan data; wherein the PET scan uses a step-by-step scan or a continuous bed-moving scan; and the static PET image reconstruction is performed asynchronously with the PET scan.

[0061] In the embodiment of the present invention, in order to improve the parameter reconstruction efficiency as much as possible, the static PET image reconstruction can be performed asynchronously with the PET scan, that is, each time a piece of PET scan data is obtained, the static PET image reconstruction is performed on the PET scan data, and there is no need to wait until the PET scan data of all time frames are obtained before performing the static PET image reconstruction. The static PET image reconstruction does not occupy the scanning time, so as to avoid the situation where the reconstruction machine is in an idle state, resulting in a waste of computing resources.

[0062] Optionally, before performing a PET scan on the scanned object, a CT scan may be performed on the scanned object to obtain a CT image of the same scanned part as the scanned part of the PET image sequence, and the CT image may be used to perform attenuation correction on static PET image reconstruction.

[0063] In the embodiment of the present invention, a three-dimensional electronic computer tomography CT image of the scanned object is obtained, and the scanning part corresponding to the CT image is the same as the scanning part corresponding to the static PET image. It should be understood that the CT image can accurately reflect the structure of human organs, so in the process of reconstructing the static PET image from the PET scan data, the CT image can be used to perform attenuation correction on the static PET image reconstruction process.

[0064] Attenuation correction refers to compensating for the loss of counts caused by the interaction of photons with the medium before they reach the detector during the reconstruction of static PET images. Photons travel through the medium before reaching the detector, and are lost or attenuated due to reflection, scattering or absorption due to interaction with the medium. Therefore, this part of the loss needs to be compensated to obtain a true nuclide distribution map. The correction method is to measure the absorption coefficient of the medium to photons and calculate it based on the attenuation law. In the actual static PET image reconstruction process, the density of each voxel point, that is, the absorption coefficient of photons, is obtained using CT images to compensate for the loss of counts caused by attenuation.

[0065] Optionally, a static PET reconstruction method is used to reconstruct a static PET image of the PET scan data. Specifically, an iterative reconstruction method can be used to reconstruct a static PET image of the PET scan data, and multiple correction factors corresponding to the reconstruction of each static PET image are saved.

[0066] In the embodiment of the present invention, the following can be used: Figure 2a The iterative reconstruction method shown in the figure reconstructs static PET images from PET scan data. Figure 2a The iterative reconstruction method shown is an iterative method based on the Ordered Subset Expectation Maximization (OSEM) algorithm. Each iterative step of the OSEM algorithm includes forward projection, data comparison, back projection, and image update steps until the iteration stop condition is met, thereby reconstructing a static PET image for the PET scan data of each time frame to obtain a static PET image corresponding to each time frame.

[0067] In the process of static PET image reconstruction, correction factors for static PET image reconstruction of each PET scan data can also be saved. The correction factors are correction data for adjusting the deviation of the scan data caused by external factors during the scanning process. In the PET scanning process, due to a series of interactions between photons and the surrounding environment, the time and space resolution limitations of the equipment, system design and other factors, the scan data obtained has certain deviations. In order to eliminate these deviations, a series of corrections are usually required, including: random correction, regularization correction, count loss correction, decay correction, attenuation correction, scattering correction, etc. In the above multiple correction process, multiple correction factors will be generated, and for the scan data corresponding to each time frame, the correction factors used in the static PET image reconstruction and parameter reconstruction processes are the same. In order to avoid repeated calculations, these correction factors can be saved in the static PET image reconstruction process, and directly loaded when the parameter reconstruction is performed later, eliminating the repeated calculation process.

[0068] Step S102, combining multiple static PET images into a dynamic PET image sequence, dividing the dynamic PET image sequence into image regions according to the temporal variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, and obtaining multiple target image regions.

[0069] In an embodiment of the present invention, a dynamic PET image sequence may include multiple static PET images obtained by scanning the same part of the scanned object at different scanning times within a preset time length. That is, each static PET image shows the same scanned part, but due to different scanning times, the distribution of the tracer in the scanned part in each static PET image is different, that is, the radioactivity corresponding to each static PET image is different. The multiple static PET images are sorted and combined according to the chronological order to obtain a dynamic PET image sequence for characterizing the time variation characteristics of the radioactivity corresponding to the voxel point within the preset time length.

[0070] Optionally, the dynamic PET image sequence is divided into image regions according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length to obtain the target image region. Specifically, the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length can be clustered to obtain the voxel point class of the scanned object; the dynamic PET image sequence is divided into image regions using the voxel point class.

[0071] In the embodiment of the present invention, in order to introduce time prior information in the parameter reconstruction process, the dynamic PET image sequence can be divided into image regions, which is actually to combine voxel points with similar metabolic rates and close positions into regions. It should be understood that each tissue in the human body has a metabolic rate, which is specifically manifested as a time activity curve (TAC) on the dynamic PET image sequence, but the TAC of a single voxel point will be affected by noise. By combining points with similar TAC curves and taking the mean radioactivity of the image region as the radioactivity value of each voxel point, noise can be effectively eliminated.

[0072] Specifically, the image region division is performed on the dynamic PET image sequence, and each voxel point in the static PET image can be used as a sample point, and the radioactivity value of each voxel point scanned in each time frame can be used as a feature point for cluster analysis, and all voxel points are divided into multiple regions, that is, (x, y, z) can be used to represent the coordinate position of each voxel point, and the voxel points with similar time variation characteristics of radioactivity within a preset time length and similar positions are clustered to generate multiple voxel point classes, and the voxel point class can represent the clustering result of each voxel point, and the same voxel point class is divided into the same image region. In the embodiment of the present invention, the cluster analysis can use clustering algorithms such as k-means clustering and k-nearest neighbor method, and the present invention does not limit this.

[0073] Optionally, after the image region is obtained by division, the image region division adjustment may be performed on the image region according to the connected domain information of the CT image to obtain the target image region.

[0074] In the embodiment of the present invention, the CT image may use the CT image of the same scanning part corresponding to the PET image sequence obtained before the above step S101. Since the CT image can accurately reflect the structure of human organs, the obtained image area can be re-divided according to the connected domain information corresponding to the coordinate position information of each voxel point. Specifically, the same image area can be divided into different areas, such as two different lungs or two kidneys, and different image areas can also be merged according to noise points or lesion points in tissues and organs. The re-divided and adjusted image area is used as the target image area.

[0075] Step S103 : obtaining the mean radioactivity value of the static PET image corresponding to each target image region as the radioactivity value corresponding to each voxel point in the target image region.

[0076] In the embodiment of the present invention, the real radioactivity value of each voxel point is used as the time prior information, that is, in order to obtain the radioactivity value with effective noise elimination, the average radioactivity value corresponding to multiple voxel points in the same target image area is used as the radioactivity value of each voxel point, and the radioactivity value of each voxel point is saved. When performing parameter reconstruction, the radioactivity value corresponding to the voxel point corresponding to each position information in each time frame can be directly extracted.

[0077] Step S104: Parameter reconstruction is performed on the PET scan data of the scanned object using the radioactivity value to obtain a parameter image for characterizing the metabolic rate in the scanned object.

[0078] In the embodiment of the present invention, the radioactivity value corresponding to the voxel point is used as time prior information and introduced into the iterative process of parameter reconstruction. In the parameter reconstruction process, the radioactivity in plasma is usually used as the input function, but the change characteristics of the radioactivity in plasma over time are significantly different from those of the radioactivity in tissues over time.

[0079] Therefore, before performing parameter reconstruction, it is also necessary to cluster the heart area of ​​the scanned object according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, so as to obtain a plasma voxel point class for characterizing the plasma in the scanned object; and perform function fitting according to the time variation characteristics of the radioactivity corresponding to each voxel point in the plasma voxel point class within a preset time length to obtain a plasma input function.

[0080] Therefore, if it is necessary to obtain a dynamic PET image sequence of the heart region of the scanned object, the same clustering analysis algorithm as in step S103 can be used to cluster the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence of the heart region within a preset time length, and obtain a voxel point class in the static PET image for characterizing the plasma in the scanned object, and then a function fitting can be performed based on the time variation characteristics of the radioactivity of each voxel point in the plasma voxel point class within a preset time length to obtain a plasma input function. It should be understood that the disclosed embodiment does not limit the function used for fitting.

[0081] In actual operation, the plasma input function can also be obtained by artificially pre-marking the aorta or heart area and measuring the activity change curve of the aorta or heart blood in the dynamic PET image sequence.

[0082] Optionally, when the parameters of the PET scan data of the scanned object are reconstructed using the radioactivity value, the correction factor mentioned in step S101 may be directly loaded, and the parameters of the PET scan data of the scanned object may be reconstructed using the correction factor.

[0083] The correction factors used in static PET image reconstruction and parameter reconstruction are the same. To avoid repeated calculations, these correction factors can be saved in the static PET image reconstruction process and directly loaded during parameter reconstruction to avoid repeated calculations.

[0084] Optionally, the parameter reconstruction of the PET scan data of the scanned object is substantially to reconstruct the parameters of the PET scan data using an iterative reconstruction method based on the radioactivity value and the plasma input function, specifically including:

[0085] The PET scan data are iteratively reconstructed using the spatiotemporal prior parameters, which are calculated using the following formula:

[0086]

[0087] Among them, R is the time-space prior parameter; β s is the spatial prior parameter; β t is the time prior parameter; γ is the empirical parameter of scanning time; N is the number of image voxel points; N i is the neighborhood of voxel point i; w ij represents the inverse of the distance between voxel point i and voxel point j; T represents the number of scanning passes; w t is the time prior weight of voxel point i; is the radioactivity value of voxel i at time point t;

[0088] In the embodiment of the present invention, the above-mentioned spatiotemporal prior parameters can be introduced to iteratively reconstruct the PET scan data, which can be specifically used as follows: Figure 2b The iterative reconstruction method shown reconstructs multiple target static PET images from PET scan data. Figure 2b The iterative reconstruction method shown is an iterative method based on the Nested Ordered Subset Expectation Maximization (Nested-OSEM) algorithm. The Nested Ordered Subset Expectation Maximization is also often referred to as the 4DOSEM algorithm. In the algorithm, each bed undergoes PET scans for multiple time frames, and each iterative step includes updating the activity image of the PET scan data corresponding to all time frames, wherein the activity image update of the PET scan data corresponding to each time frame uses the same method as each iteration in the OSEM algorithm. The parameter image update step is embedded in the OSEM iteration process and is called Nested-OSEM. Among them, in addition to outputting parameter images, 4DOSEM reconstruction can also output target static PET images and target dynamic PET image sequences for each time frame at the same time. Among them, the target static PET image, the target dynamic PET image sequence and the parameter image are all images reconstructed by introducing spatiotemporal prior information for parameter iteration. Through the Nested-OSEM algorithm, the generated target static PET image has less noise, and the change law of each voxel point of the target dynamic PET image sequence is more consistent with the real change law of the tissue to which it belongs, thereby obtaining a higher quality parameter image.

[0089] Then use the following formula to update the K image and V image in the parameter image:

[0090]

[0091]

[0092] in, Used to characterize the FDG value captured or phosphorylated by cells in the human body; Used to characterize the free FDG value in the human body; N is the number of image voxels; C p (n) is the value of the plasma input function at n; S p (n) is the integral value of the plasma input function from 0 to n; This is a static PET image.

[0093] The parameter reconstruction in the embodiment of the present invention is based on the dynamic compartment model. The compartment model divides the body into one or more independent units (tissues with the same drug transport rate are divided into the same compartment), which can simulate the absorption, distribution, elimination and other characteristics of drugs in the body. According to the complexity of the model, it is often divided into one-compartment model, two-compartment model and three-compartment model. Figure 1 The three-compartment model is shown, in which k1 represents the rate at which FDG (Fludeoxyglucose) enters the tissue from the blood, k2 represents the rate at which FDG returns from the tissue to the blood, k3 represents the rate at which FDG in the tissue is phosphorylated to FDG-6-P under the catalysis of hexose phosphokinase, and k4 represents the rate at which FDG-6-P is converted back to FDG under the catalysis of glucose-6-phosphatase. k1, k2, k3, and k4 are called individual rate constants or microparameters, and their dimensions are the inverse of the time dimension. In the process of parameter reconstruction, the radioactivity in plasma is usually used as the input function, and k3 or a combination of k3 and other rate constants is used as the parameter image.

[0094] Among them, the K image formula is It indicates the FDG captured or phosphorylated by cells in the human body. The unit is ml / g / min, which indicates the absolute value of the tracer captured by the moving volume of tissue per minute. The V image formula is It represents the free FDG in the human body, and the unit is ml / g, which represents the ratio of the free tracer in the region of interest to the tracer in the blood pool.

[0095] From the above formula, we can see that the update of K image and V image is only related to C p (n), S p (n) and related( represents a static PET image), after obtaining the input function, C p (n), S p (n) are all constants involved in the iterative process of the parameter image. Based on the update of the K image and the V image, the parameter image generated by the final reconstruction is obtained. Since the static PET image has lower noise and the voxel change is more in line with the actual metabolic law, the parameter image obtained in the embodiment of the present invention has higher quality. And in the iterative process, based on the introduction of the spatiotemporal prior, the image quality of the target static PET image, the target dynamic PET image sequence and the parameter image outputted in the final output are all improved.

[0096] Optionally, the present invention can also save, output and search CT images, target static PET images, target dynamic PET image sequences and parameter images.

[0097] The image processing method provided by the embodiment of the present invention obtains PET scanning data of multiple time frames for the scanned object, uses a static PET reconstruction method to reconstruct static PET images on the PET scanning data, obtains multiple static PET images, combines the static PET images into a dynamic PET image sequence, divides the dynamic PET image sequence into image regions according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, obtains multiple target image regions, obtains the mean radioactivity of the static PET image corresponding to each target image region, and uses the radioactivity value to perform parameter reconstruction on the PET scanning data of the scanned object, and obtains a parameter image for characterizing the metabolic rate in the body of the scanned object. The image processing method provided by the present invention uses a direct method for parameter reconstruction, avoids secondary propagation of noise, and adds spatiotemporal prior information, so that a parameter image with higher quality and more in line with the actual metabolic law can be obtained, and the PET scanning and static PET image reconstruction can also be set to be asynchronously executed, further improving the efficiency of parameter image imaging, and is an efficient, high-speed, high-quality parameter image imaging method.

[0098] Further, as Figure 1 The specific implementation of the present invention provides an image processing device, such as Figure 3 As shown, the device may include: a static image reconstruction module 310 , a dynamic image synthesis module 320 , an image parameter acquisition module 330 and a parameter image generation module 340 .

[0099] The static image reconstruction module 310 can be used to obtain PET scan data of multiple time frames for the scanned object, and use a static PET reconstruction method to perform static PET image reconstruction on the PET scan data to obtain multiple static PET images, and each voxel point in each static PET image corresponds to a radioactive activity.

[0100] The dynamic image synthesis module 320 can be used to combine multiple static PET images into a dynamic PET image sequence, and divide the dynamic PET image sequence into image regions according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length to obtain multiple target image regions.

[0101] The image parameter acquisition module 330 may be used to acquire the mean radioactivity value of the static PET image corresponding to each target image region as the radioactivity value corresponding to each voxel point in the target image region.

[0102] The parameter image generation module 340 can be used to perform parameter reconstruction on the PET scan data of the scanned object using the radioactivity value to obtain a parameter image for representing the metabolic rate in the body of the scanned object.

[0103] Alternatively, if Figure 4 As shown, an image processing device provided by an embodiment of the present invention may further include: a CT image reconstruction module 350 and an input function acquisition module 360 ​​.

[0104] The CT image reconstruction module 350 can be used to perform CT scanning on the scanned object, obtain a CT image whose corresponding scanned part is the same as the scanned part corresponding to the PET image sequence, and use the CT image to reconstruct the static PET image for attenuation correction.

[0105] The input function acquisition module 360 ​​can be used to cluster the heart area of ​​the scanned object according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, so as to obtain a plasma voxel point class for characterizing the plasma in the scanned object; and to perform function fitting according to the time variation characteristics of the radioactivity corresponding to each voxel point in the plasma voxel point class within a preset time length, so as to obtain a plasma input function.

[0106] Optionally, the dynamic image synthesis module 320 can also be used to cluster the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length to obtain the voxel point class of the scanned object; use the voxel point class to divide the dynamic PET image sequence into image regions; wherein the same voxel point class corresponds to the same image region; and adjust the image region division according to the connected domain information of the CT image to obtain the target image region.

[0107] Optionally, the static image reconstruction module 310 may also be used to perform PET scans of multiple time frames on the scanned object to obtain multiple pieces of PET scan data; wherein the PET scan uses a step-by-step scan or a continuous bed-moving scan;

[0108] A static PET image is reconstructed for each piece of PET scan data using a static PET reconstruction method, and the static PET image reconstruction is performed synchronously with the PET scan, including: each time a piece of PET scan data is acquired, a static PET image is reconstructed for the PET scan data;

[0109] The PET scan data is reconstructed into static PET images using an iterative reconstruction method, and multiple correction factors corresponding to the reconstruction of each static PET image are saved; the correction factors are correction data for adjusting the deviation of the scan data caused by external factors during the scanning process.

[0110] Optionally, the parameter image generation module 340 may also be used to load correction factors when performing parameter reconstruction on the PET scan data of the scanned object using the radioactivity value, and perform parameter reconstruction on the PET scan data of the scanned object using the correction factors.

[0111] Optionally, the parameter image generation module 340 can also be used to perform parameter reconstruction on the PET scan data using an iterative reconstruction method based on the radioactivity value and the plasma input function, specifically including:

[0112] The PET scan data are iteratively reconstructed using the spatiotemporal prior parameters, which are calculated using the following formula:

[0113]

[0114] Among them, R is the time-space prior parameter; β s is the spatial prior parameter; β t is the time prior parameter; γ is the empirical parameter of scanning time; N is the number of image voxel points; N i is the neighborhood of voxel point i; w ij represents the inverse of the distance between voxel point i and voxel point j; T represents the number of scanning passes; w t is the time prior weight of voxel point i; is the radioactivity value of voxel i at time point t; and

[0115] The K image and V image in the parameter image are updated using the following formula:

[0116]

[0117]

[0118] in, Used to characterize the FDG value captured or phosphorylated by cells in the human body; Used to characterize the free FDG value in the human body; N is the number of image voxels; C p (n) is the value of the plasma input function at n; S p (n) is the integral value of the plasma input function from 0 to n; This is a static PET image.

[0119] It should be noted that for other corresponding descriptions of the functional modules involved in the image processing device provided in the embodiment of the present invention, reference can be made to Figure 1 The corresponding description of the method shown will not be repeated here.

[0120] Based on the above Figure 1 The method shown, accordingly, an embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the image processing method described in any of the above embodiments are implemented.

[0121] Based on the above Figure 1 The method shown and Figure 4 The embodiment of the device shown in the figure, the embodiment of the present invention also provides a physical structure diagram of a computer device, such as Figure 5 As shown, the computer device may include a communication bus, a processor, a memory and a communication interface, and may also include an input / output interface and a display device, wherein each functional unit may communicate with each other through the bus. The memory stores a computer program, and the processor is used to execute the program stored in the memory and execute the steps of the image processing method described in the above embodiment.

[0122] Those skilled in the art can clearly understand that the specific working processes of the systems, devices, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and for the sake of brevity, they are not further described here.

[0123] In addition, the functional units in various embodiments of the present invention may be physically independent of each other, or two or more functional units may be integrated together, or all functional units may be integrated into one processing unit. The above integrated functional units may be implemented in the form of hardware, or in the form of software or firmware.

[0124] Those skilled in the art can understand that if the integrated functional unit is implemented in the form of software and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can essentially or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, which includes a number of instructions to enable a computing device (such as a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention when running the instructions. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk and other media that can store program codes.

[0125] Alternatively, all or part of the steps of implementing the aforementioned method embodiments may be accomplished by hardware associated with program instructions (such as a computing device such as a personal computer, a server, or a network device), and the program instructions may be stored in a computer-readable storage medium. When the program instructions are executed by a processor of a computing device, the computing device executes all or part of the steps of the methods described in the embodiments of the present invention.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that within the spirit and principles of the present invention, the technical solutions described in the aforementioned embodiments can still be modified, or some or all of the technical features therein can be replaced by equivalents. However, these modifications or replacements do not deviate from the protection scope of the present invention.

Claims

1. An image processing method, characterized in that: include: Acquire PET scan data of multiple time frames for the scanned object, and reconstruct static PET images on the PET scan data using a static PET reconstruction method to obtain multiple static PET images, wherein each voxel point in each static PET image corresponds to a radioactivity; Combining the plurality of static PET images into a dynamic PET image sequence, dividing the dynamic PET image sequence into image regions according to the temporal variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, to obtain a plurality of target image regions; Obtaining the mean radioactivity value of the static PET image corresponding to each target image region as the radioactivity value corresponding to each voxel point in the target image region; Performing parameter reconstruction on the PET scan data of the scanned object using the radioactivity value to obtain a parameter image for characterizing the metabolic rate in the body of the scanned object; Before the parameter reconstruction of the PET scan data of the scanned object is performed using the radioactivity value, the method further includes: For the heart region of the scanned object, clustering is performed according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within the preset time length to obtain a plasma voxel point class for characterizing the plasma in the scanned object; Performing function fitting according to the time variation characteristics of the radioactivity corresponding to each voxel point in the plasma voxel point class within the preset time length to obtain a plasma input function; The parameter reconstruction of the PET scan data of the scanned object includes: The PET scan data is parameter reconstructed using an iterative reconstruction method based on the radioactivity value and the plasma input function, specifically comprising: The PET scan data is iteratively reconstructed using the time-space prior parameters, and the time-space prior parameters are calculated using the following formula: Among them, the is the space-time prior parameter; is the spatial prior parameter; is the time prior parameter; is the empirical parameter of scanning time; is the number of image voxel points; is the neighborhood of voxel point i; Represents voxel points i With voxel point j The reciprocal of distance; T Indicates the number of scan passes; Voxel point i The weight of the temporal prior; For time point t When the voxel point i The radioactivity value of The K image and V image in the parameter image are updated using the following formula: in, Used to characterize the FDG value captured or phosphorylated by cells in the human body; Used to characterize the free FDG value in the human body; is the number of image voxel points; is the value of plasma input function at n; is the integral value of the plasma input function from 0 to n; This is a static PET image.

2. The method according to claim 1, characterized in that Before performing a PET scan on the scanned object, the method further includes: Performing a CT scan on the scanned object to obtain a CT image of a corresponding scanned part that is the same as the scanned part corresponding to the PET image sequence; The static PET image is reconstructed using the CT image to perform attenuation correction.

3. The method according to claim 2, characterized in that The step of dividing the dynamic PET image sequence into image regions according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time period to obtain the target image region includes: Clustering the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within the preset time length to obtain the voxel point class of the scanned object; Using the voxel point class to divide the dynamic PET image sequence into image regions; wherein the same voxel point class corresponds to the same image region; The image region is divided and adjusted according to the connected domain information of the CT image to obtain the target image region.

4. The method according to claim 1, characterized in that: The method of reconstructing a static PET image of the PET scan data using a static PET reconstruction method includes: Reconstructing static PET images of the PET scan data using an iterative reconstruction method, and saving a plurality of correction factors corresponding to the reconstruction of each static PET image; the correction factors are correction data for adjusting the deviation of the scan data caused by external factors during the scanning process; When the radioactivity value is used to perform parameter reconstruction on the PET scan data of the scanned object, the correction factor is loaded, and the correction factor is used to perform parameter reconstruction on the PET scan data of the scanned object.

5. The method according to any one of claims 1 to 4, characterized in that: The step of acquiring PET scan data of a plurality of time frames of the scanned object and reconstructing a static PET image of the PET scan data using a static PET reconstruction method comprises: Performing PET scans of multiple time frames on the scanned object to obtain multiple pieces of PET scan data; wherein the PET scan uses a step-by-step scan or a continuous bed-moving scan; A static PET reconstruction method is used to perform static PET image reconstruction on each PET scan data. The static PET image reconstruction is performed asynchronously with the PET scan, including: each time a PET scan data is acquired, static PET image reconstruction is performed on the PET scan data.

6. An image processing device, characterized in that: include: A static image reconstruction module, used to obtain PET scan data of multiple time frames for the scanned object, and to perform static PET image reconstruction on the PET scan data using a static PET reconstruction method to obtain multiple static PET images, wherein each voxel point in each of the static PET images corresponds to a radioactivity; A dynamic image synthesis module, used for combining the plurality of static PET images into a dynamic PET image sequence, dividing the dynamic PET image sequence into image regions according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within a preset time length, and obtaining a plurality of target image regions; An image parameter acquisition module, used to acquire the mean value of the radioactivity of the static PET image corresponding to each target image region as the radioactivity value corresponding to each voxel point in the target image region; A parameter image generation module, used for performing parameter reconstruction on the PET scan data of the scanned object using the radioactivity value to obtain a parameter image for characterizing the metabolic rate in the body of the scanned object; An input function acquisition module is used to cluster the heart region of the scanned object according to the time variation characteristics of the radioactivity corresponding to the voxel points in the dynamic PET image sequence within the preset time length to obtain a plasma voxel point class for characterizing the plasma in the scanned object; and to perform function fitting according to the time variation characteristics of the radioactivity corresponding to each voxel point in the plasma voxel point class within the preset time length to obtain a plasma input function; The parameter image generation module is used to perform parameter reconstruction on the PET scan data using an iterative reconstruction method based on the radioactivity value and the plasma input function, and specifically includes: The PET scan data is iteratively reconstructed using the time-space prior parameters, and the time-space prior parameters are calculated using the following formula: Among them, the is the space-time prior parameter; is the spatial prior parameter; is the time prior parameter; is the empirical parameter of scanning time; is the number of image voxel points; is the neighborhood of voxel point i; Represents voxel points i With voxel point j The reciprocal of distance; T Indicates the number of scan passes; Voxel point i The weight of the temporal prior; For time point t When the voxel point i The radioactivity value of The K image and V image in the parameter image are updated using the following formula: in, Used to characterize the FDG value captured or phosphorylated by cells in the human body; Used to characterize the free FDG value in the human body; is the number of image voxel points; is the value of plasma input function at n; is the integral value of the plasma input function from 0 to n; This is a static PET image.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the image processing method according to any one of claims 1 to 5 are implemented.

8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the image processing method according to any one of claims 1 to 5 are implemented.

Images