Three-dimensional Compton scattering imaging method and system based on a parallel hole collimator

Abstract

The three-dimensional Compton scattering imaging method and system based on a parallel-hole collimator disclosed by the present invention, on the basis of constructing a Compton imaging device according to the forward transmission model of scattered photons, uses the Compton imaging device to collect scattered projection images and transmitted projection images, then constructs a scattering problem solving model, and solves by integrating the scattered projection images and the transmitted projection images to obtain a set of high-precision electron density images, realizing three-dimensional Compton scattering imaging. This method broadens the application scope of transmission CT imaging and promotes the clinical application of CT imaging technology in low-dose diagnosis.

Description

Technical Field

[0001] The invention belongs to the field of medical engineering technology, and in particular relates to a three-dimensional Compton scattering imaging method and system based on a parallel hole collimator. Background Art

[0002] Since the discovery of X-rays by Roentgen in 1895, X-rays have been widely used in medical diagnosis, defense industry, security testing, material development and drug screening. With the increasing commercial application of X-ray imaging technology, conventional transmission imaging technology has developed to a fairly complete stage. In recent years, with the continuous improvement of detection technology and parallel computing capabilities, the development of X-ray scattering imaging technology has ushered in new opportunities.

[0003] In the diagnostic and therapeutic energy range (20keV to 10MeV), the interaction between X-rays and matter mainly includes: photoelectric absorption, scattering (Thompson scattering and Compton scattering) and electron pair effect. Among them, Compton scattering and photoelectric absorption are the two most important mechanisms in diagnostic imaging. Conventional transmission imaging is a simplified exponential decay model based on the Beer-Lambert law, that is, it is assumed that the rays propagate in a straight line from the source to the detector and the transmittance of the X-rays after passing through the object is measured. It is true that the rapid development of X-ray computed tomography (CT) technology has achieved great success in the biomedical field, but this type of imaging modality also has obvious disadvantages, mainly reflected in: (1) offline data processing, requiring subsequent CT reconstruction; (2) requiring a higher radiation dose to maintain image quality; (3) poor image contrast, insufficient to distinguish tumor boundaries. These problems related to dose and speed are essentially due to the fact that transmission imaging ignores scattered photons that propagate non-linearly. From a physical mechanism point of view, the scattering process of photons also carries information inside the material and can be developed into scattering CT imaging. Scattering imaging technology is expected to break through the dependency between irradiation and detection paths. As an effective supplement to transmission attenuation imaging, scattering imaging can make full use of the interaction information between light and matter, thereby minimizing the radiation dose while maximizing the diagnostic information. Therefore, in-depth research on the scattering imaging mechanism, building a scattering path reconstruction algorithm, and realizing Compton scattering imaging will fundamentally change the CT technical framework, improve the technical level of CT equipment from the source, and open up a new path for exploring the intrinsic structure of matter, which has important theoretical and practical significance.

[0004] According to different CT diagnosis and treatment requirements, the proportion of incident X-rays scattered in the human body is as high as 30%-60%. Under the architecture of transmission CT, this part of the radiation dose will form a blurred background on the detector surface, reducing the image signal-to-noise ratio. Therefore, scattered photons are usually regarded as "noise" pollution, and physical blocking, algorithm correction, or deep learning scattered artifact correction is adopted. The anti-scatter grid is usually placed in front of the detector to block the photons scattered by the body. By simulating and measuring the scattered photon statistical model, this "noise" can be effectively suppressed. Unfortunately, the scattered photon suppression technology also discards a considerable part of the unscattered photons. Summary of the Invention

[0005] In view of the above, the object of the present invention is to provide a three-dimensional Compton scattering imaging method and system based on a parallel-hole collimator, which realizes the scattered imaging of an object by making full use of the scattered photons of X-rays in transmission CT scanning.

[0006] To achieve the above object of the invention, the three-dimensional Compton scattering imaging system based on a parallel-hole collimator provided by the embodiment includes a Compton imaging device and an imaging calculator;

[0007] The Compton imaging device includes an X-ray source and a transmission detector arranged along the X-ray optical axis. There is a rotating platform for carrying the item to be measured between the X-ray source and the transmission detector, and it is ensured that the item to be measured is irradiated by X-rays and is completely imaged in the transmission detector. Two sets of scattered detection components are arranged along the direction perpendicular to the optical axis and on both sides of the rotating platform. Each set of scattered detection components includes a scattered detector and a parallel-hole collimator on the scattered detection optical path, and it is ensured that the two sets of scattered detection components can detect the complete item to be measured;

[0008] The imaging calculator realizes three-dimensional Compton scattering imaging calculation, including the following steps:

[0009] Step 1, obtaining a transmission empty-field projection image of the transmission detector scanning the empty field;

[0010] Step 2, obtaining two sets of scattered projection image sets and a transmission projection image set obtained by simultaneously scanning the item to be measured rotating one week by the two sets of scattered detection components and the transmission detector;

[0011] Step 3, performing standard cone-beam CT filtered back-projection reconstruction according to the transmission empty-field projection image and the transmission projection image set to obtain a set of linear attenuation coefficient distribution images of the item to be measured;

[0012] Step 4, calculating the system matrix of the forward projection based on the set of linear attenuation coefficient distribution images;

[0013] Step 5: Based on the combination of two sets of scattered projection image sets and the system matrix of forward projection, perform electron density image reconstruction to obtain the electron density image set of the item under test, thereby achieving three-dimensional Compton scattering imaging.

[0014] Preferably, the distances from the centers of the transmission detector and the two scattered detectors to the rotation axis of the rotating platform are equal.

[0015] Preferably, in each set of scattered detection components, the parallel-hole collimator is attached to the detection surface of the scattered detector.

[0016] Preferably, in Step 4, the following formula is used to calculate the system matrix of forward projection based on the linear attenuation coefficient distribution image set:

[0017]

[0018] where E represents the scattered photon energy, θ represents the photon scattering angle, I 0 represents the number of incident photons, l is the path length of the photon passing through the item under test, μ p and μ s are the linear attenuation coefficient values at the incident beam and scattered beam energies, which are obtained from the linear attenuation coefficient distribution image set, represents the value of the Klein-Nishina differential cross-section related to the scattered photon energy E and the photon scattering angle θ, ΔΩ represents the scattering area, ρ e represents the electron density, and M represents the system matrix of the forward projection to be solved.

[0019] Preferably, Step 5 includes:

[0020] Step 5-1: Based on the combination of two sets of scattered projection image sets and the system matrix of forward projection, establish the objective function for electron density image reconstruction, expressed as:

[0021]

[0022] where, represents the two sets of scattered projection image sets after vectorization, M is the system matrix of forward projection, is the electron density image set to be reconstructed, represents the optimal reconstructed electron density image set, Σ is a diagonal matrix, and the i-th element on the diagonal is the variance estimate of the i-th pixel in ‖·‖ TV is the L1 norm of the spatial gradient image, λ is the regularization term coefficient, and argmin(·) represents the value of the independent variable when the value of the objective function is minimized.

[0023] Step 5-2: Minimize the objective function using the fast iterative shrinkage-thresholding algorithm with linear search to obtain an optimal set of reconstructed electron density images.

[0024] Preferably, step 5-2 includes:

[0025] Step 5-2-1: Construct a convex function based on the objective function and convex function

[0026]

[0027]

[0028] Step 5-2-2: Modify the objective function according to convex function and convex function to:

[0029]

[0030] Step 5-2-3: Calculate the gradient of convex function

[0031]

[0032] Step 5-2-4: Perform an approximation operation on convex function to obtain an approximate function of

[0033]

[0034] where P + (·) is the projection function in the non-negative quadrant, is the maximum value between the pixel value of the image set and 0;

[0035] Step 5-2-5: Iteratively solve the objective function shown in step 5-2-2 according to the gradient and the approximate function to obtain an optimal reconstructed electron density image.

[0036] Preferably, in step 5-2-5, the iterative solution process is:

[0037]

[0038] where is an intermediate variable of the reconstructed electron density image in the (n - 1)-th iteration process, and is an intermediate variable of the reconstructed electron density image in the n-th iteration process. and represent the sets of reconstructed electron density images in the (n - 1)-th iteration and the n-th iteration respectively, and t n represents the step size of the n-th iteration, and θ n represents the positive root of the n-th iteration, and the calculation formula is:

[0039]

[0040] where θ n-1 represents the positive root of the (n - 1)-th iteration, and t n-1 represents the step size of the (n - 1)-th iteration, and t n is selected as the maximum step size that satisfies formula (9).

[0041] Preferably, in step 5 - 2 - 5, the termination condition for iterative solution is:

[0042]

[0043] where <·> represents the sum of products, represents the square of the two-norm.

[0044] To achieve the above-mentioned invention purpose, the three-dimensional Compton scattering imaging method based on a parallel-hole collimator provided by the embodiment, which applies the above-mentioned three-dimensional Compton scattering imaging system, includes the following steps:

[0045] Step 1, using a transmission detector to scan the empty field to obtain a transmission empty-field projection image;

[0046] Step 2, using two sets of scattering detection components and a transmission detector to simultaneously scan the item to be measured that rotates one week, to obtain two sets of scattering projection image sets and a transmission projection image set;

[0047] Step 3, performing filtered back-projection reconstruction based on the transmission empty-field projection image and the transmission projection image set to obtain a set of linear attenuation coefficient distribution images of the item to be measured;

[0048] Step 4, calculating the system matrix of the forward projection based on the set of linear attenuation coefficient distribution images;

[0049] Step 5, according to the two sets of scattering projection image sets and in combination with the system matrix of the forward projection, performing electron density image reconstruction to obtain a set of electron density images of the item to be measured, so as to achieve three-dimensional Compton scattering imaging.

[0050] Compared with the prior art, the beneficial effects of the present invention at least include:

[0051] On the basis of constructing a Compton imaging device according to the forward transport model of scattered photons, the Compton imaging device is used to collect scattered projection images and transmitted projection images, then a scattering problem solving model is constructed, and the scattered projection images and transmitted projection images are combined and solved to obtain a set of high-precision electron density images, realizing three-dimensional Compton scattering imaging. This method broadens the application range of transmission CT imaging and promotes the clinical application of CT imaging technology in low-dose diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0053] Figure 1 is a schematic structural diagram of the Compton imaging device provided by the embodiment;

[0054] Figure 2 is a schematic structural diagram of the parallel hole collimator provided by the embodiment;

[0055] Figure 3 is the digital phantom simulation scattered photon sine image provided by the embodiment;

[0056] Figure 4 is the digital phantom electron density reconstruction image provided by the embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] To make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and do not limit the protection scope of the present invention.

[0058] Research has found that due to the scattering of X-ray photons in all directions, placing more detectors around the object to be measured can collect the material information carried by this part of the radiation dose. Since each photon has information related to the scattering position and type, such as wave vector, polarization, and energy, etc., the collected signals are richer. In this way, the more information dimensions, the better the contrast of the imaging image, providing new evaluation capabilities for biomedical imaging and also providing new material contrast information for industrial safety detection.

[0059] The electron density is derived from the computed tomography (CT) numbers measured in X-ray transmission CT scans. However, due to the beam hardening effect and the method of converting CT values to electron density, there are uncertainties. Since the direction of X-ray scattering is different from the direction of the incident light, real-time online tomography can be achieved, so that the structural details of multiple transmission exposures can be obtained by single-scattering measurement. Through image reconstruction, the precise distribution of electron density is obtained, and thus the radiation dose within the energy range can be accurately calculated.

[0060] Through research, it has been found that Compton scattering imaging technology can be integrated with conventional transmission CT to obtain richer material information. However, Compton scattering imaging technology is not yet mature, and the main problems are as follows: the physical mechanism has not been clarified, the evaluation of the hardware architecture is insufficient, the scattering photon statistical model is imperfect, the reconstruction algorithm is lacking, and the sensitivity of the detection end is limited. To solve these problems, the embodiments provide a three-dimensional Compton scattering imaging system and method based on a parallel-hole collimator.

[0061] The three-dimensional Compton scattering imaging system based on a parallel-hole collimator provided by the embodiments includes a Compton imaging device and an imaging calculator. Among them, the Compton imaging device is used to collect transmission projection images and scattering projection images, and the imaging calculator is used to reconstruct an electron density image based on the collected images to achieve three-dimensional Compton scattering imaging.

[0062] As Figure 1 shown, the Compton imaging device provided by the embodiments includes an X-ray source 1, a transmission detector 2, a rotating platform 3, and two groups of scattering detection groups 4 and 5. Among them, the X-ray source 1 and the transmission detector 2 are arranged along the X-ray optical axis. The rotating platform 3 is used to carry the item to be measured, and is arranged between the X-ray source 1 and the transmission detector 2, and ensures that the item to be measured is irradiated by X-rays and is completely imaged in the transmission detector. The two groups of scattering detection components 4 and 5 are arranged on both sides of the rotating platform along the direction perpendicular to the optical axis, and ensure that the two groups of scattering detection components can detect the complete item to be measured.

[0063] In the embodiments, each group of scattering detection components includes a scattering detector and a parallel-hole collimator on the scattering detection optical path. In an implementation manner, the parallel-hole collimator is attached to the detection surface of the scattering detector, which is more convenient to prepare.

[0064] In the embodiments, the parallel-hole collimator is arranged on the scattering detection optical path of the scattering detector, and is used to correct the incident direction of the incident photons of the scattering detector, so that the incident directions of all incident photons are the same, and the quality of the scattering projection image is improved.

[0065] In the embodiments, the geometric dimensions of the parallel-hole collimator are designed according to the imaging resolution and imaging efficiency of the parallel-hole collimator. Figure 2 is the three-dimensional structure diagram of the optimal parallel-hole collimator designed in the embodiments. As Figure 2As shown, the total length of the parallel-hole collimator is 400.0 mm, the total width is 300.0 mm, the collimator aperture is 0.16 mm, the distance between adjacent small holes is 36.0 μm, and the collimator thickness is 5.0 mm.

[0066] In an embodiment, taking the intersection point of the rotation axis of the rotation platform 3 and the X-ray optical axis as the coordinate origin, the direction parallel to the X-ray optical axis as the positive direction of the Y-axis, the direction parallel to the rotation axis as the positive direction of the Z-axis, and the direction perpendicular to the X-ray optical axis and the rotation axis as the X-direction, an imaging coordinate system is established. An X-ray source 1, a rotation platform 3, and a transmission detector 2 are sequentially arranged along the Y-axis direction, wherein the centers of the X-ray source 1 and the transmission detector 2 are aligned. A set of scattering detection components 4 is vertically arranged in the X-axis direction on one side of the rotation platform 2, and a second set of scattering detection components 5 is vertically arranged in the X-axis direction at a symmetric position on the other side of the rotation platform 3; in each set of scattering detection components, the parallel-hole collimator is closely attached to the detection surface of the scattering detector. The distances between the centers of the two scattering detectors and the transmission detector 2 and the rotation axis of the rotation platform 3 are equal.

[0067] Based on the above Compton imaging device, a transmission empty-field projection image is obtained by scanning an empty field with the transmission detector, and the two sets of scattering projection image sets and the transmission projection image set are obtained by simultaneously scanning the item to be measured that rotates one week with the two sets of scattering detection components and the transmission detector. Based on the acquired images, the imaging calculator realizes three-dimensional Compton scattering imaging calculation, including the following steps:

[0068] Step 1, obtain the transmission empty-field projection image.

[0069] Step 2, obtain the two sets of scattering projection image sets and the transmission projection image set.

[0070] Step 3, perform standard cone-beam CT filtered back-projection reconstruction according to the transmission empty-field projection image and the transmission projection image set to obtain the linear attenuation coefficient distribution image set of the item to be measured.

[0071] Step 4, calculate the system matrix of the forward projection based on the linear attenuation coefficient distribution image set.

[0072] In the embodiment, the following formula is used to calculate the system matrix of the forward projection based on the linear attenuation coefficient distribution image set:

[0073]

[0074] Among them, E represents the scattered photon energy, θ represents the photon scattering angle, I 0 represents the number of incident photons, l is the path length of the photon passing through the item to be measured, μ p and μ sis the value of the linear attenuation coefficient at the incident beam and scattered beam energies, which is obtained from a set of linear attenuation coefficient distribution images. represents the value of the Klein-Nishina differential cross section related to the scattered photon energy E and the photon scattering angle θ, ΔΩ represents the scattering area, and ρ e represents the electron density, and M represents the system matrix of the forward projection to be solved.

[0075] Step 5: Based on the two sets of scattered projection image sets and combined with the system matrix of the forward projection, perform electron density image reconstruction to obtain a set of electron density images of the item to be measured, and achieve three-dimensional Compton scattering imaging.

[0076] In the embodiment, the process of electron density image reconstruction includes:

[0077] Step 5-1: Based on the two sets of scattered projection image sets and combined with the system matrix of the forward projection, establish an objective function for electron density image reconstruction, expressed as:

[0078]

[0079] where, represents the two sets of scattered projection image sets after vectorization, M is the system matrix of the forward projection, is the set of electron density images to be reconstructed, represents the set of optimal reconstructed electron density images, Σ is a diagonal matrix, and the i-th element on the diagonal is the variance estimate of the i-th pixel in ‖·‖ TV is the L1 norm of the spatial gradient image, λ is the regularization term coefficient, and argmin(·) represents the value of the independent variable when the value of the objective function is minimized.

[0080] Step 5-2: Use the fast iterative shrinkage-thresholding algorithm with linear search to minimize the objective function to obtain the set of optimal reconstructed electron density images.

[0081] Specifically, the process of using the fast iterative shrinkage-thresholding algorithm with linear search to minimize the objective function and solve the electron density image includes:

[0082] Step 5-2-1: Construct convex functions and convex function

[0083]

[0084]

[0085] Step 5-2-2, according to the convex function and the convex function the objective function is modified to:

[0086]

[0087] Step 5-2-3, calculate the gradient of the convex function which is

[0088]

[0089] Step 5-2-4, perform an approximation operation on the convex function to obtain the approximate function of

[0090]

[0091] where P + (·) is the projection function in the non-negative quadrant, is the maximum value between the pixel value of the image set and 0;

[0092] Step 5-2-5, according to the gradient and the approximate function iteratively solve the objective function shown in Step 5-2-2. The iterative solution process is:

[0093]

[0094] where is the intermediate variable of the reconstructed electron density image in the (n - 1)-th iteration process, and are the intermediate variables of the reconstructed electron density image in the n-th iteration process, and respectively represent the reconstructed electron density image sets in the (n - 1)-th iteration and the n-th iteration, t n represents the step size of the n-th iteration, and θ n represents the positive root of the n-th iteration. The calculation formula is:

[0095]

[0096] where θ n-1 represents the positive root of the (n - 1)-th iteration, and t n-1 represents the step size of the (n - 1)-th iteration, and t n is selected as the maximum step size that satisfies formula (9).

[0097] Step 5-2-6: When the stopping condition for the iteration of the objective function as shown in Equation (10) is reached, the optimal reconstructed electron density image is obtained.

[0098]

[0099] Wherein, <·> represents the sum of products, represents the square of the two-norm.

[0100] The three-dimensional Compton scattering imaging system based on a parallel-hole collimator provided by the embodiment constructs a forward transmission model of scattered photons, conducts physical design of cone-beam incidence and builds a Compton imaging device, and collects scattered signals perpendicular to the X-ray optical path direction. On this basis, an accurate scattering problem solving model is established and solved to break through the limitation of the random path of scattered photons and improve the signal-to-noise ratio of the scattered image. This system further broadens the application scope of transmission CT imaging and promotes the clinical application of CT imaging technology in low-dose diagnosis.

[0101] Based on the same inventive concept, the embodiment also provides a three-dimensional Compton scattering imaging method based on a parallel-hole collimator, including the following steps:

[0102] Step a: Use a transmission detector to scan the empty field to obtain a transmission empty-field projection image.

[0103] In the embodiment, after starting the X-ray source 1, use the transmission detector 2 to obtain a transmission empty-field projection image according to the exposure duration t1, and then turn off the X-ray source 1.

[0104] Step b: Use two groups of scattering detection components and a transmission detector to simultaneously scan the item to be measured for one rotation, and obtain two sets of scattered projection image sets and a transmission projection image set.

[0105] In the embodiment, set the total number of exposures T = 360° / ((t1 + t2) × v1); where, v1 represents the rotation speed of the rotation platform 3; t2 represents the interval duration.

[0106] Then, first start the rotation platform 3 so that the item to be measured starts to rotate uniformly at the rotation speed v1, and after starting the X-ray source 1, use the first scattering detector of the scattering detection component 4 to obtain the scattered projection image of the item to be measured according to the single exposure duration t1, use the second scattering detector of the scattering detection component 5 to obtain the scattered projection image of the item to be measured according to the single exposure duration t1, and at the same time use the transmission detector 2 to obtain the transmission projection image of the item to be measured according to the single exposure duration t1; thus, after T exposures, a set of transmission projection images after the rotation platform 2 rotates one week, and two sets of scattered projection image sets corresponding to the scattering detection component 4 and the scattering detection component 5 are obtained, and then turn off the X-ray source 1 and the rotation platform 3.

[0107] Step c: Perform filtered backprojection reconstruction based on the transmission empty-field projection image and the set of transmission projection images to obtain a set of linear attenuation coefficient distribution images of the item under test.

[0108] Step d: Calculate the system matrix of the forward projection based on the set of linear attenuation coefficient distribution images.

[0109] Step e: Perform electron density image reconstruction according to the two sets of scattered projection images and in combination with the system matrix of the forward projection to obtain a set of electron density images of the item under test, thereby realizing three-dimensional Compton scattering imaging.

[0110] In the embodiment, the processes of Step c to Step e are as detailed in the above Step 3 to Step 5.

[0111] Figure 3 It is the simulated scattered photon sine image of the digital phantom provided by the embodiment, that is, the image acquired by the scattered detector. The digital phantom is imaged by using the three-dimensional Compton scattering imaging method based on the parallel-hole collimator provided by the embodiment, and the reconstructed electron density image obtained is as Figure 4 .

[0112] The above-described specific embodiments have elaborated on the technical solutions and beneficial effects of the present invention. It should be understood that the above is only the most preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, supplements, equivalent replacements, etc. made within the scope of the principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A three-dimensional Compton scattering imaging system based on a parallel-hole collimator, characterized in that, it includes a Compton imaging device and an imaging calculator; The Compton imaging device includes an X-ray source and a transmission detector arranged along the X-ray optical axis. There is a rotating platform for carrying the item to be measured between the X-ray source and the transmission detector, and it is ensured that the item to be measured is irradiated by X-rays and is completely imaged in the transmission detector. Two sets of scattering detection components are arranged along the direction perpendicular to the optical axis and on both sides of the rotating platform. Each set of scattering detection components includes a scattering detector and a parallel-hole collimator on the scattering detection optical path, and it is ensured that the two sets of scattering detection components can detect the complete item to be measured; The imaging calculator realizes three-dimensional Compton scattering imaging calculation, including the following steps: Step 1, obtain the transmission empty-field projection image of the transmission detector scanning the empty field; Step 2, obtain two sets of scattering projection image sets and a transmission projection image set obtained by scanning the item to be measured rotating one week by the two sets of scattering detection components and the transmission detector simultaneously; Step 3, perform filtered back-projection reconstruction based on the transmission empty-field projection image and the transmission projection image set to obtain a set of linear attenuation coefficient distribution images of the item to be measured; Step 4, calculate the system matrix of forward projection based on the set of linear attenuation coefficient distribution images; Step 5, according to the two sets of scattering projection image sets and combined with the system matrix of forward projection, perform electron density image reconstruction to obtain a set of electron density images of the item to be measured, and realize three-dimensional Compton scattering imaging.

2. The three-dimensional Compton scattering imaging system based on a parallel-hole collimator according to claim 1, characterized in that, the distances from the centers of the transmission detector and the two scattering detectors to the rotation axis of the rotating platform are equal.

3. The three-dimensional Compton scattering imaging system based on a parallel-hole collimator according to claim 1, characterized in that, in each set of scattering detection components, the parallel-hole collimator is attached to the detection surface of the scattering detector.

4. The three-dimensional Compton scattering imaging system based on a parallel-hole collimator according to claim 1, characterized in that, in step 4, the following formula is used to calculate the system matrix of forward projection based on the set of linear attenuation coefficient distribution images: where E represents the scattered photon energy, θ represents the photon scattering angle, I 0 represents the number of incident photons, l is the path length of the photon passing through the object under test, μ p and μ s are the values of the linear attenuation coefficient at the incident beam and scattered beam energies, which are obtained from the set of linear attenuation coefficient distribution images, represents the value of the Klein-Nishina differential cross section related to the scattered photon energy E and the photon scattering angle θ, ΔΩ represents the scattering area, ρ e represents the electron density, and M represents the system matrix of the forward projection to be solved.

5. The three-dimensional Compton scattering imaging system based on a parallel-hole collimator according to claim 1, characterized in that, step 5 includes: Step 5-1, according to the two sets of scattering projection image sets and combined with the system matrix of forward projection, establish an objective function for electron density image reconstruction, expressed as: Among them, represents two sets of vectorized scattered projection image sets, M is the system matrix of forward projection, is the set of electron density images to be reconstructed, represents the set of optimal reconstructed electron density images, Σ is a diagonal matrix, and the i-th element on the diagonal is the variance estimate of the i-th pixel in ‖·‖ TV is the L1 norm of the spatial gradient image, λ is the regularization term coefficient, and argmin(·) represents the value of the independent variable when the value of the objective function is minimized; Step 5-2, use the fast iterative shrinkage thresholding algorithm with linear search to minimize the objective function to obtain the optimal reconstructed set of electron density images.

6. The three-dimensional Compton scattering imaging system based on a parallel-hole collimator according to claim 5, characterized in that, step 5-2 includes: Step 5-2-1, construct a convex function according to the objective function and the convex function : Step 5-2-2, according to the convex function and the convex function the objective function is modified to: Step 5-2-3, calculate the gradient of the convex function :​ Step 5-2-4, approximate the convex function to obtain an approximate function of : Among them, P + (·) is the projection function in the non - negative quadrant, is the maximum value between the pixel value of the image set and 0; Step 5-2-5, according to the gradient and the approximation function iteratively solve the objective function shown in Step 5-2-2 to obtain the optimal reconstructed electron density image.

7. The three-dimensional Compton scattering imaging system based on a parallel-hole collimator according to claim 6, characterized in that, in step 5-2-5, the iterative solution process is: Among them, is an intermediate variable of the reconstructed electron density image in the (n - 1)-th iteration process, and are intermediate variables of the reconstructed electron density image in the n-th iteration process, and respectively represent the sets of reconstructed electron density images in the (n - 1)-th iteration and the n-th iteration, t n represents the step size of the n-th iteration, θ n represents the positive root of the n-th iteration, and the calculation formula is: where, θ n-1 represents the positive root of the (n - 1)-th iteration, t n-1 represents the step size of the (n - 1)-th iteration, and t n is selected as the maximum step size that satisfies formula (9).

8. The three-dimensional Compton scattering imaging system based on a parallel-hole collimator according to claim 7, characterized in that, In step 5-2-5, the termination condition for iterative solution is: where <·> represents the sum of products, represents the square of the second norm.

9. A three-dimensional Compton scattering imaging method based on a parallel hole collimator, characterized in that the method is applied to the three-dimensional Compton scattering imaging system according to any one of claims 1-8, and includes the following steps: Step 1, using a transmission detector to scan an empty field to obtain a transmission empty field projection image; Step 2, using two sets of scattering detection components and a transmission detector to simultaneously scan the item to be measured for one rotation to obtain two sets of scattering projection image sets and a transmission projection image set; Step 3, performing filtered backprojection reconstruction based on the transmission empty field projection image and the transmission projection image set to obtain a set of linear attenuation coefficient distribution images of the item to be measured; Step 4, calculating the system matrix of forward projection based on the set of linear attenuation coefficient distribution images; Step 5, according to the two sets of scattering projection image sets and in combination with the system matrix of forward projection, performing electron density image reconstruction to obtain a set of electron density images of the item to be measured, so as to realize three-dimensional Compton scattering imaging.

Images