Compton camera XFCT imaging system and image reconstruction method thereof

Abstract

The invention discloses a Compton camera XFCT imaging system and an image reconstruction method thereof.The Compton camera XFCT imaging system comprises a radiation source and two sets of double-layer Compton camera detectors, the two sets of double-layer Compton camera detectors are arranged on the same straight line at intervals, and the radiation source is located beside the straight line; the two sets of double-layer Compton camera detectors are symmetrical about the ray source, and the ray source is a gamma ray source; each of the two sets of double-layer Compton camera detectors comprises a scattering detector and an absorption detector, and the scattering detectors are used for recording Compton scattering positions and deposition energy of recoil electrons; and the absorption detector is used for receiving the scattered photons and recording an absorption position and deposition energy.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging technology, and in particular to a Compton camera XFCT imaging system and an image reconstruction method thereof. Background Art

[0002] Existing X-ray Computed Tomography Technology (XCT) includes traditional spiral CT technology, static CT technology, Micro-CT technology, etc. It is based on the differences in the absorption and attenuation of X-rays by different biological tissues. When the X-ray beam passes through the human body, it is absorbed and scattered by different tissues of the human body, resulting in changes in the intensity of the X-rays after passing through the human body. The detector receives these attenuated X-ray signals and converts them into electrical signals or digital signals. The computer then processes and reconstructs these signals to ultimately generate a tomographic image of the human body.

[0003] X-ray fluorescence computed tomography (XFCT) is an imaging technology that combines X-ray fluorescence analysis and X-CT. It irradiates the sample with X-rays to excite the high-Z tracer atoms inside it to produce X-ray fluorescence photons, which are then received by the X-ray fluorescence detector, thereby providing spatial distribution information of the tracer concentration inside the sample, overcoming the shortcomings of traditional CT in detecting certain microlesions or when requiring high tissue differentiation. However, this technology still has many problems. When used, the direction of the photons needs to be mechanically collimated, which reduces the photon absorption rate, and requires long-term scanning, thereby increasing the radiation risk. The spatial resolution is also low, and there are limitations in resolving high atomic number elements. The existing image reconstruction algorithm takes a lot of time to acquire data, the calculation process is complex, and it is difficult to remove noise, resulting in low imaging accuracy.

[0004] Compton Camera (CC) uses the Compton scattering physical effect to track the direction of incident photons and perform "electronic collimation". It does not require a mechanical collimator and can achieve three-dimensional imaging through one or more scanning views. It is small in size, light in weight, and low in power consumption, making it suitable for portable and space-constrained application scenarios. However, this method also has many shortcomings, including low utilization of fluorescent photons, low detection efficiency, limited energy resolution, and high computational complexity, long time consumption, and severe noise interference when processing large amounts of data or high-resolution imaging. Summary of the invention

[0005] The object of the present invention is to provide a Compton camera XFCT imaging system and an image reconstruction method thereof, so as to solve the problems of low imaging efficiency and poor imaging quality of the existing imaging system.

[0006] A first aspect of the present invention provides a Compton camera XFCT imaging system, comprising:

[0007] X-ray source and two sets of double-layer Compton camera detectors.

[0008] The two sets of double-layer Compton camera detectors are arranged in a straight line at intervals, the ray source is located beside the straight line, the two sets of double-layer Compton camera detectors are symmetrical about the ray source, and the ray source is a gamma ray source;

[0009] Both sets of double-layer Compton camera detectors include scattering detectors and absorption detectors. The scattering detectors are used to record the location where Compton scattering occurs and the deposited energy of the recoil electrons; the absorption detectors are used to receive scattered photons and record the absorption location and deposited energy.

[0010] Furthermore, the distance between the scattering detector and the sample to be measured is 60 mm, and the distance between the scattering detector and the absorption detector is 20 mm.

[0011] Furthermore, the scattering detector is made of silicon and has a size of 20 mm×20 mm×5 mm.

[0012] Furthermore, the absorption detector is made of cadmium zinc telluride and has a size of 50 mm×50 mm×5 mm.

[0013] The second aspect of the present invention provides an image reconstruction method for a Compton camera XFCT imaging system, which is used in the above-mentioned Compton camera XFCT imaging system, wherein the sample to be tested is placed on a straight line where two sets of double-layer Compton camera detectors are arranged at intervals, and the two sets of double-layer Compton camera detectors are symmetrical about the straight line connecting the ray source and the sample to be tested, comprising:

[0014] S01: irradiating the gamma rays emitted by the ray source to the sample to be tested, and the gamma rays interact with the elements in the sample to be tested to generate fluorescent photons; wherein the fluorescent photons are incident on the scattering detector to generate Compton scattering events, and the fluorescent photons after Compton scattering are absorbed by the absorption detector to generate Compton absorption events;

[0015] S02: Receive and record the fluorescent photons when the Compton scattering event occurs through the corresponding scattering detector to obtain Compton scattering event data, wherein the Compton scattering event data includes the scattering point position (x s ,y s ,z s ) and the deposited energy E s ;

[0016] S03: Receive and record the fluorescent photons when the Compton absorption event occurs through the corresponding absorption detector to obtain Compton absorption event data, wherein the Compton absorption event data includes the absorption point position (x a ,y a ,z a ) and the deposited energy E a ;

[0017] S04: filtering the Compton scattering event data and the Compton absorption event data;

[0018] S05: Calculate the angle of the filtered Compton scattering event data and the Compton absorption event data to obtain a scattering angle θ i ;

[0019] S06: dividing the space region where the sample to be tested and the two sets of double-layer Compton camera detectors are located into a plurality of cubic units to obtain a voxel space, wherein each of the cubic units serves as a voxel;

[0020] S07: According to the scattering angle θ i Back-projecting the filtered Compton scattering event data and the Compton absorption event data into the voxel space;

[0021] S08: Calculating the propagation direction of the fluorescent photons according to the projection The intersection coordinates of the fluorescent photon and the voxel, the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij , and update the system matrix to obtain the system matrix element a ij ;

[0022] S09: According to the system matrix element a ij The intensity value f of the voxel is calculated using the MLEM algorithm. j Perform iterative updates to obtain a reconstruction result matrix, perform three-dimensional volume reconstruction on the reconstruction result matrix through a visualization software tool to obtain a reconstruction result image, and adjust iteration parameters in real time according to the reconstruction result image until the reconstruction result image meets preset requirements.

[0023] Further, S04: filtering the Compton scattering event data and the Compton absorption event data, including:

[0024] Calculate the scattering point position (x s ,y s ,z s) and the absorption point position (x a ,y a ,z a ) s-a , the distance d s-a The distance threshold d th For comparison, if d s-a <d th , then the corresponding scattering point position (x s ,y s ,z s ) and the corresponding absorption point position (x a ,y a ,z a ) are deleted from the Compton scattering event data and the Compton absorption event data, respectively.

[0025] Further, S05: performing angle calculation on the filtered Compton scattering event data and the Compton absorption event data to obtain a scattering angle θ i ,include:

[0026] From the filtered Compton scattering event data and Compton absorption event data, the scattering angle θ is calculated using the Compton scattering angle calculation formula according to the deposited energy Es at the scattering point and the deposited energy Ea at the absorption point. i , the formula is:

[0027] Among them, m e c 2 Represents the electron rest energy.

[0028] Further, S08: calculating the propagation direction of the fluorescent photon according to the projection The intersection coordinates of the fluorescent photon and the voxel, the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij , and update the system matrix to obtain the system matrix element a ij ,include:

[0029] Calculate the fluorescence photon from the scattering point position (x s ,y s ,z s ) to the absorption point position (x a ,y a ,z a ) direction vector The formula is: For the direction vector Normalized processing is performed to obtain the propagation direction of the fluorescence photon The formula is:

[0030]

[0031] Calculate the intersection coordinates of the fluorescent photon and the voxel, where the position of the voxel can be expressed as (x j ,y j ,z j );

[0032] The range of the voxel in the X-axis, Y-axis, and Z-axis directions is calculated by the range formula: j1 =X j +0.5×ΔX,X j2 =X j +1×ΔX,Y j1 =Y j +0.5×ΔY,Y j2 =Y j +1×ΔY,Z j1 =Z j +0.5×ΔZ,Z j2 =Z j +1×ΔZ, where X j1 and X j2 Indicates the range of the voxel in the X-axis direction, Y j1 and Y j2 Indicates the range of the voxel in the Y-axis direction, Z j1 and Z j2 represents the range of the voxel in the Z-axis direction, ΔX, ΔY, ΔZ represent the variable range of the voxel in the X-axis, Y-axis and Z-axis respectively, and j represents the index of the voxel;

[0033] Photon propagation formula The value range of parameter t on the X-axis, Y-axis and Z-axis is calculated by the voxel boundary inequality, and the formula is:

[0034] X0+td x ∈[X j1 ,X j2 ],Y0+td y ∈[Y j1 ,Y j2 ],Z0+td z ∈[Z j1 ,Z j2 ],in, represents the position vector of the fluorescent photon at the initial position, X0, Y0, and X0 represent The components on the X-axis, Y-axis, and Z-axis, d x d y d z Respectively Components on the X-axis, Y-axis, and Z-axis;

[0035] Substitute the value range of parameter t on the X-axis, Y-axis and Z-axis into the photon propagation formula respectively Performing calculations to obtain the coordinates of the intersection of the fluorescent photon and the voxel;

[0036] Calculate the path length d of the fluorescent photon in the voxel according to the intersection coordinates ij , the formula is:

[0037] in, and represents the boundary intersection point when the fluorescent photon intersects with the voxel;

[0038] According to the Beer-Lambert law, the intensity attenuation of the fluorescent photon after passing through the voxel and the energy deposition ratio of the fluorescent photon in the voxel are calculated. ij , the formula is:

[0039] Where I0 and I represent the energy of the fluorescence photon before and after decay in the voxel, μ j represents the attenuation coefficient of fluorescence photons in the voxel;

[0040] According to the propagation direction Intersection coordinates, path length d ij and the energy deposition ratio e ij Update the system matrix to get the system matrix element a ij , the formula is:

[0041] Further, according to the system matrix element a i j uses the MLEM algorithm to calculate the intensity value f of the voxel j Perform iterative updates to obtain the reconstruction result matrix, including:

[0042] Initialize algorithm parameters, including the number of iterations n, the initial voxel intensity estimate f j (0) , assign an initial value to the weight of each voxel in the voxel space, and iteratively update the voxel intensity value f of the voxel in combination with the MLEM algorithm j , the formula is:

[0043] Wherein, i represents the index of the observation data detected by the double-layer Compton camera detector, j represents the index of the voxel, and f j represents the intensity value of voxel j, f j (n) represents the intensity value of voxel j at n iterations, f j (n+1)represents the intensity value of voxel j at the next iteration, and m represents the upper limit of the observed data.

[0044] Furthermore, adjusting the iteration parameters in real time according to the reconstructed image until the reconstructed image meets the preset requirements includes:

[0045] Adjust the preset distance threshold d th : If artifacts occur, increase the preset distance threshold d th If the number of noise events increases, the preset distance threshold d is reduced. th , where the scattering point position (x s ,y s ,z s ) and the absorption point position (x a ,y a ,z a ) s-a , is less than the preset distance threshold d th The event is judged as a noise event;

[0046] Adjust the attenuation coefficient μ of the fluorescence photons in the voxel j Value: Set different attenuation coefficients μ according to different elements j Numeric value;

[0047] Adjust the initial voxel intensity estimate f according to the element concentration in the reconstructed image j (0) .

[0048] The present invention has at least the following beneficial effects:

[0049] A Compton camera XFCT imaging system provided in the present invention can track and detect photons from more angles, can obtain more effective Compton events, improve angular resolution, significantly reduce the half-width (FWHM) of the distribution, improve ARM performance by 19%, and reduce the situation of missing fluorescent photons due to directional restrictions, thereby reducing background interference. The system can more accurately determine the propagation direction of fluorescent photons and locate the position of the radiation source, as well as have better energy resolution, thereby improving the detection efficiency and imaging accuracy of the entire system for fluorescent photons and accelerating the imaging process.

[0050] In the present invention, the scattering angle θ is accurately calculated by filtering the Compton scattering event data and the Compton absorption event data. i and the propagation direction of the fluorescence photons The coordinates of the intersection of the fluorescent photon and the voxel, and the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ijIt effectively suppresses noise and artifacts, accurately separates edge parts and positioning elements, and makes the position and concentration distribution of high-Z elements clearly identifiable, thereby improving the image resolution and imaging accuracy, and helping to more accurately analyze the internal structure and element distribution of the measured sample. Calculate the propagation direction of fluorescent photons The coordinates of the intersection of the fluorescent photon and the voxel, and the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij Then update the system matrix to get the system matrix element a ij , the system matrix element a ij Describes the projection probability of voxel j to the observed i-th data, the system matrix element a ij The element value can be adaptively adjusted according to the scattering angle and energy of different Compton scattering events, so as to accurately reflect the scattering, absorption, propagation and energy distribution of fluorescent photons between different voxels, thereby improving the accuracy of imaging. The intensity value f of the voxel is updated by combining the MLEM iterative formula j , which closely combines the voxel intensity update with the actual element distribution, thereby improving the spatial resolution and being able to more accurately identify the element types, especially for low-concentration elements, which can be clearly displayed and improves the ability to resolve the details of element distribution.

[0051] During the imaging process, visualization software is used to display the reconstructed image, which facilitates the observation of the element distribution, noise level, resolution, etc. in the image, and timely discovers problems and makes corresponding adjustments, ensuring that the iterative process proceeds in the direction of improving image quality, further improving the imaging quality and making the entire imaging process more controllable and optimized.

[0052] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0053] Figure 1 A schematic diagram of the structure of a Compton camera XFCT imaging system provided by the present invention;

[0054] Figure 2 A flowchart of an image reconstruction method for a Compton camera XFCT imaging system provided by the present invention;

[0055] Figure 3 A schematic diagram of the photon propagation direction and its relationship with the spatial voxel in the present invention;

[0056] Figure 4 A schematic diagram of three-dimensional volume reconstruction of a Compton camera XFCT imaging system provided by the present invention;

[0057] In the figure, 1. Radiation source; 2. Sample to be tested; 3. Scattering detector; 4. Water absorption detector. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0059] In the description of the present invention, it is necessary to understand that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0060] Example 1: Combination Figure 1-4 This embodiment is described.

[0061] This embodiment is a Compton camera XFCT imaging system, comprising:

[0062] Radiation source 1 and two sets of double-layer Compton camera detectors,

[0063] The two sets of double-layer Compton camera detectors are arranged in a straight line at intervals, the ray source 1 is located beside the straight line, the two sets of double-layer Compton camera detectors are symmetrical about the ray source 1, and the ray source 1 is a gamma ray source;

[0064] Both sets of the double-layer Compton camera detectors include a scattering detector 3 and an absorption detector 4. The scattering detector 3 is used to record the location where Compton scattering occurs and the deposited energy of the recoil electrons; the absorption detector 4 is used to receive scattered photons and record the absorption location and deposited energy.

[0065] The ray source 1 is spaced apart from the sample 2 to be tested, and two sets of double-layer Compton camera detectors are symmetrically arranged on both sides of a straight line connecting the ray source 1 and the sample 2 to be tested, and the two sets of double-layer Compton camera detectors and the sample 2 to be tested are arranged in sequence in a straight line; the gamma rays emitted by the ray source 1 enter the sample 2 to be tested, and the gamma rays interact with the elements in the sample 2 to generate fluorescent photons, and the fluorescent photons incident on the Compton camera scattering detector 3 are Compton scattered in the Compton camera scattering detector 3, and the scattered fluorescent photons are received by the corresponding Compton camera absorption detector 4 and the position where the Compton scattering occurs and the deposited energy are recorded to obtain Compton scattering event data; the fluorescent photons passing through the scattering detector 3 are absorbed by the corresponding absorption detector 4, and the absorption position and the deposited energy are recorded; the Compton absorption event data are obtained; and the Compton scattering event data and the Compton absorption event data are used to reconstruct an image.

[0066] The ray source 1 is a gamma ray source 1, and the coordinates are set to (x, y, z) = (100mm, 0mm, 0mm). The angle between the ray source 1 and the double-layer Compton camera detector is 90°, and the momentum direction is randomly distributed on the disk centered at the origin. This layout and parameter setting can ensure that the ray is stably and effectively irradiated onto the sample 2 under test.

[0067] The sample 2 under test in this scheme consists of a large cylinder and seven small cylinders. The large cylinder has a radius of 25 mm and a height of 5 mm. It is filled with air and serves as the overall frame and background environment. The seven small cylinders have a radius of 2.5 mm and a height of 5 mm. They are mixed with high Au elements and water at different concentrations, with concentrations of 2.0%, 4.0%, 6.0%, 8.0%, 10.0%, and 12.0%, respectively.

[0068] Both sets of double-layer Compton camera detectors are composed of two layers: scattering detector 3 and absorption detector 4. Scattering detector 3 is made of silicon material and has a size of 20mm×20mm×5mm. It is used to record the location of Compton scattering and the deposition energy of recoil electrons. Absorption detector 4 is made of cadmium zinc telluride material and has a size of 50mm×50mm×5mm. It is used to receive scattered fluorescent photons and record their absorption location and deposition energy. Two sets of double-layer Compton camera detectors are symmetrically arranged on both sides of the straight line connecting the radiation source 1 and the sample 2 to be measured. The distance between the scattering detector 3 and the sample 2 to be measured is 60 mm, and the distance between the scattering detector 3 and the absorption detector 4 is 20 mm. This layout enables the system to track and detect photons from more angles, obtain more effective Compton events, improve angular resolution, and significantly reduce the half-maximum width (FWHM) of the distribution, thereby improving ARM performance by 19%, and reducing the situation of missing fluorescent photons due to directional limitations, thereby reducing background interference. The system can more accurately determine the propagation direction of fluorescent photons and locate the position of the radiation source, as well as better energy resolution, thereby improving the detection efficiency and imaging accuracy of the entire system for fluorescent photons and accelerating the imaging process.

[0069] In terms of material testing, accurate element distribution imaging is crucial for evaluating material uniformity, impurity distribution and other performance indicators. A Compton camera XFCT imaging system in this solution can provide higher resolution and more accurate element distribution images, helping researchers and engineers to better understand the internal microstructure of materials and guide material research and development and production processes.

[0070] Example 2

[0071] The present invention also provides an image reconstruction method for a Compton camera XFCT imaging system, which is used in the above-mentioned Compton camera XFCT imaging system, wherein the sample to be tested is placed on a straight line where two sets of double-layer Compton camera detectors are arranged at intervals, and the two sets of double-layer Compton camera detectors are symmetrical about the straight line connecting the ray source and the sample to be tested, comprising:

[0072] S01: irradiate the gamma rays emitted by the ray source 1 to the sample 2 to be tested, and the gamma rays interact with the elements in the sample 2 to generate fluorescent photons; wherein the fluorescent photons are incident on the scattering detector 3 to generate Compton scattering events, and the fluorescent photons after Compton scattering are absorbed by the absorption detector 4 to generate Compton absorption events;

[0073] S02: Receive and record the fluorescent photons when the Compton scattering event occurs through the corresponding scattering detector 3 to obtain Compton scattering event data, wherein the Compton scattering event data includes the scattering point position (x s ,y s ,z s ) and the deposited energy E s ;

[0074] S03: Receive and record the fluorescent photons when the Compton absorption event occurs through the corresponding absorption detector 4, and obtain Compton absorption event data, wherein the Compton absorption event data includes the absorption point position (x a ,y a ,z a ) and the deposited energy E a ;

[0075] S04: filtering the Compton scattering event data and the Compton absorption event data;

[0076] S05: Calculate the angle of the filtered Compton scattering event data and the Compton absorption event data to obtain a scattering angle θ i ;

[0077] S06: dividing the space region where the sample 2 and the two sets of double-layer Compton camera detectors are located into a plurality of cubic units to obtain a voxel space, wherein each of the cubic units serves as a voxel;

[0078] S07: According to the scattering angle θ i Back-projecting the filtered Compton scattering event data and the Compton absorption event data into the voxel space;

[0079] S08: Calculating the propagation direction of the fluorescent photons according to the projection The intersection coordinates of the fluorescent photon and the voxel, the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij , and update the system matrix to obtain the system matrix element a ij ;

[0080] S09: According to the system matrix element a ij The intensity value f of the voxel is calculated using the MLEM algorithm. j Perform iterative updates to obtain a reconstruction result matrix, perform three-dimensional volume reconstruction on the reconstruction result matrix through a visualization software tool to obtain a reconstruction result image, and adjust iteration parameters in real time according to the reconstruction result image until the reconstruction result image meets preset requirements.

[0081] Turn on the ray source 1 to emit gamma rays. The gamma rays enter the sample 2 and interact with the high atomic number Au element in the sample 2, exciting the atoms to produce fluorescent photons. The fluorescent photons will be incident on the scattering detector 3 to cause a Compton scattering event. After Compton scattering, the fluorescent photons will be absorbed by the absorption detector 4 to cause a Compton absorption event. The Compton camera double-sided detectors are used to detect fluorescent photons. When the fluorescent photons enter the scattering detector 3, the scattering detector 3 records the scattering point position (x s ,y s ,z s ) and the deposited energy E s , and obtain Compton scattering event data. The fluorescent photons that have undergone Compton scattering enter the absorption detector 4 and are completely absorbed. The absorption detector 4 will record the absorption point position (x a ,y a ,z a ) and the deposited energy E a , Compton absorption event data is obtained. The obtained Compton scattering event data and Compton absorption event data are filtered to obtain data with higher purity, thereby improving the imaging accuracy of the image.

[0082] Then, from the filtered Compton scattering event data and Compton absorption event data, the energy deposition E at the scattering point and the absorption point is calculated. s and E a Calculate the scattering angle θ i , and divide the space area where the sample 2 and the two sets of double-layer Compton camera detectors are located into multiple cubic units to obtain a voxel space, wherein each of the cubic units is a voxel; according to the scattering angle θ i The filtered Compton scattering event data and Compton absorption event data are back-projected into voxel space.

[0083] Then the propagation direction of the fluorescent photons is calculated based on the projection The coordinates of the intersection of the fluorescent photon and the voxel, and the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij , and update the system matrix to obtain the system matrix element a ij , the system matrix element a ij Describes the projection probability of voxel j to the i-th observed data.

[0084] Finally, according to the system matrix element a ij The intensity value f of the voxel is calculated using the MLEM algorithm. j Perform iterative updates to obtain a reconstruction result matrix, perform three-dimensional volume reconstruction on the reconstruction result matrix using a visualization software tool to obtain a reconstruction result image, and after each iterative update, use a visualization software tool such as MATLAB to reconstruct the obtained reconstruction result matrix and display the reconstruction result image. Figure 4 As shown in the figure, the visualization software tool divides the 3D reconstruction space into a 3D array of voxels, extracts 2D slices from the array and adds colors, then uses a loop to extract isosurfaces to draw a 3D cube, and finally stores colors on the voxels to build a 3D model. In this way, the changes in the reconstruction result image in each iteration can be observed intuitively, and the 3D structure of the object or scene can be restored from the 2D projection data, making it possible to fully analyze the sample 2 under test.

[0085] During the iteration process, pay close attention to the intermediate results. After each iteration, check the changes in the area of ​​interest in the reconstructed image, including whether the shape, size, intensity and other characteristics of the area are gradually becoming stable and accurate, and whether the noise in the background area is gradually decreasing. If abnormalities are found, such as increased artifacts in certain areas and unreasonable element concentration distribution, adjust the algorithm parameters in time or check whether there are problems in the data acquisition process. If the element concentration distribution is unreasonable, check whether the system matrix calculation is accurate, or adjust the initial voxel intensity estimate. Through continuous monitoring and adjustment, ensure that the iteration process can steadily move in the direction of improving image quality.

[0086] When the reconstructed image reaches the preset stop condition, the iterative update stops and the final iterative result is output. The final two-dimensional and three-dimensional reconstruction results are displayed in the form of images through visual reconstruction, which can facilitate further data processing, analysis, and comparison with other systems or research results. In practical applications, such as in the field of early cancer diagnosis, the reconstruction results can be compared and analyzed with the element distribution characteristics of normal tissues to determine whether there is abnormal element enrichment, etc., providing an important basis for disease diagnosis; in material testing, the material uniformity, impurity distribution and other performance indicators can be evaluated according to the element distribution, guiding the material research and development and production process.

[0087] In the present invention, the scattering angle θ is accurately calculated by filtering the Compton scattering event data and the Compton absorption event data. i and the propagation direction d of the fluorescent photon, the coordinates of the intersection of the fluorescent photon and the voxel, and the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij It effectively suppresses noise and artifacts, accurately separates edge parts and positioning elements, and makes the position and concentration distribution of high-Z elements clearly identifiable, thereby improving the image resolution and imaging accuracy, and helping to more accurately analyze the internal structure and element distribution of the measured sample. Calculate the propagation direction of fluorescent photons The coordinates of the intersection of the fluorescent photon and the voxel, and the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij Then update the system matrix to get the system matrix element a ij , the system matrix element a ij Describes the projection probability of voxel j to the observed i-th data, the system matrix element a ij The element value can be adaptively adjusted according to the scattering angle and energy of different Compton scattering events, so as to accurately reflect the scattering, absorption, propagation and energy distribution of fluorescent photons between different voxels, thereby improving the accuracy of imaging. The intensity value f of the voxel is updated by combining the MLEM iterative formula j , which closely combines the voxel intensity update with the actual element distribution, thereby improving the spatial resolution and being able to more accurately identify the element types, especially for low-concentration elements, which can be clearly displayed and improves the ability to resolve the details of element distribution.

[0088] During the imaging process, visualization software is used to display the reconstructed image, which facilitates the observation of the element distribution, noise level, resolution, etc. in the image, and timely discovers problems and makes corresponding adjustments, ensuring that the iterative process proceeds in the direction of improving image quality, further improving the imaging quality and making the entire imaging process more controllable and optimized.

[0089] In medical applications such as early cancer diagnosis, more accurate and efficient imaging technology is needed to detect changes in element distribution within tissues. This solution improves the imaging efficiency and quality of the CC-XFCT system, and can more accurately display the distribution of high atomic number elements in the body, providing a more reliable basis for early cancer diagnosis, improving the accuracy and timeliness of disease diagnosis, and thus improving patient treatment outcomes and prognosis.

[0090] Further, filtering the Compton scattering event data and the Compton absorption event data includes:

[0091] Calculate the scattering point position (x s ,y s ,z s ) and the absorption point position (x a ,y a ,z a ) s-a , the distance d s-a The distance threshold d th For comparison, if d s-a <d th , then the corresponding scattering point position (x s ,y s ,z s ) and the corresponding absorption point position (x a ,y a ,z a ) are deleted from the Compton scattering event data and the Compton absorption event data, respectively.

[0092] The obtained Compton scattering event data and Compton absorption event data are filtered, and for each event, the scattering point position (x s ,y s ,z s ) and the absorption point position (x a ,y a ,z a ) s-a According to the Compton scattering characteristics and the noise statistics obtained through experiments and theoretical analysis, a suitable distance threshold d is pre-set. th , the distance d s-a The distance threshold d th For comparison, when d s-a <d th When the event is determined to be a noise event or an interference event, the corresponding scattering point position (x s ,y s ,zs ) and the corresponding absorption point position (x a ,y a ,z a ) are deleted from the Compton scattering event data and the Compton absorption event data respectively. After distance filtering, data with higher purity is obtained, thereby improving the imaging accuracy of the image.

[0093] Further, S05: performing angle calculation on the filtered Compton scattering event data and the Compton absorption event data to obtain a scattering angle θ i ,include:

[0094] From the filtered Compton scattering event data and Compton absorption event data, the scattering angle θ is calculated using the Compton scattering angle calculation formula according to the deposited energy Es at the scattering point and the deposited energy Ea at the absorption point. i , the formula is:

[0095] Among them, m e c 2 Represents the electron rest energy.

[0096] Further, S08: calculating the propagation direction of the fluorescent photon according to the projection The intersection coordinates of the fluorescent photon and the voxel, the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij , and update the system matrix to obtain the system matrix element a ij ,include:

[0097] Calculate the fluorescence photon from the scattering point position (x s ,y s ,z s ) to the absorption point position (x a ,y a ,z a ) direction vector The formula is: For the direction vector Normalized processing is performed to obtain the propagation direction of the fluorescence photon The formula is:

[0098]

[0099] Assume the coordinates of the scattering point are (x s ,y s ,z s ), the absorption point coordinates are (x a ,y a ,z a ), then the fluorescence photon is scattered from the position (xs ,y s ,z s ) to the absorption point position (x a ,y a ,z a ) direction vector for For this direction vector After normalization, we get the unit vector representing the propagation direction of the fluorescence photon:

[0100] Calculate the intersection coordinates of the fluorescent photon and the voxel, where the position of the voxel can be expressed as (x j ,y j ,z j );

[0101] The range of the voxel in the X-axis, Y-axis, and Z-axis directions is calculated by the range formula: j1 =X j +0.5×ΔX,X j2 =X j +1×ΔX,Y j1 =Y j +0.5×ΔY,Y j2 =Y j +1×ΔY,Z j1 =Z j +0.5×ΔZ,Z j2 =Z j +1×ΔZ, where X j1 and X j2 Indicates the range of the voxel in the X-axis direction, Y j1 and Y j2 Indicates the range of the voxel in the Y-axis direction, Z j1 and Z j2 represents the range of the voxel in the Z-axis direction, ΔX, ΔY, ΔZ represent the variable range of the voxel in the X-axis, Y-axis and Z-axis respectively, and j represents the index of the voxel;

[0102] Photon propagation formula The value range of parameter t on the X-axis, Y-axis and Z-axis is calculated by the voxel boundary inequality, and the formula is:

[0103] X0+td x ∈[X j1 ,X j2 ],Y0+td y ∈[Y j1 ,Y j2 ],Z0+td z ∈[Z j1 ,Z j2 ],in, represents the position vector of the fluorescent photon at the initial position, X0, Y0, and X0 represent The components on the X-axis, Y-axis, and Z-axis, d x ,d y ,d z Respectively Components on the X-axis, Y-axis, and Z-axis;

[0104] Substitute the value range of parameter t on the X-axis, Y-axis and Z-axis into the photon propagation formula respectively Calculation is performed to obtain the coordinates of the intersection of the fluorescent photon and the voxel.

[0105] Determine the intersection of the fluorescent photon and the voxel. For each voxel, its position can be expressed as (x j ,y j ,z j ). Taking the X axis as an example, the range of the voxel in the X axis direction is X j1 and X j2 They are

[0106] X j1 =X j +0.5×ΔX,X j2 =X j +1×ΔX, and the photon propagation formula Joint,

[0107] And the photon propagation equation The X-axis component X0, The X-axis component d x Substituting into the voxel boundary inequality, we get X0+td x ∈[X j1 ,X j2 ], solve the inequality to get the range of the parameter t, and perform similar calculations in the Y-axis and Z-axis directions. If there is a t value that satisfies the boundary inequality in all three axis directions, it is determined that this fluorescent photon intersects with the voxel, and then the coordinates of the intersection are obtained by substituting the t value into the photon propagation formula.

[0108] Calculate the path length d of the fluorescent photon in the voxel according to the intersection coordinates ij , the formula is:

[0109] in, and represents the boundary intersection point when the fluorescent photon intersects with the voxel;

[0110] According to the Beer-Lambert law, the intensity attenuation of the fluorescent photon after passing through the voxel and the energy deposition ratio of the fluorescent photon in the voxel are calculated. ij , the formula is:

[0111] Where I0 and I represent the energy of the fluorescence photon before and after decay in the voxel, μ j represents the attenuation coefficient of fluorescence photons in the voxel;

[0112] According to the propagation direction Intersection coordinates, path length d ij and the energy deposition ratio e ij Update the system matrix to get the system matrix element a ij , the formula is:

[0113] Further, according to the system matrix element a ij The intensity value f of the voxel is calculated using the MLEM algorithm. j Perform iterative updates to obtain the reconstruction result matrix, including:

[0114] Initialize algorithm parameters, including the number of iterations n, the initial voxel intensity estimate f j (0) , assign an initial value to the weight of each voxel in the voxel space, and iteratively update the voxel intensity value f of the voxel in combination with the MLEM algorithm j , the formula is:

[0115] Wherein, i represents the index of the observation data detected by the double-layer Compton camera detector, j represents the index of the voxel, and f j represents the intensity value of voxel j, f j (n) represents the intensity value of voxel j at n iterations, f j (n+1) represents the intensity value of voxel j at the next iteration, and m represents the upper limit of the observed data.

[0116] Update the voxel intensity value f by combining the MLEM iterative formula j , which closely combines the voxel intensity update with the actual element distribution, thereby improving the spatial resolution and being able to more accurately identify the element types, especially for low-concentration elements, which can be clearly displayed and improves the ability to resolve the details of element distribution.

[0117] Furthermore, adjusting the iteration parameters in real time according to the reconstructed image until the reconstructed image meets the preset requirements includes:

[0118] Adjust the preset distance threshold d th : If artifacts occur, increase the preset distance threshold d th If the number of noise events increases, the preset distance threshold d is reduced.th , where the scattering point position (x s ,y s ,z s ) and the absorption point position (x a ,y a ,z a ) s-a Less than the preset distance threshold d th The event is judged as a noise event;

[0119] Adjust the attenuation coefficient μ of the fluorescence photons in the voxel j Value: Set different attenuation coefficients μ according to different elements j Numeric value;

[0120] Adjust the initial voxel intensity estimate f according to the element concentration in the reconstructed image j (0) .

[0121] During the iteration process, pay close attention to the intermediate results. After each iteration, check the changes in the area of ​​interest in the reconstructed image, including whether the shape, size, intensity and other characteristics of the area are gradually becoming stable and accurate, and whether the noise in the background area is gradually decreasing. If abnormalities are found, such as increased artifacts in certain areas, unreasonable element concentration distribution, etc., adjust the iteration parameters or check whether there are problems in the data acquisition process. Adjustments are mainly made in the following aspects:

[0122] Adjust the preset distance threshold d th :By adjusting the appropriate distance threshold d th , the scattering point position (x s ,y s ,z s ) and the absorption point position (x a ,y a ,z a ) s-a Less than the preset distance threshold d th The event is judged as a noise event or an interference event, and the corresponding scattering point position (x s ,y s ,z s ) and the corresponding absorption point position (x a ,y a ,z a ) are deleted from the Compton scattering event data and the Compton absorption event data, respectively, to improve the data purity. If a large number of Compton scattering events and Compton absorption events are misjudged as noise events or interference events, it will lead to image information loss and artifacts. In this case, the preset distance threshold d should be appropriately relaxed. thOn the contrary, if noise events or interference events still exist in large quantities, the image quality will be affected. In this case, the preset distance threshold d can be reduced. th To improve the image quality. Each time the distance threshold d is adjusted th After that, it is necessary to re-filter the data and perform subsequent iterative calculations, and observe the changes in image quality to determine the optimal distance threshold d. th .

[0123] Adjust the attenuation coefficient μ of the fluorescence photons in the voxel j Value: Attenuation coefficient μ when different elements are involved j The values ​​are different. You need to re-check the relevant physics manual or experimental data and use the corresponding attenuation coefficient μ j Numeric value.

[0124] Adjust the initial voxel intensity estimate f according to the element concentration in the reconstructed image j (0) : As the starting point of the iterative algorithm, the initial voxel intensity estimate f j (0) It has an important impact on the subsequent iteration process and the final imaging result. If it is found during the iteration process that the element concentration distribution is significantly different from the actual situation, the initial voxel intensity estimate f needs to be adjusted. j (0) For example, if it is known that a certain area should contain a certain element with a high concentration, and the current reconstruction result image shows that the element concentration in this area is too low, the f j (0) From the low estimate (such as 0.1) to a more reasonable value (such as 1.0 or higher), and then restart the iteration. If the approximate range of the element concentration in the measured sample 2 is known, and the current initial voxel intensity estimate f j (0) If it deviates significantly from this range, it will be corrected according to the actual situation to make it closer to this range. After adjustment, iterative calculation will be performed again to observe the improvement of element concentration distribution and ensure that the adjusted initial voxel intensity estimate f j (0) It can guide iterations in the right direction, thereby improving imaging accuracy.

[0125] It should also be noted that the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of further restrictions, the elements defined by the sentence "includes one..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements. The words "first", "second" and the like are used to indicate names, but do not indicate any specific order. The above schematically describes the invention and its implementation methods, which is not restrictive. The invention can be implemented in other specific forms without departing from the spirit or basic features of the invention. What is shown in the accompanying drawings is only one of the implementation methods of the invention, and the actual structure is not limited thereto. Any figure mark in the claims should not limit the claims involved. Therefore, if a person of ordinary skill in the art is inspired by it, and does not deviate from the purpose of the invention, and designs a structural mode and an embodiment similar to the technical solution without creativity, they should all belong to the scope of protection of this patent.

Claims

1. A Compton camera XFCT imaging system, characterized in that: include: X-ray source and two sets of double-layer Compton camera detectors. The two sets of double-layer Compton camera detectors are arranged in a straight line at intervals, the ray source is located beside the straight line, the two sets of double-layer Compton camera detectors are symmetrical about the ray source, and the ray source is a gamma ray source; Both sets of double-layer Compton camera detectors include scattering detectors and absorption detectors. The scattering detectors are used to record the location where Compton scattering occurs and the deposited energy of the recoil electrons; the absorption detectors are used to receive scattered photons and record the absorption location and deposited energy.

2. A Compton camera XFCT imaging system according to claim 1, characterized in that: The distance between the scattering detector and the sample to be measured is 60 mm, and the distance between the scattering detector and the absorption detector is 20 mm.

3. A Compton camera XFCT imaging system according to any one of claims 1-2, characterized in that: The scattering detector is made of silicon and has a size of 20 mm×20 mm×5 mm.

4. A Compton camera XFCT imaging system according to any one of claims 1-2, characterized in that: The absorption detector is made of cadmium zinc telluride and has a size of 50 mm×50 mm×5 mm.

5. A Compton camera XFCT imaging system image reconstruction method, characterized in that: A Compton camera XFCT imaging system for use in any one of claims 1 to 4, wherein the sample to be tested is placed on a straight line where two sets of double-layer Compton camera detectors are arranged at intervals, and the two sets of double-layer Compton camera detectors are symmetrical about a straight line connecting the ray source and the sample to be tested, comprising: S01: irradiating the gamma rays emitted by the ray source to the sample to be tested, and the gamma rays interact with the elements in the sample to be tested to generate fluorescent photons; wherein the fluorescent photons are incident on the scattering detector to generate Compton scattering events, and the fluorescent photons after Compton scattering are absorbed by the absorption detector to generate Compton absorption events; S02: Receive and record the fluorescent photons when the Compton scattering event occurs through the corresponding scattering detector to obtain Compton scattering event data, wherein the Compton scattering event data includes the scattering point position (x s ,y s ,z s ) and the deposited energy E s ; S03: Receive and record the fluorescent photons when the Compton absorption event occurs through the corresponding absorption detector to obtain Compton absorption event data, wherein the Compton absorption event data includes the absorption point position (x a ,y a ,z a ) and the deposited energy E a ; S04: filtering the Compton scattering event data and the Compton absorption event data; S05: Calculate the angle of the filtered Compton scattering event data and the Compton absorption event data to obtain a scattering angle θ i ; S06: dividing the space region where the sample to be tested and the two sets of double-layer Compton camera detectors are located into a plurality of cubic units to obtain a voxel space, wherein each of the cubic units serves as a voxel; S07: According to the scattering angle θ i Back-projecting the filtered Compton scattering event data and the Compton absorption event data into the voxel space; S08: Calculating the propagation direction of the fluorescent photons according to the projection The intersection coordinates of the fluorescent photon and the voxel, the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij , and update the system matrix to obtain the system matrix element a ij ; S09: According to the system matrix element a ij The intensity value f of the voxel is calculated using the MLEM algorithm. j Perform iterative updates to obtain a reconstruction result matrix, perform three-dimensional volume reconstruction on the reconstruction result matrix through a visualization software tool to obtain a reconstruction result image, and adjust iteration parameters in real time according to the reconstruction result image until the reconstruction result image meets preset requirements.

6. The image reconstruction method of a Compton camera XFCT imaging system according to claim 5, characterized in that: Filtering the Compton scattering event data and the Compton absorption event data comprises: Calculate the scattering point position (x s ,y s ,z s ) and the absorption point position (x a ,y a ,z a ) s-a , the distance d s-a The distance threshold d th For comparison, if d s-a <d th , then the corresponding scattering point position (x s ,y s ,z s ) and the corresponding absorption point position (x a ,y a ,z a ) are deleted from the Compton scattering event data and the Compton absorption event data respectively.

7. The image reconstruction method of a Compton camera XFCT imaging system according to claim 6, characterized in that: S05: Calculate the angle of the filtered Compton scattering event data and the Compton absorption event data to obtain a scattering angle θ i ,include: From the filtered Compton scattering event data and Compton absorption event data, the scattering angle θ is calculated using the Compton scattering angle calculation formula according to the deposited energy Es at the scattering point and the deposited energy Ea at the absorption point. i , the formula is: Among them, m e c 2 Represents the electron rest energy.

8. The image reconstruction method of a Compton camera XFCT imaging system according to claim 7, characterized in that: S08: Calculating the propagation direction of the fluorescent photons according to the projection The intersection coordinates of the fluorescent photon and the voxel, the path length d of the fluorescent photon in the voxel ij and the energy deposition ratio e ij , and update the system matrix to obtain the system matrix element a ij ,include: Calculate the fluorescence photon from the scattering point position (x s ,y s ,z s ) to the absorption point position (x a ,y a ,z a ) direction vector The formula is: For the direction vector Normalized processing is performed to obtain the propagation direction of the fluorescence photon The formula is: Calculate the intersection coordinates of the fluorescent photon and the voxel, where the position of the voxel can be expressed as (x j ,y j ,z j ); The range of the voxel in the X-axis, Y-axis, and Z-axis directions is calculated by the range formula: j1 =X j +0.5×ΔX,X j2 =X j +1×ΔX,Y j1 =Y j +0.5×ΔY,Y j2 =Y j +1×ΔY,Z j1 =Z j +0.5×ΔZ,Z j2 =Z j +1×ΔZ, where X j1 and X j2 Indicates the range of the voxel in the X-axis direction, Y j1 and Y j2 Indicates the range of the voxel in the Y-axis direction, Z j1 and Z j2 represents the range of the voxel in the Z-axis direction, ΔX, ΔY, ΔZ represent the variable range of the voxel in the X-axis, Y-axis and Z-axis respectively, and j represents the index of the voxel; Photon propagation formula The value range of parameter t on the X-axis, Y-axis and Z-axis is calculated by the voxel boundary inequality, and the formula is: X0+td x ∈[X j1 ,X j2 ],Y0+td y ∈[Y j1 ,Y j2 ],Z0+td z ∈[Z j1 ,Z j2 ],in, represents the position vector of the fluorescent photon at the initial position, X0, Y0, and X0 represent The components on the X-axis, Y-axis, and Z-axis, d x d y d z Respectively Components on the X-axis, Y-axis, and Z-axis; Substitute the value range of parameter t on the X-axis, Y-axis and Z-axis into the photon propagation formula respectively Performing calculations to obtain the coordinates of the intersection of the fluorescent photon and the voxel; Calculate the path length d of the fluorescent photon in the voxel according to the intersection coordinates ij , the formula is: in, and represents the boundary intersection point when the fluorescent photon intersects with the voxel; According to the Beer-Lambert law, the intensity attenuation of the fluorescent photon after passing through the voxel and the energy deposition ratio of the fluorescent photon in the voxel are calculated. ij , the formula is: Where I0 and I represent the energy of the fluorescence photon before and after decay in the voxel, μ j represents the attenuation coefficient of fluorescence photons in the voxel; According to the propagation direction Intersection coordinates, path length d ij and the energy deposition ratio e ij Update the system matrix to get the system matrix element a ij , the formula is:

9. The image reconstruction method of a Compton camera XFCT imaging system according to claim 8, characterized in that: According to the system matrix element a ij The intensity value f of the voxel is calculated using the MLEM algorithm. j Perform iterative updates to obtain the reconstruction result matrix, including: Initialize algorithm parameters, including the number of iterations n, the initial voxel intensity estimate f j (0) , assign an initial value to the weight of each voxel in the voxel space, and iteratively update the voxel intensity value f of the voxel in combination with the MLEM algorithm j , the formula is: Wherein, i represents the index of the observation data detected by the double-layer Compton camera detector, j represents the index of the voxel, and f j represents the intensity value of voxel j, f j (n) represents the intensity value of voxel j at n iterations, f j (n+1) represents the intensity value of voxel j at the next iteration, and m represents the upper limit of the observed data.

10. The image reconstruction method of a Compton camera XFCT imaging system according to claim 9, characterized in that: Adjusting iteration parameters in real time according to the reconstructed image until the reconstructed image meets preset requirements includes: Adjust the preset distance threshold d th : If artifacts occur, increase the preset distance threshold d th If the number of noise events increases, the preset distance threshold d is reduced. th , where the scattering point position (x s ,y s ,z s ) and the absorption point position (x a ,y a ,z a ) s-a , is less than the preset distance threshold d th The event is judged as a noise event; Adjust the attenuation coefficient μ of the fluorescence photons in the voxel j Value: Set different attenuation coefficients μ according to different elements j Numeric value; Adjust the initial voxel intensity estimate f according to the element concentration in the reconstructed image j (0) .

Images