Spherical spiral track CBCL imaging method and FDK reconstruction method

Abstract

The invention relates to a spherical spiral track CBCL imaging method and an FDK reconstruction method.In spherical spiral CBCL imaging, an imaging sample is fixed, a ray source and a detector rotate around a rotating shaft, and the CL angle is changed by adjusting the positions of the X-ray source and the detector; or, the imaging sample rotates around the rotating shaft, and the CL angle is adjusted by adjusting the positions of the X-ray source and the detector; alternatively, the X-ray source and the detector are fixed, the sample rotates around the rotation axis, and the CL angle is changed by adjusting the tilt angle of the rotation axis. According to the method, during spherical spiral track CBCL imaging, namely circular track rotation, the CL angle is gradually reduced, the imaging data is equivalent to adding small dip angle data on the basis of circular track CL data, a circular region with frequency domain section missing information is reduced to a spiral region, missing data information is effectively supplemented, and thus the interlayer aliasing artifacts of a reconstructed image are reduced.

Description

Technical Field

[0001] The present invention relates to the field of computer tomography, and in particular to a spherical spiral trajectory CBCL imaging method and an FDK reconstruction method. Background Art

[0002] Computed tomography (CT) is a widely used non-destructive testing technology. During CT imaging, the sample rotates 360° around the rotation axis, and projection data at various angles are collected. Then, a reconstruction algorithm is used to obtain a three-dimensional image of the internal structure of the object. However, when inspecting large plate-like objects (such as PCBs, etc.), when the X-rays are nearly parallel to the plate surface, the attenuation path is too long and the rays are difficult to penetrate, resulting in missing data at certain angles. In addition, the CT imaging structure limits the minimum distance between the rotation center of large plate-like objects and the X-ray source, which limits the imaging resolution of the object. Therefore, it is difficult for CT to achieve high-resolution imaging of large plate-like objects. Cone-beam Computed Laminography (CBCL) provides a solution to the imaging problem of large plate-like objects. The CBCL imaging trajectory can be divided into translational and rotational types. Among them, the circular trajectory rotational CBCL is the most widely used CL imaging method. The main differences between the circular trajectory CBCL and CBCT systems are as follows: Figure 1 As shown, the rotation axis of a CBCT system is parallel to the detector plane and perpendicular to the primary beam. In contrast, in a CBCL system, the rotation axis is tilted at an angle a to the detector plane, and the angle between the rotation axis and the primary beam is 90° - a. The tilted rotation axis geometry of the CBCL system overcomes object size limitations when scanning large, planar objects, shortening the distance between the rotation axis and the radiation source and achieving higher resolution. Furthermore, when imaging planar objects, radiation consistently passes only through the thickness of the plate within a 360-degree radius, resulting in a stable grayscale distribution across all projected angles. Therefore, CBCL is suitable for high-resolution imaging of large, planar samples.

[0003] Analytical reconstruction technology has always been the mainstream reconstruction method for CBCL imaging due to its advantage of fast reconstruction speed. CBCL analytical reconstruction methods include direct Fourier reconstruction, FBP-based reconstruction, FDK-based reconstruction, projection transformation followed by FDK reconstruction, etc. However, the tilted rotation axis of CL imaging will lead to information loss in the imaging data. The loss of CBCL data will cause the information in adjacent slices of the reconstructed image to be mixed, resulting in aliasing artifacts in the reconstructed image. Figure 2 As shown in Figure 3, the larger the CL angle, the more severe the aliasing artifact.

[0004] Suppressing aliasing artifacts has long been a challenging and hot topic in CBCL imaging research. Research efforts have focused on three key areas: improving analytical reconstruction methods, employing iterative approaches, and supplementing with additional scan data. Because the fundamental issue of missing data remains unresolved, analytically based methods have achieved limited success. Addressing the ill-posed problem caused by missing data, iterative reconstruction methods can introduce constraints or prior knowledge into the CBCL reconstruction process to reduce aliasing artifacts. However, iterative reconstruction methods are computationally intensive, time-consuming, and limited to specific samples, limiting their versatility and practicality. Complementing the missing CBCL data by adding data from other scan methods, such as CT scans, linear scans, and limited-angle scans, can reduce aliasing artifacts in the reconstructed image. However, this requires performing two scans, which adds an additional workload. Supplementing the missing CBCL data without adding additional workload is crucial for suppressing aliasing artifacts. Summary of the Invention

[0005] In response to the above problems, in the first aspect of the present invention, a spherical spiral trajectory CBCL imaging method is provided. In spherical spiral CBCL imaging, the imaging sample is fixed, the X-ray source and the detector rotate around the rotation axis, and the CL angle is changed by adjusting the positions of the X-ray source and the detector; or, the imaging sample rotates around the rotation axis, and the CL angle is adjusted by adjusting the positions of the X-ray source and the detector; or, the X-ray source and the detector are fixed, the sample rotates around the rotation axis, and the CL angle is changed by adjusting the inclination angle of the rotation axis.

[0006] Preferably, the distances between the ray source, the detector and the rotation center remain unchanged.

[0007] In a second aspect of the present invention, a spherical spiral trajectory CBCL imaging FDK reconstruction method is provided, the method comprising the following steps:

[0008] S1, obtain projection data and perform weighted filtering on the projection data:

[0009]

[0010] Where p(θ,a,u,v) is the projection data when the projection angle is θ, θ ranges from 0° to 360°, D is the distance from the ray source to the detector plane, u and v are the coordinates of the detector, h(u′-u) is the filter function, and a is the tilt angle;

[0011] S2, perform spatial transformation on the reconstructed volume data:

[0012] [x′,y′,z′+=[x,y,z]R(a,Δa,θ);

[0013] Among them, x, y, z are the coordinates of the spatial transformation precursor data, R(a, Δa, θ) is the rotation factor, and R(a, Δa, θ) is specifically:

[0014]

[0015] S3, back-project the volume data:

[0016]

[0017] S4, traverse each rotation angle θ, repeat steps S1-S3, and accumulate the back-projected volume data:

[0018]

[0019] Preferably, the method is applied to the reconstruction of projection data acquired as described in the first aspect.

[0020] In a third aspect of the present invention, a spherical spiral trajectory CBCL imaging FDK reconstruction system is provided, the system comprising the following modules:

[0021] The weighted filtering module is used to obtain projection data and perform weighted filtering on the projection data:

[0022]

[0023] Where p(θ,a,u,v) is the projection data when the projection angle is θ, θ ranges from 0° to 360°, D is the distance from the ray source to the detector plane, u and v are the coordinates of the detector, h(u′-u) is the filter function, and a is the tilt angle;

[0024] The spatial transformation module is used to reconstruct volume data and perform spatial transformation:

[0025] [x′,y′,z′+=[x,y,z]R(a,Δa,θ);

[0026] Among them, x, y, z are the coordinates of the volume data before spatial transformation, and x′, y′, z′ are the coordinates of the volume data after spatial transformation.

[0027] R(a,Δa,θ) is the rotation factor, and R(a,Δa,θ) is specifically:

[0028]

[0029] Back projection module, used to back project volume data:

[0030]

[0031] The accumulation module is used to traverse each rotation angle θ, repeat steps S1-S3, and accumulate the back-projected volume data:

[0032]

[0033] Preferably, the system is applied to reconstruct projection data acquired as described in the first aspect.

[0034] This invention uses SHCL imaging to reduce missing data from circular regions of frequency domain slices to spiral regions by supplementing low-angle data, effectively supplementing the missing data information in CL imaging. Furthermore, the FDK reconstruction algorithm is improved to address the trajectory characteristics of SHCL, enabling SHCL image reconstruction and reducing inter-slice aliasing artifacts in CL imaging. The SHCL imaging trajectory provided by this invention is simple and easy to implement without increasing the scanning workload. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 Schematic diagram of the imaging structure of the circular trajectory CBCL and CBCT system;

[0036] Figure 2 Schematic diagram of CL aliasing artifact;

[0037] Figure 3 Schematic diagram of the comparison between CBCL and parallel beam CL projection;

[0038] Figure 4 Schematic diagram of the trajectory of different imaging methods;

[0039] Figure 5 Schematic diagram of the transformation of the information missing area as the CL angle decreases;

[0040] Figure 6 Schematic diagram of the amount of data supplemented for SHCL;

[0041] Figure 7 Schematic diagram of the method for realizing the spherical spiral trajectory;

[0042] Figure 8 Schematic diagram of space transformation;

[0043] Figure 9 Schematic diagram of the circular phantom and multi-layer phantom for simulation experiments;

[0044] Figure 10 Reconstructed images and cross-sectional images of the circular phantom;

[0045] Figure 11 The horizontal slice (50th layer) and longitudinal slice of the reconstructed image;

[0046] Figure 12 It is a cross-sectional view of the transverse section;

[0047] Figure 13 It is a schematic diagram of image frequency domain information;

[0048] Figure 14 It is a self-made 4-layer PCB circuit board;

[0049] Figure 15 Reconstruct the image for the entire PCB;

[0050] Figure 16 Comparison of local reconstructed images of 6°CL and 6°-30°SHCL PCB;

[0051] Figure 17 Comparison of local reconstructed images of 30°CL and 6°-30°CL PCB. DETAILED DESCRIPTION

[0052] In the embodiments of the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of this application should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner to facilitate understanding.

[0053] It will be understood that the “embodiment” mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, the various embodiments throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It will be understood that in the various embodiments of the present application, the size of the sequence number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application.

[0054] In the present invention, unless otherwise specified, the same or similar parts between the various embodiments can refer to each other. In the various embodiments of the present invention, and the various implementation methods / implementation methods / implementation methods in each embodiment, if there is no special explanation and logical conflict, the terms and / or descriptions between different embodiments and the various implementation methods / implementation methods / implementation methods in each embodiment are consistent and can be referenced to each other. The technical features in different embodiments and the various implementation methods / implementation methods / implementation methods in each embodiment can be combined to form new embodiments, implementation methods, implementation methods, or implementation methods according to their inherent logical relationships. The implementation methods of the present application described below do not constitute a limitation on the scope of protection of the present application.

[0055] According to the 3D Fourier slice theorem, the Fourier transform of a 3D function f(x, y, z) at a certain angle is equal to the slice of the 3D Fourier transform of the original 3D function on a plane perpendicular to the projection direction. Figure 3 As shown in (a), the 2D Fourier transform of the object at angle θ corresponds to the slice of the 3D Fourier transform of the object at angle θ. When θ takes a full circle (0°-360°), the frequency domain information of the object is complete. The frequency domain information of the parallel beam CL projection is as follows: Figure 3 As shown in (b), the tilt of the rotation axis leads to the tilt of the Fourier slice, where a represents the CL angle. When the angle θ takes a full circle (0°-360°), the frequency domain space constructed by the projected Fourier transform has a conical area with missing frequency domain information as shown in the green area. The larger the CL angle a is, the more missing information there is.

[0056] The difference between cone-beam CL projection and parallel beam is that the frequency domain information corresponding to parallel beam two-dimensional projection is planar, while the frequency domain information corresponding to cone-beam projection is arc-shaped, such as Figure 3 As shown in (c), O is the rotation center and R is the distance from the ray source to the rotation center. The missing information in the frequency domain space composed of 360° CBCL projection is as follows: Figure 3 As shown in (d), the yellow area is the missing data area in the three-dimensional frequency domain space. Similar to the parallel beam projection frequency domain space, the larger the CL angle a, the larger the radius r of the data missing area, and the more data missing. The data of the transverse central slice of the CBCL frequency domain is complete, and there is information missing in the circular area of the non-central slice. The farther away from the central slice, the larger the circular area of missing information.

[0057] In CT imaging, the radiation source and detector are fixed, and the object rotates around the rotation axis, such as Figure 4 As shown in (a), it is equivalent to the object being stationary, and the ray source and detector move in a circular trajectory relative to the object, as shown in Figure 4 For ease of analysis, let the distance from the object's rotation center to the ray source and the distance from the rotation center to the detector be R. Then, the trajectories of the ray source and the detector in 4(d) are both circles with a radius of R.

[0058] In circular trajectory CL imaging, if the CL angle is a, it is equivalent to the object being stationary, and the ray source and detector are moving in circular trajectories above and below the object, respectively, as shown in Figure 2. Figure 4 As shown in (b) and (e), the radius of the circle is R*cosa, and the distance between the circle plane and the object plane is R*sina. Therefore, the larger the CL angle, the smaller the rotation radius r of the ray source, and the farther the ray source trajectory plane is from the object plane, while the distance from the ray source to the rotation center is always R.

[0059] From the frequency domain analysis of CL data, it can be seen that different CL angles have different missing information. When performing CL imaging, it is usually difficult to determine the optimal CL angle for different samples. This paper proposes spherical spiral trajectory CL imaging with multi-angle data fusion to supplement the missing data information. Spherical spiral trajectory imaging means that in CL imaging, the CL angle is gradually reduced while the rotation axis rotates, such as Figure 4 As shown in (c), it is equivalent to the object being stationary, and the ray source and detector moving around the object in a spherical spiral trajectory, as shown in Figure 4 (f) is shown. The spherical radius of the spherical spiral trajectory is the distance R from the source and detector to the sample center. The CL angle of any point on the trajectory is the angle between the line connecting the point and the rotation center and the object center plane. The object center plane passes through the rotation center and is perpendicular to the rotation axis. If the rotation axis rotates 360 degrees, its imaging trajectory is one cycle of the spherical spiral, as shown in Figure 4 (g) is also Figure 4 (f) Part of the trajectory marked in yellow.

[0060] In spherical spiral trajectory CL imaging, since the CL angle gradually decreases during the 360-degree projection, the area where the frequency domain information is missing will be a spiral area with a gradually decreasing radius. Figure 5 As shown in (a), SHCL can reduce the information missing area of the circular trajectory CL from a cone to a spiral cone area marked in yellow. Its frequency domain cross-slice information is as follows Figure 5 As shown in (b), the circular areas are all areas where circular trajectory CL information is missing, the red areas are areas where spiral trajectory CL information is missing, and the blue areas are areas where information is supplemented by spiral circular trajectory CL. It can be seen that SHCL can significantly supplement the missing CL data. In one embodiment, Δa is reduced each time. In another embodiment, the reduced angle is related to time. Preferably, a(t) = a0 - Δa*t, where a(t) is the CL angle at time t during the scan, a0 is the CL angle at the initial time, and t is time.

[0061] During SHCL imaging, the CL angle gradually decreases. In theory, the smaller the minimum spiral inclination angle, the more supplementary information is obtained. However, in actual imaging, due to the limitations of CL equipment and samples, small-angle imaging near 0 degrees cannot be performed. In addition, when projecting at small angles, hardening artifacts will occur because the rays cannot penetrate the plate-like objects. Therefore, the minimum spiral inclination angle will vary depending on the specific equipment and sample. If the minimum spiral inclination angle is 5°, within the commonly used CL angle range of 10° to 45°, Figure 6 The amount of missing data, amount of SHCL supplementary data and data supplementation rate at different CL angles were quantified. The missing data were normalized to 45 degrees. Figure 6It can be seen that within the CL angle range of 10°-45°, SHCL can supplement about 42-69% of the missing CL data. In the present invention, unless otherwise specified, CL imaging refers to circular trajectory CBCL imaging.

[0062] The spherical spiral CL trajectory can be achieved by changing the CL angle while rotating. According to the specific structure of the CL imaging device, there are three specific implementation methods: one is to fix the imaging sample, rotate the X-ray source and detector around the rotation axis, and change the CL angle by adjusting the position of the X-ray source and detector, such as Figure 7 (a) As shown. The second is that the imaging sample rotates around the rotation axis and the CL angle is adjusted by adjusting the position of the X-ray source and detector, as shown in Figure 7 (b) As shown in the third example, the X-ray source and detector are fixed, the sample rotates around the rotation axis, and the CL angle is changed by adjusting the tilt angle of the rotation axis, as shown in Figure 7 (c) In the above three implementations, the distances between the ray source, detector and the rotation center remain unchanged.

[0063] The FDK algorithm is simple and fast, and is the most commonly used analytical reconstruction algorithm for CT image reconstruction. In its algorithm formula (1), f(x, y, z) is the reconstructed three-dimensional volume data, p(θ, u, v) is the projection data at an angle of θ, and the value range of θ is 0°-360°.

[0064]

[0065] The spherical spiral trajectory FDK reconstruction algorithm is based on the FDK reconstruction algorithm and is improved for spherical spiral trajectory. It can be divided into four steps: weighted filtering, spatial transformation, back projection, and superposition. The specific steps are as follows:

[0066] Step 1. Weighted filtering

[0067]

[0068] Step 2. Spatial transformation

[0069] This step is an improvement on the spherical spiral trajectory based on the FDK reconstruction of CT images. In the CT back-projection direction, only the rotation angle θ changes within the range of 360 degrees, while the CL back-projection direction has an additional tilt angle a on the basis of the rotation angle θ. The tilt angle Δa of the spherical spiral trajectory CL is also a variable value, such as Figure 8 As shown, before back-projection, the back-projected data space needs to be transformed, as shown in formula (3). Where R(a, Δa, θ) is the rotation factor, which is determined by the starting CL angle a, the spiral CL angle change Δa and the rotation angle θ, as shown in formula (4).

[0070] [x′,y′,z′]=[x,y,z]R(a,Δa,θ) (3)

[0071]

[0072] Step 3. Back-projection

[0073]

[0074] Step 4. Accumulation

[0075] Traverse each rotation angle θ, repeat steps 1-4, and accumulate the back-projected volume data.

[0076]

[0077] Simulation experiments were conducted on a circular phantom and a multilayer phantom, and imaging experiments were performed on a custom-made PCB. The experiments were programmed in MATLAB and PYCHARM and implemented on a computer with a 2.30 GHz Intel Xeon Silver 4316 CPU processor and an NVIDIA GTX3090 graphics card.

[0078] Circular phantoms and multi-layer phantoms such as Figure 9 As shown in Figure 1, the multilayer phantom includes three layers of patterns: dots, diagonal stripes, and vertical stripes, from top to bottom. Each layer is 5 pixels thick, and the distance between layers is 6 pixels. The simulation parameters are shown in Table 1.

[0079] Table 1 Simulation experiment parameters

[0080]

[0081] The spiral range refers to the range of change of the CL angle during the rotation of the spherical spiral scan, which is also the height of the spiral, such as Figure 4 As shown in (g), the spiral range is determined by two CL angles, namely the ①CL angle at the starting point of the spiral and the ②CL angle at the ending point of the spiral. The CL angle at the ending point is the minimum CL angle of the spiral.

[0082] The circular phantom is projected and reconstructed with 360-degree spherical spirals of different spiral ranges. If the minimum CL angle of the spiral is 5° and the CL angle of the starting point is changed, the reconstructed image and cross-sectional diagram are as follows: Figure 10 As shown in (a), it can be seen that when the starting point CL angle is large, the internal distortion of the image becomes more serious. If the starting point CL angle is taken as 45° and the minimum CL angle of the spiral is changed, the reconstructed image and profile diagram are as follows Figure 10As shown in Figure (b), the smaller the minimum spiral CL angle, the less image distortion. This is because larger CL angles result in more data loss and more severe image aliasing. A smaller minimum spiral angle provides more supplemental data and significantly improves aliasing artifacts. Therefore, in spherical spiral CL imaging, the smallest possible minimum spiral angle should be selected.

[0083] The multi-layer model was imaged and reconstructed by performing 10°, 20°, 30°, and 40° CL imaging and corresponding spiral CL imaging. In spiral CL imaging, the minimum spiral angle was 5°. The transverse slice (50th layer) and longitudinal slice of the reconstructed image are shown in Figure 2. Figure 11 As shown, the transverse slice is located at the center slice of the oblique stripe layer.

[0084] As can be seen from the figure, when the CL angle increases, the artifacts of the circular trajectory CL and SHCL reconstructed images become more and more serious. From the horizontal slice diagram, the aliasing artifacts of the superposition of dots and vertical stripes become more and more obvious, such as the area in the red frame. They are caused by the upper dot layer and the lower vertical stripe layer respectively. From the vertical slice diagram, it can be observed that the original strip-shaped figure diffuses into a hollow circular figure. The longitudinal diffusion causes hollow artifacts in the layer itself and aliasing artifacts in the upper and lower layers. The section diagram marked with yellow lines is as follows Figure 12 As shown in FIG, the data manifestation of the hollow artifact is that the originally uniform gray value has a central depression, and the data manifestation of the aliasing artifact is that the area with the original gray value of 0 has gray.

[0085] Comparing the images reconstructed by circular trajectory and spherical spiral trajectory CL, we can see that for each angle, the aliasing artifacts in the SHCL transverse slices are significantly reduced, and from the longitudinal slices, the diffusion of the sample shape is significantly shrunk. Figure 12 It can also be seen from the cross-sectional diagram in that the distortion of SHCL data at each angle is smaller than that of CL data.

[0086] From the frequency domain information of the reconstructed images of the circular trajectories CL and SHCL at various angles, as shown in Figure 13 As shown in the figure, SHCL has richer frequency domain information, as shown in the red box. The RMSE, PSNR, SSIM, and CC metrics are calculated to evaluate the quality of the reconstructed image. As can be seen in Table 2, all metrics of SHCL are significantly better than those of circular trajectory CL.

[0087] Table 2. Comparison of imaging indicators

[0088]

[0089]

[0090] CL and SHCL imaging was performed on a homemade PCB using a cone-beam CL imaging system. The system's radiation source was a FineTec 160.02ETT with a beam angle of up to 170°, and the detector was a Teledynedalsa Rad-icon 3030 with dimensions of 307 mm x 306 mm. The imaging parameters are shown in Table 3.

[0091] Table 3 PCB imaging parameters

[0092] parameter value Voltage 125kV Current 80uA SOD 442mm SDD 1429mm Detector size 99um Number of probes 3104*3096 PCB sample size 106mm*106mm*1.8mm

[0093] The optical image and design diagram of the circuit elements of each layer of the self-made 4-layer PCB circuit board are as follows Figure 14 shown.

[0094] Perform 6°-30° SHCL imaging on the PCB, and the first layer of the reconstruction result is as follows: Figure 15 To further observe the imaging details and compare the imaging quality, the three local areas ①②③ were imaged and reconstructed using 6°CL, 30°CL and 6°-30°SHCL. Due to the different nature of artifacts in 6°CL and 30°CL images, the 6°-30°SHCL imaging results were compared with the 6°CL and 30°CL results, as shown in the figure. Figure 15 、 Figure 16 shown.

[0095] Figure 16 The 6°CL and 6°-30°SHCL reconstruction images were compared. In the 6°CL image, hardening artifacts can be clearly seen. This is caused by the small CL angle, the long penetration path of the multi-energy X-ray, and the absorption of low-energy rays by the object. It mainly manifests as "cup"-shaped artifacts, blurred edges, banding or radial artifacts, etc. Figure 16 As shown in the red box. In the 6°-30° SHCL image, as the CL angle gradually increases, the ray penetration path becomes shorter and the hardening artifact is not obvious, but the PCB interlayer aliasing artifact caused by the missing CL data can be clearly seen, such as Figure 16 As shown in the green box in the middle. Therefore, the main causes and manifestations of artifacts in 6°CL and 6°-30°SHCL imaging are different. A comprehensive comparison of the image average gradient, image entropy, and local contrast indicators shows from Table 4 that the image quality of 6°-30°SHCL is better than that of 6°CL images.

[0096] Table 4 Image quality comparison

[0097]

[0098]

[0099] Figure 16The images reconstructed by 30°CL and 6°-30°SHCL were compared. Figure 16 The reconstruction results for local area ① show that the grayscale of the 6°-30° SHCL image is smoother, and the hollow artifacts are significantly smaller than those in the 30° CL image. The reconstruction results for local areas ② and ③ show that the intensity of aliasing artifacts in the SHCL image is smaller than that in the CL image in both the component image and the blank area. The cross-sectional view of the area marked by the yellow line also shows that the grayscale distortion of the 6°-30° SHCL image is significantly smaller than that of the 30° CL image.

[0100] Although the present application has been described with reference to specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present application. Accordingly, this specification and the drawings are merely illustrative of the present application as defined by the appended claims and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, those skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, the present application is intended to include such modifications and variations if they fall within the scope of the claims of the present application and their equivalents.

Claims

1. A spherical spiral trajectory CBCL imaging method, characterized in that: In spherical spiral CBCL imaging, the imaging sample is fixed, the X-ray source and detector rotate around the rotation axis, and the CL angle is changed by adjusting the position of the X-ray source and detector; alternatively, the imaging sample rotates around the rotation axis, and the CL angle is adjusted by adjusting the position of the X-ray source and detector; alternatively, the X-ray source and detector are fixed, the sample rotates around the rotation axis, and the CL angle is changed by adjusting the inclination angle of the rotation axis.

2. The method according to claim 1, wherein The distances between the ray source, the detector and the rotation center remain unchanged.

3. An improved FDK reconstruction method suitable for spherical spiral trajectory CBCL imaging, characterized in that: The method comprises the following steps: S1, obtain projection data and perform weighted filtering on the projection data: Where p(θ,a,u,v) is the projection data when the projection angle is θ, θ ranges from 0° to 360°, D is the distance from the ray source to the detector plane, u and v are the coordinates of the detector, h(u′-u) is the filter function, and a is the tilt angle; S2, perform spatial transformation on the reconstructed volume data: [x′,y′,z′+=[x,y,z]R(a,Δa,θ); Among them, x, y, z are the coordinates of the spatial transformation precursor data, R(a, Δa, θ) is the rotation factor, and R(a, Δa, θ) is specifically: S3, back-project the reconstructed volume data: S4, traverse each rotation angle θ, repeat steps S1-S3, and accumulate the back-projected volume data:

4. The method according to claim 3, wherein The method is applied to the reconstruction of projection data acquired by the method according to claim 1 or 2 .

5. An improved FDK reconstruction system suitable for spherical spiral trajectory CBCL imaging, characterized in that: The system includes the following modules: The weighted filtering module is used to obtain projection data and perform weighted filtering on the projection data: Where p(θ,a,u,v) is the projection data when the projection angle is θ, θ ranges from 0° to 360°, D is the distance from the ray source to the detector plane, u and v are the coordinates of the detector, h(u′-u) is the filter function, and a is the tilt angle; The spatial transformation module is used to perform spatial transformation on the reconstructed volume data: [x′,y′,z′+=[x,y,z]R(a,Δa,θ); Among them, x, y, z are the coordinates of the spatial transformation precursor data, R(a, Δa, θ) is the rotation factor, and R(a, Δa, θ) is specifically: Back projection module, used to back project volume data: The accumulation module is used to traverse each rotation angle θ, repeat steps S1-S3, and accumulate the back-projected volume data:

6. The system according to claim 5, wherein: The system reconstructs the projection data collected by the method according to claim 1 or 2.

Images