Geometric calibration methods, devices, equipment, systems and media for multi-source and multi-probe systems

Abstract

This application relates to the field of medical imaging technology, and in particular to a method, apparatus, device, system, and medium for geometric calibration of a multi-source multi-detector system. The method includes: determining the coupling relationship between the X-ray source and detector of the multi-source multi-detector system based on the irradiation relationship between the X-ray source and detector; dividing the X-ray source and detector of the multi-source multi-detector system into coupled X-ray source subsets, coupled detector subsets, and uncoupled subsets based on the coupling relationship; designing a calibration phantom based on the design parameters of the multi-source multi-detector system and constructing a calibration dataset; and iteratively solving the geometric parameters of the multi-source multi-detector system sequentially based on the coupled X-ray source subset, coupled detector subset, uncoupled subset, and an objective function constructed using the calibration dataset, to complete the calibration of the overall geometric parameters of the multi-source multi-detector system. This solves the problem of high computational complexity in the geometric parameter calibration of multi-source multi-detector systems in related technologies.

Description

Technical Field

[0001] This application relates to the field of medical imaging technology, and in particular to a geometric calibration method, device, equipment, system and medium for a multi-source multi-probe system. Background Technology

[0002] Currently, the temporal resolution of CT (Computed Tomography) systems has reached its limit, but this resolution is still insufficient to scan objects that are constantly in high-speed motion, such as the human heart. Therefore, reducing or even avoiding the rotation of the source probe will help improve the temporal resolution of CT systems. With the rotation speed of the gantry approaching its limit, increasing the arrangement and number of source probes is one of the development trends of this technology. Therefore, research on multi-source, multi-probe systems has begun to gradually develop, hoping to obtain sufficiently complete projection images for reconstruction under smaller rotation angles, thereby achieving higher resolution.

[0003] However, after scanning an object, accurate geometric parameters are required for successful and clear reconstruction of its internal structure using any reconstruction algorithm. Inaccurate geometric parameters will result in severe geometric artifacts in the reconstructed image, causing significant blurring of the object's internal structure and greatly affecting the system's spatial resolution performance. These errors are unacceptable for reconstruction; therefore, various methods are needed to calibrate the geometric parameters of the constructed multi-source, multi-detector CT system.

[0004] Related technologies utilize steel ball phantoms based on projection matrix relationships for multi-source and multi-detector parameter calibration. However, in actual experiments, due to undesirable factors such as errors in steel ball processing positions, turntable rotation angles, and centroid identification errors in the projection diagrams, the equations established based on the projection matrix become ill-conditioned. This results in multiple geometric parameter solutions with similar reconstruction effects within a given parameter space, impacting the reconstruction performance of the multi-source and multi-detector system. While it's possible to establish a unified set of equations constraining the correspondences of all sources and detectors, and then simultaneously optimize the relevant parameters of all X-ray sources and detectors using this constraint, this process increases computational complexity, requiring greater computational resources. Furthermore, this calibration process cannot utilize the calibrated parameters to optimize the position parameters of the small balls within the calibration phantom or the rotation angle parameters during rotation. Summary of the Invention

[0005] This application provides a geometric calibration method, apparatus, equipment, system, and medium for multi-source and multi-probe systems to solve the problem of high computational complexity in the calibration of geometric parameters of multi-source and multi-probe systems in related technologies.

[0006] The first aspect of this application provides a geometric calibration method for a multi-source multi-detector system, comprising the following steps: determining the coupling relationship between the radiation source and the detector of the multi-source multi-detector system based on the irradiation relationship between the radiation source and the detector; dividing the radiation source and detector of the multi-source multi-detector system into a coupled radiation source subset, a coupled detector subset, and an uncoupled subset based on the coupling relationship between the radiation source and the detector; designing a calibration phantom based on the design parameters of the multi-source multi-detector system and constructing a calibration dataset; and iteratively solving the geometric parameters of the multi-source multi-detector system sequentially based on the coupled radiation source subset, the coupled detector subset, the uncoupled subset, and the objective function constructed using the calibration dataset, to complete the calibration of the overall geometric parameters of the multi-source multi-detector system.

[0007] Optionally, the geometric parameters of the multi-source multi-detector system are iteratively solved sequentially based on the coupled X-ray source subset, coupled detector subset, uncoupled subset, and an objective function constructed using a calibration dataset. This includes: performing parameter iteration on the coupled detector subset based on the objective function to obtain the first pre-estimated coupled detector parameters, coupled X-ray source parameters, and calibration phantom pose parameters within the coupled detector subset; performing parameter iteration on the first pre-estimated coupled detector parameters, coupled X-ray source parameters, calibration phantom pose parameters, and objective function to obtain the calibration phantom coordinates and rotation angle; performing parameter iteration on the coupled detector subset based on the calibration phantom coordinates, rotation angle, and objective function to obtain the second pre-estimated coupled detector parameters, coupled X-ray source parameters, and calibration phantom pose parameters within the coupled detector subset; and performing parameter iteration on the calibration phantom... The coupled detector subset is iterated again using coordinates, rotation angle, calibration phantom pose parameters, and objective function to obtain coupled detector parameters. The coupled ray source subset is iterated using the coupled detector parameters, calibration phantom coordinates, rotation angle, calibration phantom pose parameters, and objective function to obtain coupled ray source parameters. Alternating cyclic parameter iterations are performed using coupled ray source parameters and coupled detector parameters until the overall average projection error is less than or equal to the error threshold, at which point iteration stops, resulting in optimized coupled ray source parameters and optimized coupled detector parameters. Finally, the uncoupled subset is iterated using the optimized coupled ray source parameters, optimized coupled detector parameters, calibration phantom coordinates, rotation angle, phantom pose parameters, and objective function to obtain uncoupled ray source parameters and uncoupled detector parameters.

[0008] Optionally, after obtaining the coordinates and rotation angle of the calibration phantom by iterating the parameters based on the first pre-estimated parameters of the coupled detector, the coupled X-ray source, the calibration phantom pose parameters, and the objective function, the method further includes: obtaining the coordinates and rotation angle of the calibration phantom obtained by iterating the parameters of all subsets of the coupled detector; and averaging all the coordinates and rotation angles of the calibration phantom to obtain the final coordinates and rotation angle of the calibration phantom.

[0009] Optionally, after obtaining the second pre-estimated coupled detector parameters, coupled ray source parameters, and calibrated phantom pose parameters within the coupled detector subset by iterating the parameters of the coupled detector subset based on the calibration phantom coordinates, rotation angle, and objective function, the method further includes: obtaining the calibrated phantom pose parameters obtained by iterating all coupled detector subsets; and averaging all calibrated phantom pose parameters to obtain the final calibrated phantom pose parameters.

[0010] Optionally, the calibration mold is a mold made of metal or high-density material spheres, and the spheres in the calibration mold are arranged in an alternating spiral pattern on opposite sides.

[0011] A second aspect of this application provides a geometric calibration device for a multi-source multi-detector system, comprising: a determination module for determining the coupling relationship between the radiation source and the detector of the multi-source multi-detector system based on the irradiation relationship between the radiation source and the detector; a partitioning module for partitioning the radiation source and detector of the multi-source multi-detector system into a coupled radiation source subset, a coupled detector subset, and an uncoupled subset based on the coupling relationship between the radiation source and the detector; a design module for designing a calibration phantom based on the design parameters of the multi-source multi-detector system and constructing a calibration dataset; and a calibration module for iteratively solving the geometric parameters of the multi-source multi-detector system sequentially based on the coupled radiation source subset, the coupled detector subset, the uncoupled subset, and an objective function constructed using the calibration dataset, to complete the calibration of the overall geometric parameters of the multi-source multi-detector system.

[0012] Optionally, the calibration module is further configured to: perform parameter iteration on the coupled detector subset according to the objective function to obtain the first pre-estimated coupled detector parameters, coupled ray source parameters, and calibration phantom pose parameters within the coupled detector subset; perform parameter iteration on the first pre-estimated coupled detector parameters, coupled ray source parameters, calibration phantom pose parameters, and objective function to obtain the calibration phantom coordinates and rotation angle; perform parameter iteration on the coupled detector subset according to the calibration phantom coordinates, rotation angle, and objective function to obtain the second pre-estimated coupled detector parameters, coupled ray source parameters, and calibration phantom pose parameters within the coupled detector subset; and perform parameter iteration on the coupled detector subset according to the calibration phantom coordinates, rotation angle, calibration phantom pose parameters, and objective function. The parameters of the coupled detector subset are iterated again to obtain the coupled detector parameters. Based on the coupled detector parameters, the coordinates of the calibrated phantom, the rotation angle, the pose parameters of the calibrated phantom, and the objective function, the parameters of the coupled ray source subset are iterated to obtain the coupled ray source parameters. Alternating cyclic parameter iterations are performed based on the coupled ray source parameters and the coupled detector parameters until the overall average projection error is less than or equal to the error threshold, at which point iteration stops, resulting in optimized coupled ray source parameters and optimized coupled detector parameters. Based on the optimized coupled ray source parameters, optimized coupled detector parameters, the coordinates of the calibrated phantom, the rotation angle, the pose parameters of the phantom, and the objective function, the parameters of the uncoupled subset are iterated to obtain uncoupled ray source parameters and uncoupled detector parameters.

[0013] Optionally, it further includes: a first averaging module, used to obtain the coordinates and rotation angles of the calibration phantom obtained by parameter iteration based on the first pre-estimated coupled detector parameters, coupled X-ray source parameters, calibration phantom pose parameters and objective function; and to average all calibration phantom coordinates and all rotation angles to obtain the final calibration phantom coordinates and final rotation angle.

[0014] Optionally, it also includes: a second averaging module, used to obtain the calibration phantom pose parameters obtained by iterating the parameters of the coupled detector subset according to the calibration phantom coordinates, rotation angle and objective function to obtain the second pre-estimated coupled detector parameters, coupled ray source parameters and calibration phantom pose parameters within the coupled detector subset; and to average and update all calibration phantom pose parameters to obtain the final calibration phantom pose parameters.

[0015] Optionally, the calibration mold is a mold made of metal or high-density material spheres, and the spheres in the calibration mold are arranged in an alternating spiral pattern on opposite sides.

[0016] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the program to implement the geometric calibration method for a multi-source multi-probe system as described in the above embodiments.

[0017] The fourth aspect of this application provides a geometric calibration system for a multi-source multi-probe system, comprising: a calibration phantom, wherein the calibration phantom is a phantom made of metal or high-density material spheres, and the spheres in the calibration phantom are arranged in an alternating spiral pattern on opposite sides; a rotating stage for placing and rotating the calibration phantom; and the electronic equipment described in the above embodiment.

[0018] A fifth aspect of this application provides a computer-readable storage medium having a computer program or instructions stored thereon, which are executed by a processor to implement the geometric calibration method for a multi-source, multi-probe system as described in the above embodiments.

[0019] A sixth aspect of this application provides a computer program, including computer programs or instructions, which, when executed, are used to implement the geometric calibration method for a multi-source, multi-probe system as described in the above embodiments.

[0020] Therefore, this application has the following beneficial effects:

[0021] This application's embodiments can divide the X-ray source and detector into coupled X-ray source subsets, coupled detector subsets, and uncoupled subsets based on the coupling relationship between the X-ray source and detector. A calibration phantom and a calibration dataset are designed and constructed. The geometric parameters of the multi-source multi-detector system are iteratively solved sequentially based on the coupled X-ray source subset, coupled detector subset, uncoupled subset, and the objective function constructed using the calibration dataset. This completes the overall geometric parameter calibration of the multi-source multi-detector system. The complex problem of calibrating all geometric parameters as a whole is subdivided into the problem of calibrating the coupled X-ray source subset, coupled detector subset, and uncoupled subset sequentially. This reduces the complexity of geometric calibration for multi-source multi-detector systems, effectively reduces computational resources in calibration, and improves calibration efficiency. Therefore, it solves the technical problem of high computational complexity in the geometric parameter calibration of multi-source multi-detector systems in related technologies.

[0022] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0023] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0024] Figure 1This is a flowchart of a geometric calibration method for a multi-source, multi-probe system provided according to an embodiment of this application;

[0025] Figure 2 This is a schematic diagram of a multi-source, multi-detector CT system provided according to an embodiment of this application;

[0026] Figure 3 This is a schematic diagram illustrating the definition of Euler angles for a detector according to an embodiment of this application;

[0027] Figure 4 This application provides an overall flowchart for overall geometric calibration according to embodiments of the present application;

[0028] Figure 5 This is an example diagram of a geometric calibration device for a multi-source, multi-probe system provided according to an embodiment of this application;

[0029] Figure 6 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application;

[0030] Figure 7 This is a schematic diagram of the geometric calibration system of the multi-source multi-probe system provided according to an embodiment of this application. Detailed Implementation

[0031] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0032] Before describing the solution of this application, let's first introduce some relevant issues about CT systems.

[0033] CT imaging technology, based on the principle that different materials and objects attenuate X-rays differently, is widely used in medical, security inspection, and industrial non-destructive testing fields. A typical CT system consists of an X-ray source, a detector, a rotating gantry, and related electronic components. Utilizing the characteristic that X-rays have different linear attenuation coefficients after passing through different objects, the detector collects the signal after the X-ray light emitted from the X-ray source penetrates the object; this signal is usually called a projection image. Then, analytical reconstruction algorithms such as FBP and FDK, or iterative reconstruction algorithms such as ART, can be used to reconstruct the internal structure of the object, achieving the goal of non-destructive observation of the object's internal structure.

[0034] Since the advent of CT, research and exploration of CT systems have been continuously developing. Currently, widely used cone-beam CT systems, classified according to the source probe's motion trajectory, include helical track, circular track, and C-arm CBCT systems. These CT systems all involve a process of acquiring more angular projection images by rotating the source probe. After years of development, due to the challenges posed by the centrifugal force of the source probe rotating around the object, the rotation speed of the gantry has reached its limit. Therefore, the temporal resolution of the CT system has also reached its limit, but this temporal resolution is still insufficient to support scanning objects that are constantly in high-speed motion, such as the human heart. Therefore, reducing or even avoiding the source probe rotation process will help improve the temporal resolution of CT systems, enabling the scanning and reconstruction of objects like the heart that are constantly moving. With the gantry rotation speed approaching its limit, increasing the arrangement and number of source probes is one of the development trends in this technology. Therefore, researchers have begun to gradually develop multi-source, multi-probe systems, hoping to obtain sufficiently complete projection images for reconstruction under smaller rotation angles, thereby achieving higher resolution. In recent years, the development of carbon nanotube cold cathode X-ray sources has brought possibilities for the realization and construction of static CT systems. It is hoped that a ring of detectors and X-ray sources can be arranged around the object to be imaged, and the X-ray sources can be sequentially emitted by electrical signals, without rotation, to acquire sufficiently complete projection image signals for reconstruction. Such a system is called static CT. Theoretically, the temporal resolution of static CT systems will be significantly improved compared to gantry-based CT systems, making it possible to image high-speed moving objects such as the heart. Therefore, static CT is also considered the future "fourth generation CT".

[0035] However, regardless of whether it's rotational or static CT, accurate geometric parameters are essential for successful and clear reconstruction of internal structures using any reconstruction algorithm after scanning an object. Inaccurate geometric parameters will result in severe geometric artifacts in the reconstructed image, causing significant blurring of the object's internal structure and drastically impacting the system's spatial resolution performance. In the actual construction of a CT system, due to the mechanical assembly of multiple components, the assembled CT system rarely reflects the accurate geometric relationships established in the initial system design. Errors will exist near the designed geometric parameter values, which are unacceptable for reconstruction. Therefore, various methods are needed to calibrate the geometric parameters of the assembled CT system. A CT system's key components include the X-ray source, detector, and stage; the relationship between these three components significantly affects the CT image reconstruction results.

[0036] Numerous effective calibration methods exist for the geometric parameter calibration of single-source, single-detector rotational CT systems. These methods can be broadly categorized into two types: those based on specific phantoms and those that do not require a specific phantom. Among phantom-based methods, those using steel ball phantoms are the most widely studied and employed. These methods utilize the projected positions of the steel ball phantom in the projection image, establishing geometric relationships based on the positional information between the specially designed balls to directly calculate geometric parameters. Other methods utilize the elliptical trajectory formed after the steel ball phantom rotates and scans one revolution on a circular track to directly calculate geometric parameters. Still others utilize the projection relationship between the three-dimensional coordinates of the steel balls and the centers of the balls in the projection image to establish a projection matrix, which is then calculated analytically or iteratively optimized to obtain the geometric parameters. There are also calibration methods based on line reference phantoms, using the relationship between a three-dimensional line reference and a two-dimensional straight line in the projection image to analytically or iteratively optimize and solve for geometric parameters. Finally, there are methods based on head models with known structures, using forward projection registration to calibrate geometric parameters. The limitation of these methods is that they require separate phantom designs for specific geometries for different systems, resulting in insufficient robustness. Another type of calibration method that does not require a specific phantom includes methods that utilize the symmetry of the projection onto the detector after a full scan of a circular track to calibrate geometric parameters; others use information such as the entropy of the reconstructed image, the roundness of the sphere, and the sharpness of the edges to calculate geometric parameters. These methods do not require a specific phantom, which helps in online calibration of CT systems. However, some methods are still limited to circular track CT systems, and these methods are significantly affected by poor imaging conditions such as noise, resulting in high computational costs. However, there is currently limited research on methods for multi-source, multi-detector systems. Some methods use steel ball phantoms based on projection matrix relationships for multi-source, multi-detector parameter calibration. However, in actual experiments, due to errors in the steel ball processing position, the rotation angle of the turntable, and the centroid identification error in the projection image, the equations established based on the projection matrix become a somewhat ill-conditioned problem, no longer a geometric parameter optimization process with a unique analytical solution. This poses a challenge for parameter calibration when there are X-ray sources simultaneously hitting multiple detectors (coupled X-ray sources) and detectors simultaneously receiving signals from multiple coupled X-ray sources (coupled detectors), making it no longer a problem that can be solved independently for a single source and single detector.

[0037] As described above, one method with significant reference value and good adaptability for multi-source, multi-detector systems is the method based on steel ball phantoms and projection matrices. By simply designing the arrangement of the steel balls, it can be ensured that each pair of source-detector relationships has enough balls projected onto the detector, providing sufficient correspondence between the three-dimensional coordinates of the steel balls and the coordinates of the steel balls projected onto the two-dimensional detector. This establishes an optimization problem based on the projection matrix, thereby optimizing and calculating the geometric parameters of the system. In the practical application of this calibration method, due to undesirable factors such as errors in the processing position of the steel balls, errors in the rotation angle of the turntable, and errors in centroid identification in the projection image, the equations established based on the projection matrix become a problem with a certain degree of ill-conditioning. This results in multiple geometric parameter solutions with similar reconstruction effects within a certain parameter space. In the geometric calibration of single-source, single-detector CT systems, this ill-conditioning has a limited impact on the final reconstruction effect because only the optimized source-detector geometric parameters are used during the projection process. However, for multi-source, multi-detector systems, if independent calibration of single-source, single-detector systems continues, the resulting problems will lead to further issues. When projected signals from various sources are used together to reconstruct an object, a mismatch arises between the independent single-source, single-detector systems. For example, if a single X-ray source simultaneously hits two detectors (called a coupled X-ray source), after single-source, single-detector calibration of this source and each detector, it's found that the position parameters of the X-ray source cannot be aligned. Similarly, if a detector can sequentially receive signals from two coupled X-ray sources on either side (called a coupled detector), after single-source, single-detector calibration of this detector and each coupled X-ray source, it's found that the position and angle parameters of the detector cannot be aligned. One direct solution to this problem is to establish a single system of equations constraining the correspondences between all sources and detectors, and then use this constraint to simultaneously optimize the relevant parameters of all X-ray sources and detectors. However, this process increases computational complexity, requiring more computational resources. Furthermore, this calibration process cannot utilize the calibrated parameters to optimize the position parameters of the spheres within the calibration phantom or the rotation angle parameters during rotation.

[0038] To address the aforementioned issues, this application provides a geometric calibration method for a multi-source, multi-detector system. Based on the coupling relationship between the X-ray source and the detector, the X-ray source and detector are divided into coupled X-ray source subsets, coupled detector subsets, and uncoupled subsets. A calibration phantom and a calibration dataset are designed. The geometric parameters of the multi-source, multi-detector system are iteratively solved sequentially based on the coupled X-ray source subset, coupled detector subset, uncoupled subset, and the objective function constructed using the calibration dataset. Parameter updates are performed independently and sequentially within each subset, and calculations between different subsets are independent. This effectively reduces the number and complexity of parameters in each parameter iteration, and reduces the consumption of computational resources and computation time.

[0039] Specifically, Figure 1 This is a schematic flowchart illustrating the geometric calibration method for a multi-source, multi-probe system provided in an embodiment of this application.

[0040] like Figure 1 As shown, the geometric calibration method for this multi-source, multi-probe system includes the following steps:

[0041] In step S101, the coupling relationship between the X-ray source and the detector of the multi-source multi-detector system is determined based on the irradiation relationship between the X-ray source and the detector of the multi-source multi-detector system.

[0042] It is understood that, according to the embodiments of this application, the coupling relationship between the X-ray source and the detector of the multi-source multi-detector system can be determined.

[0043] In step S102, based on the coupling relationship between the X-ray source and the detector of the multi-source multi-detector system, the X-ray source and detector of the multi-source multi-detector system are divided into a coupled X-ray source subset, a coupled detector subset, and an uncoupled subset.

[0044] It should be noted that in a multi-source, multi-detector system, some X-ray sources can only be received by one detector; these sources are called uncoupled X-ray sources. Uncoupled X-ray sources that can be received by the same detector can be called a group of uncoupled X-ray sources. Conversely, some X-ray sources can be received by two detectors simultaneously; these sources are called coupled X-ray sources. Coupled X-ray sources that can be received by two identical detectors can be called a group of coupled X-ray sources. Because the X-ray beam emitted by the sources has a certain linewidth during actual system setup, and a collimation system is generally designed to limit the linewidth range of the X-ray beam, it is virtually impossible for one X-ray source to simultaneously hit three X-ray sources. Among the detectors, some can receive signals from at least one group of coupled X-ray sources; these detectors are called coupled detectors. Other detectors can only receive signals from at least one group of uncoupled X-ray sources; these detectors are called uncoupled detectors.

[0045] Furthermore, this application primarily addresses the calibration problem of multi-source, multi-detection systems with coupled source probes, mainly targeting systems such as... Figure 2The system shown (a multi-source multi-detector CT system) has X-ray sources arranged closely on one side and detectors arranged closely on the opposite side. The close arrangement of X-ray sources facilitates the acquisition of signals from as many angles as possible during a single rotation, while the close arrangement of detectors on the opposite side further facilitates the acquisition of signals from all X-ray sources within a single rotation angle. Therefore, in such a system, each detector can be considered a coupled detector, and a set of coupled X-ray sources exists between two adjacent detectors. These coupled X-ray source groups and coupled detectors form a complete alternating coupled source-detector link or even a loop. If there is no coupled X-ray source between two intermediate detectors, this multi-source multi-detector system can be divided into multiple subsystems, allowing the coupled sources within each subsystem to form a complete alternating coupled source-detector link. Independent parameter calibration can then be performed within each subsystem.

[0046] Therefore, in this embodiment, the X-ray sources and detectors of a multi-source multi-detector system can be classified into multiple coupled X-ray source subsets, multiple coupled detector subsets, and uncoupled subsets based on the coupling relationship between the X-ray sources and detectors in the multi-source multi-detector system. This allows for the orderly and iterative updating of the source-detector subsets (the collective term for the X-ray source subsets and detector subsets) in subsequent steps. The complex problem requiring the solution of all geometric parameters is subdivided into multiple sub-problems for sequential solution, reducing the complexity of solving the geometric parameter problem of multi-source multi-detector systems. This can effectively reduce the consumption of computing resources and improve the efficiency of calibration work.

[0047] Among them, the X-ray source subsets S-Subsets: each subset contains a set of coupled X-ray sources and coupled detectors that can receive the signals from that set of X-ray sources, denoted as S. n The number is n, and the total number is N;

[0048] Detector Subsets (D-Subsets): Each subset contains a coupled detector and a coupled ray source that can illuminate that detector, denoted as D. m The number is m, and the total number is M.

[0049] In other words, after dividing the source probe subsets, each calibration using the projection signal will be an optimization calculation of the source probe parameters within the subset. The subsets are independent of each other. At the same time, it will be found that adjacent coupled X-ray source subsets will contain the same coupled detector, and adjacent coupled detector subsets will contain the same set of coupled detectors.

[0050] In step S103, a calibration phantom is designed based on the design parameters of the multi-source multi-probe system, and a calibration dataset is constructed.

[0051] In this embodiment, the calibration mold can be made of metal or high-density material spheres. The spheres in the calibration mold are arranged in an alternating spiral arrangement on opposite sides. The calibration mold can be designed according to specific circumstances, or it can be other molds. No specific limitation is made in this regard.

[0052] It is understood that the embodiments of this application can design a calibration model for the design parameters of a multi-source multi-probe system and construct a calibration dataset so that the geometric parameters of the multi-source multi-probe system can be solved subsequently.

[0053] It should be noted that the arrangement of the small balls in the calibration phantom of this application can effectively avoid overlapping during projection signal acquisition, thus affecting geometric calibration.

[0054] Specifically, this application designs a calibration phantom in which small spheres of metal or high-density material are arranged in an alternating spiral pattern on opposite sides of an plexiglass column. The alternating spiral arrangement of the spheres on opposite sides can effectively avoid overlap during projection signal acquisition. The angular arrangement of the spheres is uniform, but their distribution radius, distribution height, angular interval, and number of spheres can be simply given according to the system design parameters. After selecting appropriate parameters, the projection of the phantom on the detector will be arranged as comprehensively as possible, ensuring that almost every area has a complete projection of the spheres. Furthermore, the distribution of the sphere phantom in the parameter coordinate system will cover a larger area, which helps to reduce the ill-conditioned problems in solving geometric parameters.

[0055] It should be noted that reference coordinates need to be established during subsequent parameter calculations. Figure 2 Taking a multi-source, multi-detector CT system as an example, the following four coordinate systems are defined during parameter calculation: system coordinate system, phantom coordinate system, center coordinate system, and projection coordinate system, as follows:

[0056] System coordinate system: Select a detector, take the geometric center of its center pixel as the origin, take the row direction of the detector as the X-axis, take the column direction of the detector as the Y-axis, and take the center perpendicular line of the detector plane as the Z-axis. The positive direction of the Z-axis is consistent with the direction in which the detector receives the X-ray. This coordinate system does not require any device other than the source detector. That is, after the source detector is installed, the geometric parameters of the source detector position in this coordinate system will not change. Therefore, it is defined as the system coordinate system, denoted as the XYZ coordinate system.

[0057] Mold coordinate system: The coordinate system that defines the mold. The positions of the small balls inside the mold are designed and machined under this coordinate system. Once the mold is machined, the mold coordinate system is determined, denoted as the X′Y′Z′ coordinate system. The small balls inside the mold are arranged in an alternating spiral pattern on opposite sides under this coordinate system.

[0058] Central coordinate system: For multi-source, multi-probe systems requiring a rotation axis, to facilitate the description of rotational motion around the axis and related parameters, the rotation axis and calibration phantom are selected as references. The rotation axis is taken as the Y-axis, and the positive direction of the Y-axis is chosen to be basically in the same direction as the positive direction of the Y′ axis of the phantom coordinate system. The height of the origin of the Y-axis is consistent with the height of the origin of the phantom coordinate system. The X-axis of this coordinate system is parallel to the X′ axis of the phantom coordinate system, and it is denoted as the central coordinate system (X″Y″Z″). For multi-source, multi-probe systems without a rotation axis, the phantom coordinate system is used as the central coordinate system.

[0059] Projected coordinate system: This coordinate system conveniently represents the projection relationship between a point in space and the two-dimensional coordinates on the detector plane. It is used to construct the projection matrix. Each source-detector pair can form a projected coordinate system. This coordinate system takes the position of the ray source as the origin, the perpendicular line from the ray source to the detector plane as the Z-axis (the positive direction of the Z-axis is from the ray source to the detector), the axis parallel to the center row of the system detector as the X-axis, and the axis parallel to the center column of the system detector as the Y-axis, thus establishing the projected coordinate system, denoted as the xyz coordinate system. This coordinate system will not be used for parameter description, but only for the process of constructing the projection matrix.

[0060] Under each defined coordinate system, the multi-source multi-probe system has the following source and probe geometric parameters:

[0061] (1) Locations of all X-ray sources: (S x,i ,S y,i ,S z,i ) T (i = 1, 2, 3, ...);

[0062] (2) Location of all detector centers: the geometric center of the detector's center pixel (D x,j D y,j D z,j ) T (j = 1, 2, 3, ...);

[0063] (3) Euler angles of all detectors: yaw angle θ j Twist angle α j Pitch angle Euler angles are defined as follows: Figure 3 As shown.

[0064] The above source parameters are called source direct geometric parameters, denoted as . However, in general reconstruction algorithms, source probe distance or axis probe distance and detector offset are typically used to represent geometric parameters. Therefore, for the final reconstruction problem, under various defined coordinate systems, multi-source multi-probe systems have the following source probe geometric parameters, which are called source probe reconstruction geometric parameters:

[0065] (1) Locations of all X-ray sources: (S x,i ,S y,i ,S z,i ) T (i = 1, 2, 3, ...);

[0066] (2) Axis distance: ODD j (j=1,2,3,…), the perpendicular distance between the origin on the coordinate axis and all detector planes, where the axis distance of the reference detector in the system coordinate system is 0;

[0067] (3) Detector offset: (U 0,j V 0,j ) T The pixel coordinates of the detectors where the foot of the perpendicular from the origin on the coordinate axis to all detector planes is located, where the offset of the reference detector in the system coordinate system is the pixel coordinate of the center pixel.

[0068] (4) Euler angles of the detector itself: the yaw angle θ of each detector j Twist angle α j Pitch angle

[0069] In the reconstruction of geometric parameters, all parameters are uniformly denoted as... The relevant parameters of the radiation source are denoted as follows: The relevant parameters of the detector are denoted as

[0070] It should be noted that research shows that a deviation of about 1° will have almost no impact on the reconstruction effect regarding the elevation angle of the detector, and this level of accuracy is easy to achieve for installation. Therefore, the elevation angle will be assumed to be 0 in the following geometric parameter calculations.

[0071] To facilitate the coordinate transformation process during calculation, direct geometric parameters will be used. After all parameters have been calculated, the direct geometric parameters will be converted into reconstructed geometric parameters. The conversion relationship is shown in equation (1):

[0072]

[0073] Among them, T para It is a parametric transformation matrix. The rotation matrices around the Y-axis, X-axis, and Z-axis are respectively, as shown in equation (2):

[0074]

[0075] in, p represents the coordinates of the pixel at the geometric center of the detector. u p vThese represent the pixel sizes of the detector in the horizontal (u) and vertical (v) directions, respectively, while the Euler angles of the detector remain consistent.

[0076] During the calibration of the multi-source, multi-detector system, the phantom will be positioned as close as possible to the center of the system's field of view. If a rotation axis is introduced to add data, its position will also be positioned as close as possible to the center of the system's field of view. During the system fine-tuning phase, the central coordinate system can be used as a reference coordinate system for geometric calculations. This will visually demonstrate the relationship between the source detector's installation position and the center of the field of view or the platform, directly providing a reference for fine-tuning. Therefore, subsequent iterative calculations will all be performed in the central coordinate system. After the geometric parameters are calculated, to avoid incomparable differences in parameter values ​​due to different phantom placements, the system coordinate system can be transformed to a system coordinate system established with the detector as a reference. The coordinate system transformation is a simple rotation and translation process. For example, if the detector selected as the reference in the system coordinate system is numbered 0, after the geometric parameters are calculated, the center position of this detector in the central coordinate system is: (D x″,0 D y″,0 D z″,0 ) T The Euler angles of the detector are: Therefore, the transformation of coordinate parameters from the central coordinate system to the system coordinate system is expressed as equation (3):

[0077]

[0078] Where R is the rotation matrix and t0 is the translation matrix, as shown in equation (4):

[0079]

[0080] The transformation of Euler angles is expressed as equation (5):

[0081]

[0082] In step S104, the geometric parameters of the multi-source multi-detector system are solved iteratively according to the coupled X-ray source subset, coupled detector subset, uncoupled subset, and objective function constructed using the calibration dataset, so as to complete the calibration of the overall geometric parameters of the multi-source multi-detector system.

[0083] It is understood that, in the embodiments of this application, the geometric parameters of the multi-source multi-detector system can be solved iteratively according to the coupled X-ray source subset, the coupled detector itself, the uncoupled subset, and the objective function constructed using the calibration dataset. The calibration of the overall geometric parameters of the multi-source multi-detector system is divided into multiple subsets for solving, which reduces the computational complexity, effectively reduces the computational resources in the calibration, and improves the calibration efficiency.

[0084] Specifically, in this embodiment of the application, after scanning the calibration phantom and obtaining the calibration dataset, a target function is constructed. When iteratively updating the parameters within the subset, the target function is established based on the projection relationship between the three-dimensional coordinates of the sphere and the pixel coordinates of its projection center, as follows:

[0085] During the design of the mold body machining, the position (X′) of each small ball in the mold body coordinate system is as follows: k ,Y k ′,Z′ k ) T It is definite, where k is the ball index. Of course, there will be some error in the processing. After correction and estimation, a more accurate ball position can be obtained. The calibration phantom is placed within the system's field of view. X-rays emitted sequentially from all X-ray sources penetrate the ball phantom and generate a series of projected signals of the balls on the detector. If a rotation axis is needed to increase calibration or projection data, the calibration phantom can be rotated to collect more projection signals. The Canny algorithm can be used to divide the ball projections using gradient thresholding in the projection image. The centroid of each ball projection is calculated to obtain the pixel coordinates (u) of the projected position of the ball's center on the detector. k ,v k After matching the three-dimensional coordinates of the center of the ball with the pixel coordinates of the projected center of the ball on the detector, multiple sets of correspondences can be obtained, forming a calibration dataset. Each source-detector relationship that can receive signals will have its own calibration dataset. Since the data can be increased by introducing a rotation axis, this dataset is sufficient.

[0086] During the process of X-ray irradiation of an object to form a projection, the relationship between the three-dimensional coordinates on the object and the pixel coordinates of the projection on the detector is similar to that of transmission projection. The relationship can be linked by the projection matrix (P), as shown in equation (6), introducing a homogeneous coordinate system:

[0087]

[0088] Among them, w k It is a distance weighting factor, (x″) k ,y″ k ,z″ k ) T The three-dimensional coordinates of the ball in the central coordinate system can also be denoted as X". k ′,(u k ,v k T represents the pixel coordinates of the projected position of the center of the ball on the detector, and P is a 3×4 projection matrix containing the complete geometric parameters of the entire perspective relationship system. Once the projection matrix of a system is determined, the geometric parameters of the system are also determined. The projection matrix can be decomposed into equation (7):

[0089] P = K[R|t]; (7)

[0090] Wherein, K is an upper triangular matrix containing the geometric information of the system’s built-in construction, as shown in equation (8);

[0091]

[0092] In the K matrix, SDD is the perpendicular distance from the ray source to the detector plane, p u p v These are the pixel sizes of the detector in the horizontal (u direction) and vertical (v direction), respectively, and are generally constants. (u0, v0) are the detector coordinates where the foot of the perpendicular from the ray source to the detector plane is located. Generally, detectors are conventional rectangles, so there is no pixel distortion factor.

[0093] R is the rotation matrix and t is the translation matrix, which contains the external geometric information of the system. After rotation and translation, the coordinates of the ball are transformed from the central coordinate system to the projected coordinate system. R is an orthogonal matrix containing the three Euler angles of the detector, as shown in equation (9):

[0094]

[0095] t is a translation matrix containing the position information of the ray source, as shown in equation (10):

[0096]

[0097] Some parameters of the K matrix are inconsistent with the geometric parameters defined in 3.3, but they can be transformed by the translation matrix t, as shown in equation (11):

[0098]

[0099] The reconstructed geometric parameters in equation (11) can be derived from the direct geometric parameters through equation (1). Thus, the parameters in the projection matrix are derived from the reconstructed geometric parameters. The projection matrix can be constructed based on the corresponding geometric parameters of any source probe.

[0100] For a multi-source, multi-probe system without a rotation axis, the central coordinate system is the same as the phantom coordinate system. Therefore, the position of the sphere in the central coordinate system (x″) k ,y″ k ,x″ k ) T That is, the position of the small ball (x′) in the phantom coordinate system. k ,y′ k ,z′ k ) T .

[0101]

[0102] Therefore, for a multi-source multi-probe system that does not require a rotation axis, its projection matrix relationship (6) can be constructed by combining equations (8), (9), (10), and (12), as shown in equation (13):

[0103]

[0104] Among them, w k As a distance weighting factor, (u k ,v k ) T Let (x″) be the pixel coordinates of the projected position of the center of the ball on the detector. k ,y″ k ,z″ k ) T The three-dimensional coordinates of the ball in the central coordinate system can also be denoted as X″. k P is a 3×4 projection matrix containing complete geometric parameters of the entire perspective relationship system.

[0105] For multi-source and multi-probe systems that require the introduction of a rotation axis, in actual experiments, the vertical central axis of the placed model is somewhat deviated from the rotation axis, that is, the central coordinate system and the model coordinate system do not coincide. Therefore, it is necessary to set rotation and translation parameters to represent the position and orientation of the model in the central coordinate system, thereby transforming the three-dimensional coordinates of the ball from the model coordinate system to the central coordinate system, as shown in equation (14):

[0106]

[0107] Where Rt is the rigid rotation and translation matrix of the model, which will be constructed according to equation (15):

[0108] Rt=T p R x R z (15)

[0109]

[0110]

[0111] According to the definitions of the phantom coordinate system and the central coordinate system, the transformation from the phantom coordinate system to the central coordinate system is determined by θ. x θ z Two rotation angles and x p0 z p0 Two translations determine these parameters, which are denoted as... These parameters also characterize the pose information of the phantom in the central coordinate system.

[0112] In practical experiments, it may be impossible to obtain a sufficient calibration dataset due to single-angle phantom scanning. Therefore, rotating the phantom can increase the projection of the sphere, thereby obtaining a sufficient correspondence between the 3D coordinates of the sphere and the pixel coordinates of its projection center, thus expanding the calibration dataset. That is, after the phantom pose is determined, it is then rotated θ around the rotation axis. y″ (All rotation angle parameters are denoted as) The three-dimensional coordinates (x″) of the ball in the central coordinate system. k ,y" k ,z" k ) T Based on equation (14), a rotation transformation will be performed, as shown in equation (16):

[0113]

[0114] Therefore, for a multi-source, multi-probe system that requires the introduction of a rotation axis, its projection matrix relationship (6) can be constructed by combining equations (8), (9), (10), (15), and (16), as shown in equation (17):

[0115]

[0116] In actual experiments, the three-dimensional coordinates of the ball are usually referenced using the ball's manufacturing design values, the rotation angle is referenced using the values ​​set in the experiment, and the center of the ball is estimated by calculating the centroid of the ball's projection area divided using the Canny algorithm. However, during the projection acquisition process, there will be non-ideal factors such as noise, the size of the X-ray source focal spot, centroid algorithm error, ball manufacturing error, and rotation angle error, making the solution of (13) and (17) a minimization optimization problem.

[0117] The objective function of this application is to minimize the reprojection error. The L2 norm of the error between the projected position and the true projected position after the three-dimensional coordinates of the ball are calculated by the projection relationship (Equation (13) or Equation (17)) is as shown in Equation (18):

[0118]

[0119] in For the set of reconstructed geometric parameters to be determined for the system, The angle parameter for the rotation of the mold body around the rotation axis. This is the set of parameters for the phantom's pose. To calibrate the set of three-dimensional coordinates of the spheres in the correspondence within the dataset, the coordinates of the spheres in the phantom coordinate system are generated, where... This represents the set of positions of the small ball in the phantom coordinate system. This is the set of projection center pixel coordinates in the correspondence of the calibration dataset.

[0120] In the partitioned subset, The subset should include the parameters of all X-ray sources and detectors within it, and use the calibration dataset between all sources and detectors included in this subset to construct the objective function of the subset. The objective function is solved using the LM algorithm, with the system design value as the initial value, and the geometric parameters are solved iteratively.

[0121] In this embodiment, the geometric parameters of the multi-source multi-detector system are iteratively solved sequentially based on a subset of coupled X-ray sources, a subset of coupled detectors, a subset of uncoupled components, and an objective function constructed using a calibration dataset. This includes: iterating the parameters of the coupled detector subset according to the objective function to obtain the first pre-estimated coupled detector parameters, coupled X-ray source parameters, and calibration phantom pose parameters within the coupled detector subset; iterating the parameters of the calibration phantom according to the first pre-estimated coupled detector parameters, coupled X-ray source parameters, calibration phantom pose parameters, and the objective function to obtain the calibration phantom coordinates and rotation angle; iterating the parameters of the coupled detector subset again according to the calibration phantom coordinates, rotation angle, and the objective function to obtain the second pre-estimated coupled detector parameters, coupled X-ray source parameters, and calibration phantom pose parameters; and iterating the parameters of the calibration phantom subset according to the calibration phantom coordinates, rotation angle, and the objective function to obtain the second pre-estimated coupled detector parameters, coupled X-ray source parameters, and calibration phantom pose parameters; and iterating the parameters of the calibration phantom subset according to the calibration phantom coordinates. The parameters of the coupled detector subset are obtained by iterating the parameters again using the rotation angle, calibration phantom pose parameters, and objective function. The parameters of the coupled ray source subset are then obtained by iterating the parameters of the coupled detector subset using the coupled detector parameters, calibration phantom coordinates, rotation angle, calibration phantom pose parameters, and objective function. Alternating cyclic parameter iterations are performed on the coupled ray source parameters and coupled detector parameters until the overall average projection error is less than or equal to the error threshold, at which point the iteration stops, resulting in optimized coupled ray source parameters and optimized coupled detector parameters. Finally, the parameters of the uncoupled subset are obtained by iterating the parameters of the uncoupled ray source and uncoupled detector using the optimized coupled ray source parameters, optimized coupled detector parameters, calibration phantom coordinates, rotation angle, calibration phantom pose parameters, and objective function.

[0122] It is understood that, in the embodiments of this application, the coupled detector subset can be initially iterated based on the objective function to obtain the first pre-estimated coupled detector parameters, coupled X-ray source parameters, and calibration phantom pose parameters within the coupled detector subset; the calibration phantom coordinates and rotation angles can be obtained through parameter iteration based on the first pre-estimated coupled detector parameters, coupled X-ray source parameters, calibration phantom pose parameters, and the objective function; the coupled detector subset can be initially iterated based on the calibration phantom coordinates, rotation angles, and the objective function to obtain the second pre-estimated coupled detector parameters, coupled X-ray source parameters, and calibration phantom pose parameters within the coupled detector subset; the coupled detector subset can be iterated based on the calibration phantom coordinates, rotation angles, and the objective function. The parameters of the coupled detector are obtained by iterating through the set of parameters again. The coupled ray source subset is then iterated through based on the coupled detector parameters, the coordinates of the calibrated phantom, the rotation angle, the pose parameters of the calibrated phantom, and the objective function to obtain the coupled ray source parameters. Alternating cyclic parameter iteration is performed based on the coupled ray source parameters and the coupled detector parameters until the overall average projection error is less than or equal to the error threshold, at which point iteration stops, resulting in optimized coupled ray source parameters and optimized coupled detector parameters. Finally, the uncoupled subset is iterated through based on the optimized coupled ray source parameters, optimized coupled detector parameters, the coordinates of the calibrated phantom, the rotation angle, the pose parameters of the calibrated phantom, and the objective function to obtain uncoupled ray source parameters and uncoupled detector parameters.

[0123] In this embodiment of the application, after obtaining the coordinates and rotation angle of the calibration phantom by iterating the parameters based on the first pre-estimated parameters of the coupled detector, the coupled X-ray source, the calibration phantom pose parameters, and the objective function, the method further includes: obtaining the coordinates and rotation angle of the calibration phantom obtained by iterating the parameters of all subsets of the coupled detector; and averaging all the coordinates and rotation angles of the calibration phantom to obtain the final coordinates and rotation angle of the calibration phantom.

[0124] Since the calibration phantom has processing errors and rotary table errors, the embodiments of this application use the parameters of different subsets to estimate the processing errors of the calibration phantom and the rotation deviation of the rotary table, obtained by parameter iteration within each subset. Finally, the average is taken to obtain the final calibration phantom coordinates and the final rotation angle. This effectively estimates the processing errors of the small ball phantom and the rotation deviation of the rotary table, improves the calibration accuracy of the calibration method, and brings better reconstruction results. Then, based on the final calibration phantom coordinates, the final rotation angle, and the objective function, the parameters of the coupled detector subset are iterated again to obtain the coupled detector parameters.

[0125] In this embodiment, after obtaining the second pre-estimated coupled detector parameters, coupled ray source parameters, and calibrated phantom pose parameters within the coupled detector subset by iterating the parameters based on the calibration phantom coordinates, rotation angle, and objective function, the method further includes: obtaining the calibrated phantom pose parameters obtained by iterating through all coupled detector subsets; and averaging all calibrated phantom pose parameters to obtain the final calibrated phantom pose parameters.

[0126] Since the phantom pose parameters should remain consistent after calculation of the coupled detector subsets, and in actual calculations, it is found that the phantom pose parameters are basically consistent across different subsets, but there are still some differences, this embodiment of the application can average the calibration phantom pose parameters obtained by iterating through all coupled detector subsets to obtain the final calibration phantom pose parameters. Then, based on the final calibration phantom coordinates, the final rotation angle, the final calibration phantom pose parameters, and the objective function, the coupled detector subsets are iterated again to obtain updated coupled detector parameters. Finally, based on the updated coupled detector parameters, the final calibration phantom coordinates, the final rotation angle, and the objective function, the coupled X-ray source subsets are iterated to obtain coupled X-ray source parameters, thus ensuring the accuracy of parameter iteration.

[0127] In this embodiment, after obtaining the uncoupled X-ray source parameters and uncoupled detector parameters by iterating the parameters of the uncoupled subset based on the optimized coupled X-ray source parameters, optimized coupled detector parameters, calibrated phantom coordinates, rotation angle, phantom pose parameters, and objective function, the method further includes: performing parameter conversion on the multi-source multi-detector system based on the optimized coupled detector parameters, optimized coupled X-ray source parameters, uncoupled X-ray source parameters, and uncoupled detector parameters, converting the computational geometric parameters into reconstructed geometric parameters.

[0128] Since the parameters calculated in the above iterative calculations are computational geometric parameters, the embodiments of this application can perform parameter conversion on the multi-source multi-detector system using the final optimized coupled detector parameters, optimized coupled X-ray source parameters, uncoupled X-ray source parameters, and uncoupled detector parameters, converting the computational geometric parameters into reconstructed geometric parameters, thus completing the geometric calibration of the multi-source multi-detector system.

[0129] Specifically, after designing and fabricating the phantom, dividing the source probe subset, defining the system coordinate system and the central coordinate system, and scanning and constructing the calibration dataset, this embodiment of the application can use this geometric calibration method for a multi-source multi-probe system with coupled source probes based on ordered source probe subsets to complete the estimation of all geometric parameters.

[0130] For multi-source, multi-probe systems that incorporate a rotating axis, the specific calibration procedure is as follows: Figure 4 As shown, the specific steps are as follows:

[0131] (1) The first pre-estimation iteration of geometric parameters is performed within the subset of the detector subset D-Subsets, using the designed small spherical coordinate processing values. Angle rotation setting value Based on the source probe combinations contained in the subset, and using the calibration dataset composed of all corresponding relationships under the corresponding source probe combinations, the objective function is constructed as shown in equation (19):

[0132]

[0133] in, Let represent the parameters of all coupled X-ray sources in the m-th detector subset. Let be the parameters of the coupled detectors in the m-th detector subset. Let be the phantom pose parameters in the m-th detector subset. Set the angle rotation value. To use the designed spherical coordinates for machining values, Let X be the set of three-dimensional coordinates of the sphere in the calibration dataset of the m-th detector subset, generated from the designed three-dimensional coordinate value X0′ of the sphere. Let Rt be the set of pixel coordinates of the projection center of the sphere in the phantom within the calibration dataset of the m-th detector subset, and Rt be the rigid rotation and translation matrix of the phantom. After this calculation, the coupled detector and coupled ray source within each subset have geometric parameters, the phantom pose parameters are also calculated, and the reprojection errors are all small.

[0134] (2) Using the iteration parameters estimated in the first step of D-Subsets in (1), the coordinates of the small ball are calculated. Rotation angle The parameters are iterated, and the objective function is as shown in equation (20):

[0135]

[0136] Within each subset, the ball's coordinates and rotation angle will be estimated once. After all subsets have been calculated, the average will be taken, as shown in equation (21):

[0137]

[0138] The actual position coordinates of the ball have now been estimated. The actual rotation angle of the system This will effectively improve accuracy, and subsequent iterations will use the estimated ball coordinates and rotation angles from this calculation.

[0139] (3) Using the estimated ball coordinates Rotation angle Re-evaluate the coupled detector parameters within the D-Subsets subset. With coupling X-ray source parameters phantom pose parameters The second preliminary estimation iteration calculation, the objective function is as shown in equation (22):

[0140]

[0141] (4) After the second pre-estimation calculation of D-Subsets, the phantom pose is calculated. The results calculated from different subsets should be consistent. In actual calculations, it is found that the results from different subsets are basically consistent, but there are still some differences. In order to ensure the consistency of the phantom pose between different subsets, an average update is performed here. At this point, it is determined that the phantom pose will no longer change. The objective function is as shown in equation (23):

[0142]

[0143] (5) Using spherical coordinates Rotation angle estimation Other poses of the phantom Re-evaluate the coupled detector parameters within the D-Subsets subset. With coupling X-ray source parameters The iterative calculation is performed, and the objective function is shown in equation (24):

[0144]

[0145] This calculation is for the detector parameters in D-Subsets of round 0. calculate.

[0146] (6) Using spherical coordinates Rotation angle estimation Other poses of the phantom Simultaneously, the detector parameters in ⑤ will be... Used as known parameters, the coupled ray source parameters within the S-Subsets subset are... The iterative calculation is performed, and the objective function is shown in equation (25):

[0147]

[0148] in, Let X' be the set of three-dimensional coordinates of the sphere in the calibration dataset of the nth detector subset, defined by the design value X′ of the sphere's three-dimensional coordinates. est generate, This is the set of pixel coordinates of the projection centers of the spheres in the calibration dataset of the nth detector subset. This calculation represents the detector parameters in D-Subsets for round 0. Based on the calculations, calculate the radiation source parameters for each subset of radiation sources. This initial round will yield a set of overall parameters for a multi-source, multi-probe system, and will provide the magnitude of the overall average reprojection error under these parameters.

[0149] (7) Couple detector parameters in the detector subset for the i-th round. The calculation uses spherical coordinates. Rotation angle estimation Other poses of the phantom X-ray source parameters calculated in the previous S-Subsets step The objective function is calculated using known quantities, as shown in equation (26):

[0150]

[0151] This calculation is for the detector parameters in the i-th round of D-Subsets. The calculation.

[0152] (8) Couple the ray source parameters in the ray source subset S-Subsets of the i-th round. The calculation uses spherical coordinates. Rotation angle estimation phantom pose The detector parameters calculated in this round's D-Subsets step The objective function is calculated using known quantities, as shown in equation (27):

[0153]

[0154] This calculation is for the ray source parameters in the i-th round of S-Subsets. The calculation will yield a set of overall parameters for a multi-source, multi-probe system, and provide the magnitude of the overall average reprojection error under these parameters.

[0155] (9) Compare whether the overall average reprojection error of the i-th round continues to decrease compared to the overall average reprojection error of the (i-1)-th round. If the decrease is greater than 1e-5, it is considered that the overall parameters have not converged and the loop continues (7)(8). If the decrease is less than or equal to 1e-5, it is considered that the overall parameters have converged and the loop stops (7)(8).

[0156] (10) Using spherical coordinates Rotation angle estimation phantom pose Coupled detector parameters calculated in the D-Subsets step of the convergence round Perform X-ray source parameters of uncoupled source The iterative calculation is performed, and the objective function is shown in equation (28):

[0157]

[0158] in, Let be the set of three-dimensional coordinates of the sphere in the calibration dataset of the l-th uncoupled ray source, derived from the design values ​​of the sphere's three-dimensional coordinates. generate, This is the set of pixel coordinates corresponding to the projection centers of the small spheres in the calibration dataset of the l-th uncoupled X-ray sources. At this point, the geometric parameter calibration calculations for all X-ray sources and detectors in the multi-source multi-detector system have been completed.

[0159] (11) In the calculations of the previous steps, the reconstructed geometric parameters of all X-ray sources and detectors in the multi-source multi-detector system are in the central coordinate system. Next, according to equations (3) and (5), all source and detector parameters in the central coordinate system are transformed to source and detector parameters in the system coordinate system. Thus, the direct geometric parameters of the source and detector in the system coordinate system of the multi-source multi-detector system are obtained. Furthermore, according to equation (1), the direct geometric parameters of the source and detector can be converted into reconstructed geometric parameters of the source and detector, which can be directly used in the reconstruction process. Thus, the calibration work is completed.

[0160] For multi-source, multi-probe systems that do not require the introduction of a rotating shaft, compared to processes with a rotating shaft, Includes only the initial angle of 0°. All parameters are 0. Removing the angle estimation in step (2) and steps (4) and (5) completes the calibration process for a multi-source, multi-probe system without a rotation axis. Figure 4 Simply remove the dotted lines from the process flow.

[0161] In summary, the geometric calibration method for multi-source multi-detector systems in this application is applicable to the geometric calibration of multi-source multi-detector CT systems with coupled source probes. The specific steps include:

[0162] 1. Design of calibration phantom (taking small ball calibration phantom as an example);

[0163] 2. Divide the source probes into subsets based on their coupling relationships;

[0164] 3. Establish a reference coordinate system and design geometric parameters;

[0165] 4. Scan the calibration phantom, obtain the calibration dataset, and construct the objective function;

[0166] 5. Overall geometric calibration process.

[0167] The geometric calibration method of the multi-source multi-probe system of this application is illustrated below through a specific embodiment. Figure 2 Take the multi-source, multi-detector CT system as an example.

[0168] In this system, due to the small source detection coverage and field of view, a rotation axis was incorporated into the design, located near the center of the system's field of view. The system consists of carbon nanotube cold cathode X-ray sources, narrow strip detectors, a rotating stage, and other supporting electronic facilities. The carbon nanotube cold cathode X-ray array sources and narrow strip detectors are arranged closely on opposite sides around the rotation axis of the rotating stage, ensuring that all X-rays emitted from the X-ray sources are received by the detectors. There are seven carbon nanotube cold cathode X-ray array sources, numbered 1-7. Each array source contains seven focal points arranged in a uniform straight line, also numbered 1-7. Therefore, number 1.1 represents focal point 1 on array source 1. There are also three narrow strip detectors, numbered 1-3.

[0169] 1. Design and fabricate the mold.

[0170] Based on the field of view of the design system, a suitable distribution radius and height of the small balls were selected to cover the system's field of view as much as possible. The small ball phantom was designed and fabricated. Since the system's radiation source and detector are misaligned in height, i.e., the radiation is incident at an angle, the overlap of the small ball phantom's projection will be further reduced.

[0171] 2. Divide the data into subsets based on the source-probe coupling relationship.

[0172] Based on the above subset partitioning rules, for example... Figure 2 The system first classifies the radiation sources:

[0173] 1. All the rays emitted from the focal points of array sources 1 and 2 only illuminate detector 3. These focal points are numbered as the first group of uncoupled ray sources.

[0174] The rays emitted from the focal points 1, 2, 3, and 4 on array sources 2 and 3 will simultaneously illuminate detectors 2 and 3. These focal points are numbered as the first group of coupled ray sources.

[0175] 3. The rays emitted from the 5th, 6th, and 7th focal points on array source 3, all focal points on array source 4, and the 1st, 2nd, and 3rd focal points on array source 5 all illuminate detector 2. These focal points are numbered as the second group of uncoupled ray sources.

[0176] The rays emitted from focal points 4, 5, 6, and 7 on array sources 4 and 6 will simultaneously illuminate detectors 1 and 2. These focal points are designated as the second group of coupled ray sources.

[0177] The rays emitted from all the focal points on array sources 5, 6, and 7 only illuminate detector 1. These focal points are numbered as the third group of uncoupled ray sources.

[0178] Detectors 1, 2, and 3 in the system are all coupled detectors, among which:

[0179] 1. Detector No. 1 can receive radiation signals from the second set of coupled radiation sources and the third set of uncoupled radiation sources;

[0180] 2. Detector No. 2 can receive radiation signals from the first set of coupled radiation sources, the second set of coupled radiation sources, and the second set of uncoupled radiation sources;

[0181] 3. Detector No. 3 can receive radiation signals from the first set of coupled radiation sources and the first set of uncoupled radiation sources.

[0182] Therefore, the three sets of coupled X-ray sources and three coupled detectors are divided into a D-subset of coupled detectors and a S-subset of coupled X-ray sources:

[0183] 1. Coupled detector subsets D-Subsets:

[0184] (1) The coupled detector subset D1 includes: detector No. 3 and the first group of coupled X-ray sources;

[0185] (2) The coupled detector subset D2 includes: detector No. 2, the first group of coupled X-ray sources, and the second group of coupled X-ray sources;

[0186] (3) The coupled detector subset D3 includes: detector No. 1 and the second set of coupled X-ray sources.

[0187] 2. Coupled ray source subsets S-Subsets:

[0188] (1) The coupled X-ray source subset S1 includes: detectors No. 2 and No. 3, and the first set of coupled X-ray sources;

[0189] (2) The coupling X-ray source subset S2 includes: detectors 1 and 2, and the second set of coupling X-ray sources.

[0190] III. Establish a reference coordinate system and design geometric parameters.

[0191] Based on the above definition of coordinate systems, the system coordinate system, the phantom coordinate system, and the center coordinate system are established respectively. In the system coordinate system, the reference detector is selected as detector No. 2, and the direct geometric parameters and reconstruction geometric parameters are designed.

[0192] 4. Scan the calibration phantom to obtain the calibration dataset.

[0193] The calibration phantom is placed on the stage, with its central axis as close as possible to the rotation axis, and then rotated systematically. The rotational scanning is performed in a step-by-step manner; after each rotation, all X-ray sources emit beams sequentially, and the corresponding detectors acquire the projection signals in turn. After one complete rotation, the calibration data acquisition is finished.

[0194] Using the collected calibration data, post-processing is performed to establish a calibration dataset. In the projection image, there is a projection of a small spherical phantom. After gradient threshold segmentation using the Canny algorithm, the projection region of the sphere can be divided, and the centroid of this projection is calculated. The centroid is considered to be the two-dimensional pixel coordinates of the three-dimensional coordinates of the center of a small sphere projected onto the detector plane. After a one-to-one matching between the spheres on the phantom and their projections on the detector, the correspondence between the three-dimensional coordinates of the spheres and the two-dimensional pixel coordinates of their projection centers is obtained. Each set of source detectors capable of receiving signals will have numerous correspondences, which together constitute the calibration dataset for that set of source detectors. The calibration dataset for each subset consists of the calibration datasets of all source detectors contained within that subset.

[0195] V. Overall geometric calibration process.

[0196] 1. The first preliminary iteration of the geometric parameters is performed within the subset of the detector subset D-Subsets. The objective function of the iteration is shown in Equation (19). After this step of calculation, the coupled detector and coupled X-ray source in each subset have geometric parameters, the phantom pose parameters are also calculated, and the reprojection error is small.

[0197] 2. Using the initial iteration parameters estimated in step 1, iterate the ball coordinates and rotation angle parameters. The objective function is calculated as shown in equation (20). The calculation results of each subset are averaged as shown in equation (21). Thus, the ball coordinates and rotation angle are estimated, which will effectively improve the accuracy of the ball coordinates and rotation angle used in the iterative calculation. Subsequent iterations will use the estimated ball coordinates and rotation angle for calculation.

[0198] 3. Using the estimated ball coordinates and rotation angle, perform the second iterative calculation of the pre-estimation of the coupled detector parameters, coupled ray source parameters, and phantom pose parameters within the D-Subsets subset. The objective function is shown in equation (22).

[0199] 4. After calculation in step 3, in order to ensure the consistency of the phantom pose among different subsets, the pose calculation results of different subsets in step 3 are used to perform an average update as shown in equation (23), and the pose of the phantom is thus determined.

[0200] 5. Using the estimated sphere coordinates, rotation angle, and phantom pose, iteratively calculate the coupled detector parameters and coupled X-ray source parameters within the D-Subsets subset again. The objective function for the calculation is shown in equation (24). This calculation is for the detector parameters in the 0th round of the D-Subsets.

[0201] 6. Using the estimated sphere coordinates, rotation angle, and phantom pose, and taking the five detector parameters as known parameters, iterative calculations of the X-ray source parameters within the S-Subsets subset are performed. The objective function is shown in equation (25). After this initial round, a set of overall parameters for the multi-source multi-detector system will be obtained, and the magnitude of the overall average reprojection error under these parameters will be given.

[0202] 7. Calculate the coupled detector parameters in the detector subset of the i-th round. Use the estimated sphere coordinates, rotation angle estimate, phantom pose, and X-ray source parameters calculated in the previous round's S-Subsets step as known quantities for calculation. The objective function is shown in equation (26). This calculation is for the detector parameters in the D-Subsets of the i-th round.

[0203] 8. Calculate the coupled X-ray source parameters in the X-ray source subset of the i-th round. Use the estimated sphere coordinates, rotation angle, phantom pose, and detector parameters calculated in the D-Subsets step of the i-th round as known quantities for calculation. The objective function is shown in equation (27). This calculation is for the X-ray source parameters in the S-Subsets of the i-th round. After this round of calculation, the overall parameters of the multi-source multi-detector system in the i-th round can be obtained, and the magnitude of the overall average reprojection error under these parameters can be given.

[0204] 9. Compare whether the overall average reprojection error of round i is decreasing compared to the overall average reprojection error of round i-1. If the decrease is greater than 1e-5, it is considered that the overall parameters have not converged, and the cycle of 7 and 8 continues. If the decrease is less than or equal to 1e-5, it is considered that the overall parameters have converged, and the cycle of 7 and 8 is no longer repeated.

[0205] 10. Using the estimated sphere coordinates, rotation angle, phantom pose, and the coupled detector parameters calculated in the D-Subsets convergence steps, iteratively solve for the parameters of the uncoupled source. The objective function is shown in equation (28). At this point, the geometric parameter calibration calculations for all X-ray sources and detectors in the multi-source multi-detector system have been completed.

[0206] 11. In the calculations of the previous steps, the parameters of all X-ray sources and detectors in the multi-source multi-detector system are in the central coordinate system. Next, according to equations (3) and (5), all source and detector parameters in the central coordinate system are transformed to the system coordinate system. Thus, the direct geometric parameters of the source and detector in the system coordinate system of the multi-source multi-detector system are obtained. Furthermore, according to equation (1), the direct geometric parameters of the source and detector can be converted into reconstructed geometric parameters of the source and detector, which can be directly used in the reconstruction process. Thus, the calibration work is completed.

[0207] In summary, 1. This application designs a calibration phantom based on metal or high-density material spheres. The spheres are arranged in an alternating double helix pattern on opposite sides of the surface of an plexiglass column or plexiglass tube. By rotating the appropriate sphere distribution radius and height according to the system design parameters, the possible overlap between the sphere projection images is reduced. At the same time, the distribution of the sphere projections on the detector is made as dispersed as possible, reducing the ill-conditioned nature of the problem in solving the geometric parameters.

[0208] 2. This application designs a geometric calibration method for multi-source, multi-detector systems with coupled sources and detectors based on ordered source-detector subsets, effectively solving the complex geometric parameter calibration problem of multi-source, multi-detector systems. Simultaneously, this method innovatively introduces the concept of ordered subsets, classifying and dividing the X-ray sources and detectors into subsets. The complex problem requiring the simultaneous solution of all geometric parameters is subdivided into multiple sub-problems for sequential solution, reducing the complexity of solving the geometric parameter problem for multi-source, multi-detector systems. This effectively reduces the consumption of computational resources and improves the efficiency of calibration work.

[0209] 3. This application provides an effective method for estimating the machining error of a small ball phantom and the rotation deviation of a rotary table. By dividing the multi-source multi-probe system into multiple subset subsystems, the machining error of the small ball phantom and the rotation deviation of the rotary table are estimated using parameters from different subsets during the parameter estimation process. The average is then taken, effectively estimating the machining error of the small ball phantom and the rotation deviation of the rotary table, improving the calibration accuracy of the calibration method, and bringing better reconstruction results. That is, if the calibration data under a single viewpoint is insufficient due to the small field of view in the multi-source multi-probe system, or if the reconstruction effect is poor due to insufficient projection angle, a rotary table that drives the object to rotate can be introduced at the same time. The calibration data and projection data can be increased by rotating the phantom. Similarly, the rotation angle error can be estimated during the iteration of system parameters.

[0210] 4. The geometric calibration method proposed in this application can be effectively adapted and used regardless of the arrangement of the X-ray source and detector, providing a solution for the geometric calibration of future static CT systems.

[0211] The geometric calibration method for a multi-source multi-detector system proposed in this application can divide the X-ray source and detector into coupled X-ray source subsets, coupled detector subsets, and uncoupled subsets based on the coupling relationship between the X-ray source and detector. A calibration phantom and a calibration dataset are designed and constructed. The geometric parameters of the multi-source multi-detector system are solved iteratively according to the coupled X-ray source subset, coupled detector subset, uncoupled subset, and the objective function constructed using the calibration dataset. This completes the overall geometric parameter calibration of the multi-source multi-detector system. The complex problem of calibrating all geometric parameters as a whole is subdivided into the problem of calibrating the coupled X-ray source subset, coupled detector subset, and uncoupled subset sequentially. This reduces the complexity of geometric calibration of the multi-source multi-detector system, effectively reduces the computational resources in calibration, and improves calibration efficiency.

[0212] Next, the geometric calibration device for a multi-source multi-probe system according to an embodiment of this application is described with reference to the accompanying drawings.

[0213] Figure 5 This is a block diagram of the geometric calibration device of the multi-source multi-probe system according to an embodiment of this application.

[0214] like Figure 5 As shown, the geometric calibration device 10 of the multi-source multi-probe system includes: a determination module 100, a division module 200, a design module 300, and a calibration module 400.

[0215] The system comprises the following modules: Determination module 100 determines the coupling relationship between the X-ray source and detector of the multi-source multi-detector system based on the irradiation relationship between the X-ray source and detector; Division module 200 divides the X-ray source and detector of the multi-source multi-detector system into coupled X-ray source subset, coupled detector subset, and uncoupled subset based on the coupling relationship between the X-ray source and detector; Design module 300 designs a calibration phantom based on the design parameters of the multi-source multi-detector system and constructs a calibration dataset; and Calibration module 400 iteratively solves the geometric parameters of the multi-source multi-detector system based on the coupled X-ray source subset, coupled detector subset, uncoupled subset, and the objective function constructed using the calibration dataset, thereby completing the calibration of the overall geometric parameters of the multi-source multi-detector system.

[0216] In this embodiment, the calibration module 400 is further configured to: perform parameter iteration on the coupled detector subset according to the objective function to obtain the first pre-estimated coupled detector parameters, coupled ray source parameters, and calibration phantom pose parameters within the coupled detector subset; perform parameter iteration on the first pre-estimated coupled detector parameters, coupled ray source parameters, calibration phantom pose parameters, and objective function to obtain the calibration phantom coordinates and rotation angle; perform parameter iteration on the coupled detector subset according to the calibration phantom coordinates, rotation angle, and objective function to obtain the second pre-estimated coupled detector parameters, coupled ray source parameters, and calibration phantom pose parameters within the coupled detector subset; and perform parameter iteration on the calibration phantom coordinates, rotation angle, calibration phantom pose parameters, and objective function... The parameters of the coupled detector subset are iterated again to obtain the coupled detector parameters. The parameters of the coupled ray source subset are iterated based on the coupled detector parameters, the coordinates of the calibrated phantom, the rotation angle, the pose parameters of the calibrated phantom, and the objective function to obtain the coupled ray source parameters. Alternating cyclic parameter iteration is performed based on the coupled ray source parameters and the coupled detector parameters until the overall average projection error is less than or equal to the error threshold, at which point the iteration stops, resulting in optimized coupled ray source parameters and optimized coupled detector parameters. The parameters of the uncoupled subset are iterated based on the optimized coupled ray source parameters, optimized coupled detector parameters, the coordinates of the calibrated phantom, the rotation angle, the pose parameters of the phantom, and the objective function to obtain uncoupled ray source parameters and uncoupled detector parameters.

[0217] In this embodiment of the application, the apparatus 10 further includes a first averaging module.

[0218] The first averaging module is used to obtain the coordinates and rotation angles of the calibration phantom obtained by parameter iteration based on the first pre-estimated parameters of the coupled detector, coupled X-ray source, calibration phantom pose, and objective function. Then, it averages all the calibration phantom coordinates and rotation angles to obtain the final calibration phantom coordinates and final rotation angle.

[0219] In this embodiment of the application, the apparatus 10 further includes a second averaging module.

[0220] The second averaging module is used to obtain the calibration phantom pose parameters obtained by iterating the parameters of the coupled detector subset according to the calibration phantom coordinates, rotation angle and objective function to obtain the second pre-estimated coupled detector parameters, coupled ray source parameters and calibration phantom pose parameters within the coupled detector subset; and to obtain the final calibration phantom pose parameters by averaging all calibration phantom pose parameters.

[0221] In this embodiment, the design module 300 is further configured to: scan the calibration phantom in the multi-source multi-probe system and construct a calibration dataset.

[0222] In this embodiment of the application, the calibration mold is a mold made of small balls of metal or high-density material, and the small balls in the mold are arranged in an alternating spiral pattern on opposite sides.

[0223] It should be noted that the explanation of the geometric calibration method embodiment for the multi-source multi-probe system described above also applies to the geometric calibration device of the multi-source multi-probe system in this embodiment, and will not be repeated here.

[0224] The geometric calibration device for a multi-source multi-detector system proposed in this application can divide the X-ray source and detector into coupled X-ray source subsets, coupled detector subsets, and uncoupled subsets according to the coupling relationship between the X-ray source and the detector. A calibration phantom and a calibration dataset are designed and constructed. The geometric parameters of the multi-source multi-detector system are solved iteratively according to the coupled X-ray source subset, coupled detector subset, uncoupled subset, and the objective function constructed using the calibration dataset. This completes the overall geometric parameter calibration of the multi-source multi-detector system. The complex problem of calibrating all geometric parameters as a whole is subdivided into the problem of calibrating the coupled X-ray source subset, coupled detector subset, and uncoupled subset sequentially. This reduces the complexity of geometric calibration of the multi-source multi-detector system, effectively reduces the computational resources in calibration, and improves calibration efficiency.

[0225] Figure 6 This is a schematic diagram of the structure of an electronic device 20 provided in an embodiment of this application. The electronic device 20 may include:

[0226] The memory 601, the processor 602, and the computer program stored on the memory 601 and capable of running on the processor 602.

[0227] When the processor 602 executes the program, it implements the geometric calibration method for the multi-source multi-probe system provided in the above embodiments.

[0228] Furthermore, the electronic device 20 also includes:

[0229] Communication interface 603 is used for communication between memory 601 and processor 602.

[0230] The memory 601 is used to store computer programs that can run on the processor 602.

[0231] The memory 601 may include high-speed RAM (Random Access Memory) memory, and may also include non-volatile memory, such as at least one disk storage.

[0232] If the memory 601, processor 602, and communication interface 603 are implemented independently, then the communication interface 603, memory 601, and processor 602 can be interconnected via a bus to complete communication between them. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0233] Optionally, in a specific implementation, if the memory 601, processor 602, and communication interface 603 are integrated on a single chip, then the memory 601, processor 602, and communication interface 603 can communicate with each other through an internal interface.

[0234] The processor 602 may be a CPU (Central Processing Unit), an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of this application.

[0235] This application also provides a geometric calibration system for a multi-source, multi-probe system.

[0236] like Figure 7 As shown, the geometric calibration system 30 of the multi-source multi-probe system includes: a calibration phantom 31, a rotating stage 32, and electronic equipment 20.

[0237] The calibration model 31 is a model made of small spheres of metal or high-density material, and the small spheres in the model are arranged in alternating spirals on opposite sides; the rotating stage 32 is used to place the calibration model 31 and rotate the calibration model 31; the electronic device 20 is used to implement the geometric calibration method of the multi-source multi-probe system described above.

[0238] Since the geometric calibration system and the geometric calibration method of the multi-source multi-probe system have the same effect, they will not be described in detail here.

[0239] This application also provides a computer-readable storage medium storing a computer program or instructions thereon, which, when executed by a processor, implements the geometric calibration method of the multi-source multi-probe system described above.

[0240] This application also provides a computer program product, including a computer program or instructions, which, when executed, are used to implement the geometric calibration method for the multi-source multi-probe system described above.

[0241] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0242] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0243] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0244] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (FPGAs), field-programmable gate arrays (FPGAs), etc.

[0245] Those skilled in the art will understand that all or part of the steps of the methods implementing the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0246] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A geometric calibration method for a multi-source, multi-probe system, characterized in that, Includes the following steps: Based on the irradiation relationship between the X-ray source and the detector in the multi-source multi-detector system, the coupling relationship between the X-ray source and the detector in the multi-source multi-detector system is determined. Based on the coupling relationship between the X-ray source and the detector in the multi-source multi-detector system, the X-ray source and detector in the multi-source multi-detector system are divided into a coupled X-ray source subset, a coupled detector subset, and an uncoupled subset. A coupled X-ray source is a X-ray source received by multiple detectors, an uncoupled X-ray source is a X-ray source received by one detector, a coupled detector is a detector that receives at least one set of coupled X-ray sources, and an uncoupled detector is a detector that receives at least one set of uncoupled X-ray sources. Each subset of the coupled X-ray source subset contains a set of coupled X-ray sources and a coupled detector capable of receiving the coupled X-ray sources. Each subset of the coupled detector subset contains one coupled detector and a coupled X-ray source capable of irradiating the coupled detector. A calibration phantom is designed based on the design parameters of the multi-source multi-probe system, and a calibration dataset is constructed. The geometric parameters of the multi-source multi-detector system are iteratively solved sequentially based on the coupled X-ray source subset, the coupled detector subset, the uncoupled subset, and the objective function constructed using the calibration dataset, so as to complete the calibration of the overall geometric parameters of the multi-source multi-detector system.

2. The geometric calibration method for a multi-source, multi-probe system according to claim 1, characterized in that, The geometric parameters of the multi-source, multi-detector system are solved iteratively based on the coupled X-ray source subset, the coupled detector subset, the uncoupled subset, and the objective function constructed using the calibration dataset, including: Based on the objective function, the parameters of the coupled detector subset are iterated to obtain the first pre-estimated coupled detector parameters, coupled ray source parameters, and calibration phantom pose parameters within the coupled detector subset; Based on the first pre-estimated parameters of the coupled detector, the parameters of the coupled X-ray source, the pose parameters of the calibration phantom, and the objective function, the coordinates and rotation angle of the calibration phantom are obtained through parameter iteration. Based on the calibration phantom coordinates, the rotation angle, and the objective function, the parameters of the coupled detector subset are iterated to obtain the second pre-estimated coupled detector parameters, coupled ray source parameters, and calibration phantom pose parameters within the coupled detector subset; The parameters of the coupled detector are obtained by iterating the parameters of the subset of the coupled detectors again based on the coordinates of the calibrated phantom, the rotation angle, the pose parameters of the calibrated phantom, and the objective function; The coupled ray source parameters are obtained by iterating the parameters of the coupled detector, the coordinates of the calibration phantom, the rotation angle, the pose parameters of the calibration phantom, and the objective function on the coupled ray source subset. The parameters of the coupled X-ray source and the coupled detector are iterated alternately until the overall average projection error is less than or equal to the error threshold, and the iteration stops to obtain the optimized coupled X-ray source parameters and the optimized coupled detector parameters. Based on the optimized coupled X-ray source parameters, optimized coupled detector parameters, calibrated phantom coordinates, rotation angle, phantom pose parameters, and objective function, the uncoupled subset is iterated to obtain the uncoupled X-ray source parameters and uncoupled detector parameters.

3. The geometric calibration method for a multi-source, multi-probe system according to claim 2, characterized in that, After obtaining the coordinates and rotation angle of the calibration phantom by iterating through the parameters based on the first pre-estimated parameters of the coupled detector, the parameters of the coupled X-ray source, the pose parameters of the calibration phantom, and the objective function, the process further includes: Obtain the coordinates and rotation angles of the calibration phantom obtained through iterative analysis of all coupled detector subset parameters; The final calibration phantom coordinates and final rotation angles are obtained by averaging all calibration phantom coordinates and all rotation angles.

4. The geometric calibration method for a multi-source, multi-probe system according to claim 2, characterized in that, After obtaining the second pre-estimated coupled detector parameters, coupled ray source parameters, and calibrated phantom pose parameters within the coupled detector subset by iterating the parameters of the coupled detector subset based on the calibration phantom coordinates, the rotation angle, and the objective function, the method further includes: Obtain the pose parameters of the calibration phantom obtained through iteration of all coupled detector subsets; The final calibration phantom pose parameters are obtained by averaging and updating the pose parameters of all calibration phantoms.

5. The geometric calibration method for a multi-source, multi-probe system according to claim 1, characterized in that, The calibration mold is a mold made of metal or high-density material spheres, and the spheres in the mold are arranged in an alternating spiral pattern on opposite sides.

6. A geometric calibration device for a multi-source, multi-probe system, characterized in that, include: The determination module is used to determine the coupling relationship between the X-ray source and the detector of the multi-source multi-detector system based on the irradiation relationship between the X-ray source and the detector. The partitioning module is used to partition the X-ray sources and detectors of the multi-source multi-detector system into a coupled X-ray source subset, a coupled detector subset, and an uncoupled subset based on the coupling relationship between the X-ray sources and detectors of the multi-source multi-detector system. A coupled X-ray source is a X-ray source received by multiple detectors, an uncoupled X-ray source is a X-ray source received by one detector, a coupled detector is a detector that receives at least one set of coupled X-ray sources, and an uncoupled detector is a detector that receives at least one set of uncoupled X-ray sources. Each subset of the coupled X-ray source subset contains a set of coupled X-ray sources and a coupled detector capable of receiving the coupled X-ray sources. Each subset of the coupled detector subset contains one coupled detector and a coupled X-ray source capable of irradiating the coupled detector. The design module is used to design a calibration phantom based on the design parameters of the multi-source multi-probe system and to construct a calibration dataset. The calibration module is used to iteratively solve the geometric parameters of the multi-source multi-detector system according to the coupled X-ray source subset, the coupled detector subset, the uncoupled subset, and the objective function constructed using the calibration dataset, so as to complete the calibration of the overall geometric parameters of the multi-source multi-detector system.

7. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the geometric calibration method for the multi-source multi-probe system according to any one of claims 1-5.

8. A geometric calibration system for a multi-source, multi-probe system, characterized in that, include: A calibration mold, wherein the calibration mold is a mold made of metal or high-density material spheres, and the spheres in the mold are arranged in an alternating spiral pattern on opposite sides; A rotating stage is used to place the calibration model and rotate the calibration model; The electronic device according to claim 7.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed, they implement the geometric calibration method for the multi-source multi-probe system according to any one of claims 1-5.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed, they implement the geometric calibration method for the multi-source multi-probe system according to any one of claims 1-5.

Images