A correction device for geometry and focal point drift of a cone beam CT system and a measuring method thereof

Abstract

This invention relates to the field of geometry and drift correction technology for cone-beam CT systems, specifically providing a device and method for synchronously correcting the geometric position and focus drift of a cone-beam CT system. The device includes a correction phantom, a radiation source unit, a moving unit, and a detection unit. The phantom consists of a transparent pressure plate and positioning steel balls, positioning steel bars, and a cross steel bar sandwiched between them. The positioning steel balls are arranged in a square area, with a pair of vertical sides arranged at equal and non-equal intervals to provide angular identification features. The positioning steel bars are located in the middle of the horizontal sides, and the cross steel bar is located at the center of the area, with the intersection point serving as the drift measurement reference. This invention, through its integrated phantom design, can simultaneously complete geometric parameter calibration and focus drift measurement in a single scan without moving or replacing the phantom. The correction method integrates algorithms such as elliptical trajectory fitting, feature point matching, and bicubic spline interpolation, enabling efficient and accurate calculation of the dynamic curves of detector position, attitude, and focus drift.

Description

Technical Field

[0001] This invention relates to the field of cone-beam CT system correction technology, and more particularly to a correction device and measurement method for geometric position and focus drift of a cone-beam CT system. Background Technology

[0002] Defects can be introduced at every step of the chip manufacturing process, and these defects are one of the main causes of random chip failures, potentially leading to serious consequences. Therefore, defect detection in the chip production process is crucial. It can promptly identify and remove products with defects in previous processes, preventing them from flowing into subsequent processes, thereby effectively controlling costs and material waste, and has significant technical and economic implications. Traditional chip defect detection methods typically require destructive processing of the chip or can only observe its surface structure. However, the widely used Computed Tomography (CT) technology can achieve non-destructive three-dimensional inspection while maintaining the chip's integrity.

[0003] A typical CT system mainly consists of an X-ray source, a rotating sample stage, and a high-resolution area array detector. Among them, micro-CT, with its excellent three-dimensional structural visualization capabilities, is gradually becoming an important tool in micro- and nano-scale detection. The CT imaging process mainly includes four key stages: data acquisition, image reconstruction, image display and processing, and result analysis. In the data acquisition stage, the sample rotates at a constant speed with the rotating stage, or the X-ray source and detector rotate synchronously around the sample. The system pauses at each angular position and acquires two-dimensional projection images of the sample, ultimately obtaining a set of X-ray attenuation projection datasets from multiple angles.

[0004] After projection acquisition, the two-dimensional projection sequence needs to be converted into three-dimensional volume data using a three-dimensional reconstruction algorithm. Cone-beam CT (CBCT) reconstruction widely adopts the FDK algorithm proposed by Feldkamp, ​​Davis, and Kress in 1984. This algorithm is an approximate reconstruction method, and its approximation is mainly reflected in the handling of cone angle effects, enabling the direct reconstruction of three-dimensional tomographic images from two-dimensional projection images. In recent years, GPU-accelerated reconstruction technologies (such as the TIGRE toolbox) have significantly improved reconstruction efficiency. By using CUDA to port the computationally intensive forward and backward projection operations to NVIDIA GPUs, the speed can be increased by tens to hundreds of times compared to traditional CPU implementations, allowing iterative reconstruction tasks that originally took hours to complete to be completed in minutes.

[0005] When applying cone-beam computed tomography (CBCT) for chip defect detection, the FDK algorithm requires certain ideal geometric conditions for reconstruction: the X-ray source focal point, the center of the rotating stage, and the center of the detector plane should be strictly collinear, with the connecting line perpendicular to the detector plane, and the rotation axis parallel to the detector plane. However, in actual system installation and debugging, limitations in mechanical precision make it difficult to perfectly achieve micron or even nanometer-level alignment requirements, leading to varying degrees of geometric errors. These errors introduce geometric artifacts, manifesting as non-realistic structures such as stripes, rings, and bands in the reconstructed image, severely impacting image quality and detection reliability.

[0006] Besides installation geometric errors, X-ray source focus drift is another key factor affecting reconstruction quality. Focus drift causes the line connecting the source, rotation center, and detector center to deviate from the ideal position, thus undermining the reconstruction premise of the FDK algorithm. Especially during long-duration scanning or high-power operation, the X-ray tube target material is prone to deformation due to heat, causing thermal drift of the focus position, resulting in blurred projected images, reduced contrast, and loss of detail. This effect is particularly significant under high magnification imaging conditions. Therefore, achieving high-precision calibration of system geometric parameters and real-time monitoring of focus drift is of great importance for improving reconstruction quality.

[0007] Currently, micro-CT systems typically require separate geometric and focus drift corrections before scanning. Existing correction phantoms are often single-function; geometric correction phantoms cannot simultaneously measure focus drift. Using two separate phantoms for correction significantly increases scan time and X-ray exposures, reducing detection efficiency and accelerating X-ray source filament wear. Furthermore, existing methods still have shortcomings in measuring dynamic focus drift at multiple angles and comprehensively correcting detector attitude. For example:

[0008] Patent CN117075177A proposes to calculate the focus drift by monitoring the drift of the projection center point of the structural component, but it does not consider the possibility that the focus drift may change with the angle during the sample rotation.

[0009] Patent CN103784160B provides a geometric correction device that can correct in-plane rotation errors of the detector, but it is difficult to handle out-of-plane rotation angle deviations of the detector.

[0010] Patent CN118914253B designs a calibration plate based on a sphere array, which is suitable for system geometric parameter calibration, but does not integrate focus drift monitoring function.

[0011] To address the aforementioned issues, this paper proposes a novel integrated correction phantom and corresponding method, aiming to achieve simultaneous and efficient measurement of geometric parameters and focus drift in cone-beam CT systems. This design can simultaneously acquire geometric correction and focus drift information in a single scan, thereby reducing the number of exposures, improving correction accuracy and efficiency, and further enhancing image reconstruction quality through parameter interpolation compensation. Summary of the Invention

[0012] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0013] Addressing the shortcomings of existing cone-beam CT systems, which separate geometric correction and focus drift detection, resulting in cumbersome processes and low efficiency, this invention proposes an integrated correction scheme.

[0014] Therefore, the purpose of this invention is to provide a synchronous correction device for the geometric position and focus drift of a cone-beam CT system, so as to solve the problems of complex correction process and inability to perform geometric parameter calibration and focus drift measurement simultaneously in the prior art.

[0015] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0016] A synchronous correction device for geometric position and focus drift of a cone-beam CT system includes a correction phantom, a radiation source unit, a moving unit, and a detection unit; wherein, the correction phantom includes a transparent pressure plate and a plurality of positioning steel balls disposed between two of the transparent pressure plates, positioning steel bars and cross steel bars distributed along the vertical centerline of the transparent pressure plates.

[0017] As a preferred embodiment of the correction device for geometric position and focus drift of the cone-beam CT system described in this invention, the positioning steel balls form a square area on the plane of the transparent pressure plate, and on the two vertical sides of the square area, a number of positioning steel balls are distributed at equal intervals on one side and a number of positioning steel balls are distributed at non-equal intervals on the other side; the positioning steel strips are distributed in the middle of the two horizontal sides of the square area; the cross steel strips are located in the middle of the square area, and their cross intersection point is offset from the center point of the square area.

[0018] As a preferred embodiment of the correction device for geometric position and focus drift of the cone-beam CT system described in this invention, the positioning steel ball, positioning steel bar and cross steel bar are all made of stainless steel material with a high attenuation coefficient for X-rays.

[0019] As a preferred embodiment of the correction device for geometric position and focus drift of the cone-beam CT system of the present invention, it further includes: a radiation source unit, which includes a radiation source and a support frame for fixing the radiation source; a moving unit, which includes a three-axis moving stage and a rotary stage disposed in the height direction of the three-axis moving stage, wherein the correction phantom is fixed to the rotary stage by a transition plate; and a detection unit, which includes a detector and a support platform for fixing the detector, wherein the correction phantom and the detector are sequentially located in the light extension direction of the radiation source.

[0020] Another objective of this invention is to provide a method for correcting and measuring the geometric position and focal drift of a cone-beam CT system, thereby solving the problems of complex correction processes and the inability to simultaneously measure the focal drift of the X-ray source during geometric correction using the aforementioned correction device.

[0021] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for correcting and measuring the geometric position and focus drift of a cone-beam CT system, comprising the following steps:

[0022] S1: Construct a correction device for the geometric position and focus drift of the cone-beam CT system as described above;

[0023] S2: Initialize the calibration device and measure the initial physical parameters;

[0024] S3: Start the calibration device and perform the geometric calibration operation procedure;

[0025] S4: After the geometric correction operation is completed, perform the focus offset correction operation.

[0026] S5: After the focus offset correction operation is completed, the deflection angle measurement operation of the detector is performed.

[0027] As a preferred embodiment of the correction measurement method for the geometric position and focus drift of the cone-beam CT system described in this invention, the initialization of the correction device and the measurement of initial physical parameters include the following steps:

[0028] Place the X-ray source, rotating stage, and detector in ideal positions, ensuring that the detector plane is perpendicular to the X-ray emitted by the X-ray source, and that the normal direction of the correction phantom is perpendicular to the normal direction of the rotating stage.

[0029] Measure the distance from the X-ray source to the calibration phantom and the distance from the X-ray source to the detector;

[0030] A UV plane coordinate system is established in the detector plane with horizontal and vertical coordinates, and the incident point of the central ray beam generated by the ray source to the detector plane is recorded as the coordinate of the central incident point; at the correction phantom, a three-dimensional XYZ coordinate system is established with the incident direction of the cone beam, the direction perpendicular to the cone beam, and the rotation axis of the rotary table.

[0031] Establish on the plane where the radiation source is located - Planar coordinate system, and - The coordinate system plane is parallel to the UV coordinate system plane;

[0032] Maintain the pixel coordinates of the cone-beam ray passing through the intersection of the crossbars within the calibration model to the detector plane close to the center incident point coordinates. .

[0033] As a preferred embodiment of the correction measurement method for the geometric position and focus drift of the cone-beam CT system described in this invention, the geometric correction operation procedure for the correction device includes the following steps:

[0034] Perform dark field correction on the detector;

[0035] Turn on the X-ray source and continue to perform dead pixel correction and gain correction on the detector;

[0036] The cone-shaped beam of X-ray emitted by the X-ray source is directed toward the correction phantom along the X-axis. The correction phantom is rotated one full turn, and its projection images at various angles are collected. The exposure time of the X-ray source is 2 seconds, and the scanning step size is 0.5°.

[0037] After the data acquisition is completed, the pixel coordinates of the center coordinates of the two selected positioning steel balls on the calibration phantom are calculated. The motion trajectory of the pixel coordinates of each positioning steel ball center is fitted to obtain the corresponding two ellipse equations.

[0038] The equations of the major and minor axes of the ellipse, the coordinates of the two intersection points of the minor axis with the ellipse, and the coordinates of the intersection point of the major axis with the ellipse are calculated. Based on the spatial geometric relationship between the calibration phantom and the detector, the distance from the X-ray source to the detector, the distance from the X-ray source to the calibration phantom, the coordinates of the center incident point of the central X-ray beam emitted by the X-ray source on the detector plane, and the rotation angle of the detector's normal vector are calculated for geometric calibration of the calibration device.

[0039] As a preferred embodiment of the method for correcting and measuring the geometric position and focal drift of the cone-beam CT system described in this invention, the focal offset correction operation includes the following steps:

[0040] The two-dimensional projection dataset of the calibration phantom collected during the geometric calibration operation is used as the actual projection image and saved as slow scan data;

[0041] Without changing the scanning parameters and correcting the phantom position, perform another CT scan with a small number of viewing angles and a short duration to obtain a set of reference projection data, which is then saved as fast scan data.

[0042] By comparing the actual projected image with the reference projected image at the corresponding viewpoint, and through adaptive feature region matching, the drift of the centroid of the crossbar in the U and V directions is calculated, and then the focal point is calculated. - The amount of drift within a planar coordinate system;

[0043] The focus shift at other viewpoints is calculated using bicubic spline interpolation, thereby obtaining the focus shift of the actual projected image at all viewpoints. Finally, the focus is fitted to determine the focal point at... direction and The curve showing how the directional drift changes over time.

[0044] As a preferred embodiment of the method for correcting and measuring the geometric position and focus drift of the cone-beam CT system described in this invention, the procedure for measuring the deflection angle of the detector includes the following steps:

[0045] Turn on the X-ray source, fine-tune the direction of the correction phantom to ensure that the cone-beam X-ray incident on the correction phantom is perpendicular to the plane of the correction phantom, and measure the coordinates of point a on the cross bar inside the correction phantom. ) and the coordinates of point b ( After the detector plane is rotated by a preset angle around its plane normal, the coordinates of points a and b are respectively ( )and( Then, based on geometric relationships, calculate the column rotation angle of the detector plane around its central column. .

[0046] As a preferred embodiment of the method for correcting and measuring the geometric position and focus drift of the cone-beam CT system described in this invention, the procedure for measuring the deflection angle of the detector further includes the following steps:

[0047] The orientation of the correction phantom was fine-tuned again to ensure that the cone-beam rays incident on the correction phantom were perpendicular to the plane of the correction phantom. The coordinates of point c of the cross bar inside the correction phantom were then measured. ) and the coordinates of point d ( ), Correcting the rotation angle of the normal vector of the detector plane Sum and column rotation angle Afterwards, the coordinates of points c and d are respectively ( ) and( Then, based on geometric relationships, calculate the row rotation angle of the detector plane around its center. .

[0048] The beneficial effects of this invention are as follows:

[0049] Compared to existing technologies, this invention features a simple and easy-to-implement structural design, enabling simultaneous geometric parameter correction of the CT system and dynamic measurement of X-ray source focus drift during a single scan without moving or replacing the phantom. This integrated correction process not only significantly improves correction efficiency but also effectively estimates the spatiotemporal distribution characteristics of focus drift throughout the entire scanning process.

[0050] This invention enables precise geometric correction of the micro-CT system, significantly suppressing reconstruction artifacts caused by system installation deviations and geometric mismatches, thereby improving the spatial accuracy and structural clarity of 3D images. Simultaneously, by compensating for and correcting the reconstruction algorithm based on measured focus drift, it further reduces image blurring and contrast degradation caused by X-ray source instability, providing a more reliable and high-quality imaging foundation for the non-destructive testing of precision devices such as chips. Attached Figure Description

[0051] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0052] Figure 1 This is a schematic diagram of the correction phantom structure of the correction device for the geometric position and focus drift of the cone-beam CT system of the present invention;

[0053] Figure 2 This is a schematic diagram of the sandwich planar structure distribution of the correction phantom of the correction device for the geometric position and focus drift correction device of the cone-beam CT system of the present invention;

[0054] Figure 3 This is a schematic diagram of the overall structure of the correction device for geometric position and focus drift in the cone-beam CT system of the present invention;

[0055] Figure 4 This is a schematic diagram illustrating the geometric relationship of the elliptical trajectory formed by the movement of the steel ball in an embodiment of the present invention;

[0056] Figure 5 This is a simplified three-dimensional model schematic diagram of the correction device for geometric position and focus drift of the cone-beam CT system of the present invention;

[0057] Figure 6 This is a schematic diagram of the elliptical trajectories and positions of positioning steel balls 1 and 2 fitted in the detector plane during measurement using the calibration device of the present invention.

[0058] Figure 7This is a schematic diagram illustrating the principle of converting the CT image offset measured by the detector into the X-ray source focus offset in this invention.

[0059] Figure 8 A schematic diagram illustrating the principle of two-dimensional geometric correction for calibrating phantoms;

[0060] Figure 9 Two-dimensional CT images obtained by irradiating the correction phantom 100 with a microfocus X-ray source 201;

[0061] Figure 10 This is a system image of the motion trajectories of steel ball No. 1 and steel ball No. 2 in an embodiment of the present invention, which are fitted to form ellipses.

[0062] Figure 11 This is a slice image of the cone-beam X-ray microscopy system in this embodiment of the invention, after focus drift and geometric correction, to correct the three-dimensional reconstruction of the phantom and remove metal artifacts;

[0063] Figure 12 This is a schematic diagram illustrating the change in focal drift of the microfocus X-ray source in the X and Y directions with the scanning angle in an embodiment of the present invention. Detailed Implementation

[0064] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0065] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0066] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0067] Secondly, the present invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not according to the usual scale. Furthermore, the schematic diagrams are merely examples and should not limit the scope of protection of the present invention. In addition, actual fabrication should include three-dimensional spatial dimensions of length, width, and depth.

[0068] Reference Figures 1-12As a specific embodiment of the present invention, a device for synchronously correcting the geometric position and focal drift of a cone-beam CT system is provided. This device, through its integrated design, enables efficient and accurate calibration of the system's geometric parameters and the amount of X-ray source focal drift, thereby improving the accuracy and reliability of the correction results.

[0069] Combined with appendix Figures 1-4 As shown, the correction device includes a correction phantom 100, a radiation source unit 200, a moving unit 300, and a detection unit 400, each unit being mounted or fixed on a flat and stable optical platform. The correction phantom 100 is a dedicated structure suitable for both geometric correction and focus drift measurement; the radiation source unit 200 generates cone-beam X-rays; the moving unit 300 supports, positions, and adjusts the attitude and position of the correction phantom 100; and the detection unit 400 receives the X-ray signal penetrating the correction phantom 100, forms a projection image, and performs subsequent parameter calculations and correction analysis based on this image.

[0070] Specifically, the calibration phantom 100 includes two parallel transparent pressure plates 101, a number of positioning steel balls 102 sandwiched between the two, two positioning steel bars 103 distributed along the vertical center line of the transparent pressure plates 101, and a cross steel bar 104.

[0071] Furthermore, the positioning steel balls 102 are arranged in a square area within the plane of the transparent pressure plate 101. A positioning steel ball 102 is placed at each of the four vertices of this square area to define the area boundary. Several positioning steel balls 102 are arranged on the vertical sides A and B of the square area: the balls on one side are evenly spaced, while those on the other side are not evenly spaced; this asymmetrical arrangement helps to establish unique spatial reference features in the projection. Two positioning steel bars 103 are respectively positioned at the midpoints of the two horizontal sides C and D of the square area. A cross steel bar 104 is located at the geometric center of the square area; its intersection point is both the center point of the square area and the centroid of the cross steel bar 104, serving as a key reference marker for measuring the focal drift of the X-ray source. To ensure clear identification and positioning of the cross intersection point even when the calibration phantom 100 is rotated to 90°, the cross steel bar 104 is designed with an asymmetrical structure. Specifically, its northwest and southeast sides have shorter segments, while its southeast and south sides have longer segments, as shown in the attached diagram. Figure 2 , Figure 5 As shown in the image.

[0072] At the bottom of the calibration phantom 100, a support body for fixing and connecting is provided. It should be noted that this support body should be made of a material with a low X-ray attenuation coefficient, such as plexiglass, polyvinyl chloride, or resin. The transparent pressure plate 101 can be made of acrylic or other sheet-like materials with good light transmittance; in this embodiment, acrylic is preferred. The positioning steel balls 102, positioning steel strips 103, and cross steel strips 104 must be made of materials with high X-ray attenuation coefficients, such as tungsten, copper, molybdenum, lead, zirconium, or stainless steel. Considering that lead is relatively soft and the resulting lead balls or strips are prone to deformation, affecting geometric accuracy, stainless steel is preferred in this invention.

[0073] Furthermore, the X-ray source unit 200 includes an X-ray source 201 and a support frame 202 for fixing the X-ray source 201. The X-ray source 201 is preferably a microfocus X-ray source for generating cone-beam X-rays; the support frame 202 is used to stably support the X-ray source 201 and ensure its stable position.

[0074] The moving unit 300 includes a three-axis moving stage 301 and a rotary stage 302 mounted vertically (on the Z-axis) of the moving stage. The calibration phantom 100 is fixed to the rotary stage 302 via an adapter plate. The three-axis moving stage 301 can achieve electric translation along three mutually perpendicular directions (X, Y, and Z); the rotary stage 302 is fixed in the Z-axis direction, forming the rotation axis of the system. Thus, the calibration phantom 100 has four degrees of freedom (three-axis translation and rotation around the axis).

[0075] The detection unit 400 includes a detector 401 and a support platform 402 for fixing the detector. The detector 401 is used to receive the X-ray signal after penetrating the correction phantom 100 and form a projection image; the support platform 402 is used to securely mount the detector 401. The correction phantom 100 and the detector 401 are arranged sequentially in the direction of the cone beam extension of the X-ray source 201.

[0076] Reference Figures 1-11 The present invention provides a method for correcting and measuring the geometric position and focus drift of a cone-beam CT system, comprising the following steps:

[0077] S1: Construct the aforementioned cone-beam CT system's geometric position and focus drift correction device; such as... Figure 3 As shown in the image.

[0078] S2: Initialize the calibration device and measure the initial physical parameters;

[0079] Before performing geometric correction, the correction device needs to be initialized. First, the X-ray source 201, the rotating stage 302, and the detector 401 are adjusted to near the theoretical ideal position using conventional calibration methods, so that the plane of the detector 401 is as perpendicular as possible to the plane of the optical platform, while ensuring that the plane of the detector 401 is as perpendicular as possible to the X-rays emitted by the X-ray source 201, and that the normal direction of the correction phantom 100 is perpendicular to the axial direction of the rotating stage 302.

[0080] Before turning on the X-ray source 201, dark field correction must be performed on the detector 401; then the X-ray source 201 is turned on, and bad pixel correction and gain correction are performed on the detector 401 to eliminate the influence of the detector's inherent noise and uneven response on image quality.

[0081] After the detector 401 is calibrated, a two-dimensional projection image is captured on the calibration phantom 100. For example... Figure 4 As shown, when the outlines of the positioning steel ball 102 and the cross steel bar 104 can be clearly distinguished in the image, it indicates that the calibration device has completed initialization.

[0082] The initial physical parameter measurements include: determining the distance from the radiation source 201 to the detector 401 (denoted as...). ), and establish the corresponding coordinate system, such as Figure 5 As shown, the ray source 201 is considered as a point light source S. Specifically, a two-dimensional UV coordinate system with pixels as the unit is established on the plane of detector 401, and the incident point of the central ray beam generated by the ray source 201 to the plane of detector 401 is denoted as the coordinate of the central incident point. A three-dimensional coordinate system XYZ is established at the location of the calibration phantom 100, where the X-axis is the ray incident direction, the Z-axis is parallel to the rotation axis of the rotary table 302, and the Y-axis is perpendicular to the ZX plane.

[0083] Establish an xy-plane coordinate system in the plane where the radiation source 201 is located, and make the xy-plane coordinate system parallel to the UV-plane coordinate system. It should be noted that the x and y axes are in the same direction as the U and V axes.

[0084] Furthermore, before geometric correction begins, the X-ray source 201 is kept on, and projected images of the correction phantom 100 are continuously captured. By fine-tuning the position of the correction phantom 100, the pixel coordinates of the intersection of the crossbars 104 on the detector 401 are made as close as possible to the coordinates of the center incident point corresponding to the central X-ray beam. Based on this, the phantom is used to perform subsequent geometric corrections on the micro-CT imaging system.

[0085] S3: Start the calibration device and perform the geometric calibration operation procedure;

[0086] Furthermore, before performing the formal geometric correction operation, a simple trajectory verification step can be performed first:

[0087] The X-ray source 201 emits a cone-beam X-ray, incident along the X-axis. The rotary table 302 is controlled to rotate the calibration phantom 100 360° around the Z-axis, continuously acquiring projection images from various angles during the rotation. Any positioning steel ball 102 on the calibration phantom 100 is selected, and its pixel position in a series of projection images is tracked to verify that its trajectory conforms to an elliptical shape. This verification can be used to preliminarily confirm the basic rationality of the system's rotation and projection geometry. After successful verification, the following geometric correction operation procedure is performed:

[0088] Keeping the positions of the X-ray source 201 and detector 401 fixed, the cone-beam X-ray emitted by the X-ray source 201 irradiates the correction phantom 100 along the X-axis. The correction phantom 100 is rotated one revolution by the rotary table 302, and its projection images at various angles are acquired. In this embodiment, the X-ray source 201 is preferably set with an exposure time of 2 seconds and a scanning angle step of 0.5°. Positioning steel balls 102 at two specific locations on the correction phantom 100 (marked as No. 1 and No. 2) are selected, and the pixel coordinates of their centers in the projection images at each angle are extracted and denoted as (…). , Based on continuously collected coordinate data, the motion trajectories of the centers of the two steel balls are fitted to obtain the corresponding ellipse equations. The general equation of an ellipse is shown in equation (1):

[0089] (1)

[0090] Further, based on the fitted ellipse equations, the equations of the major and minor axes of the two ellipses, the coordinates of the intersection points of the minor axis and the ellipse, and the coordinates of the intersection points of the major axis and the ellipse are calculated respectively. Combining the spatial geometric relationship between the correction phantom 100 and the detector 401, the following system geometric parameters can be calculated sequentially: the distance from the X-ray source 201 to the detector 401. The distance from the radiation source 201 to the calibration phantom 100 Coordinates of the central incident point and the normal vector rotation angle of detector 401 Ultimately, these parameters can be used to perform geometric correction on the cone-beam CT system.

[0091] Specifically, based on the fitted equation of the elliptical projection curve of the positioning steel ball 102, the coordinates of the minor axis of the ellipse can be calculated. ( , )and ( , ), and the major axis coordinates of the ellipse. ( , )and ( , The length of the minor axis of the ellipse was calculated to be... The length of the major axis is .

[0092] The ordinate of detector 401 was obtained through fitting. Distance between X-ray source and detector The equation is shown in equation (2):

[0093] (2)

[0094] By substituting the corresponding parameters of the two ellipses into the equation, the solution can be obtained by solving the system of equations simultaneously. and .

[0095] minor axis coordinates of the ellipse ( , )and ( , The equation of the straight line determined by the given equation is shown in equation (3):

[0096] (3)

[0097] Where the slope With intercept can be ( , )and ( , The coordinates of the two points are calculated. Since this line lies along the minor axis of the ellipse, and is the ideal projection center of the system (i.e., the center of incidence point), the coordinates are... It lies on this straight line, and therefore satisfies the condition shown in equation (4):

[0098] (4)

[0099] The obtained Substituting into equation (4), the solution can be obtained. The value of .

[0100] The distance from the radiation source 201 to the detector 401 was calculated. The distance from the X-ray source 201 to the correction phantom 100 can then be calculated based on geometric relationships. As shown in equation (5), where Let be the radius of rotation of the steel ball.

[0101] (5)

[0102] Furthermore, the rotation angle of the normal vector of detector 401 is calculated. .

[0103] Ignore the column rotation angle of the detector 401 plane about its central column. and the row rotation angle around its center The resulting impact, calculate the rotation angle of the detector 401 plane around its plane normal vector. The formula is shown in equation (6):

[0104] (6)

[0105] S4: After the geometric correction operation is completed, perform the focus offset correction operation.

[0106] During the geometric correction operation described above, the focal position of the X-ray source 201 may shift due to factors such as the thermal effect of the X-ray tube. Therefore, it is necessary to simultaneously correct the focal drift. The specific procedure is as follows:

[0107] S41: The two-dimensional projection dataset of the calibration phantom 100 acquired during the geometric calibration process is used as the actual projection image and saved as slow scan data. In this step, the X-ray source 201 scans in steps of 0.5°, acquiring a total of 721 projection images;

[0108] S42: Subsequently, without changing any scanning parameters or correcting the position of phantom 100, a short-duration, small-angle rapid CT scan is performed to obtain a set of reference projection data, which is saved as fast scan data. The exposure time of X-ray source 201 can be set to 2 seconds, but the scan step size is 18°, and a total of 21 projection images are acquired. Since the fast scan time is extremely short, it can be approximately assumed that the focus does not drift during this period;

[0109] S43: After the fast scan is completed, save the two-dimensional projection images of the calibration phantom 100 obtained by the slow scan and the fast scan to the corresponding folders respectively.

[0110] The Canny edge detection algorithm is used to identify the outline of the cross steel bar 104 and locate the pixel coordinates of its centroid. The method for locating the pixel coordinates of the centroid of the cross steel bar 104 is as follows: first, find the midpoints of the horizontal and vertical widths of the cross steel bar 104 and draw two straight lines. The two lines have an intersection point. The pixel value of the intersection point is the centroid coordinate of the cross steel bar 104. It should be noted that since the position of the intersection point 'a' in the cross steel bar 104 is exactly its centroid, the centroid is replaced by the intersection point 'a' in this scheme.

[0111] S44: By comparing the pixel coordinate difference between the intersection point a of the crossbar 104 in the U and V directions of the detector 401 plane in the slow and fast scan images at the same scanning angle, the focal point of the X-ray source 201 as a function of scanning time can be deduced. - The amount of drift in the coordinate system plane;

[0112] S45: The slow scan folder contains a total of 721 two-dimensional projection images of the calibration phantom 100. Starting from the first image, one image is selected every 36 images, for a total of 21 slow scan images (the same number as the fast scan images). Slow scan and fast scan images from the same viewpoint are then named uniformly.

[0113] S46: By comparing the actual projected image (slow scan) and the reference projected image (fast scan) under the same viewpoint, the adaptive feature region matching algorithm is used to calculate the pixel drift of the intersection point a of the cross bar 104 in the U and V directions of the detector 401 plane, and then the drift of the focal point of the X-ray source 201 under this viewpoint is deduced.

[0114] S47: Based on the focus drift calculated from these 21 viewpoints, the drift corresponding to the remaining 700 viewpoints is estimated using bicubic spline interpolation, thus obtaining focus drift data for all 721 scanning angles. Finally, the focus is fitted to... direction and A drift curve that changes in direction over time (or scanning sequence).

[0115] Specifically, the principle behind deriving the drift of the X-ray source 201's focal spot over scanning time is as follows: When the X-ray source 201 drifts downwards by a distance ΔS, the projected image of the workpiece on the detector 401 shifts upwards by a distance ΔC. The distance from the focal spot of the X-ray source 201 to the calibration phantom 100 is... The distance from the focal point of X-ray source 201 to detector 401 is ,like Figure 7 As shown, based on the geometric relationship of similar triangles, the formula for calculating the focus drift ΔS is as shown in equation (7):

[0116] (7)

[0117] Specifically, the method for calculating the upward movement of the projected image of the workpiece on detector 401 by a distance ΔC is as follows:

[0118] (1) Take the average value of the actual U coordinate of the centroid of the cross bar 104 in the slow scan image and the U coordinate in the fast scan reference image, and subtract the two to obtain the pixel drift ΔU in the U direction; similarly, subtract the average value of the actual V coordinate and the reference V coordinate to obtain the pixel drift ΔV in the V direction.

[0119] (2) For the actual projected image and the reference projected image at each corresponding scanning angle, an adaptive feature region matching method is used, combined with bicubic spline interpolation, to calculate the overall offset ΔC of the projected image at that angle. Based on this, the actual focal drift of the X-ray source 201 at that viewing angle can be further calculated. .

[0120] S5: After completing the focus drift correction operation, measure the deflection angle of detector 401.

[0121] Specifically, after completing the focal drift test of X-ray source 201, it is necessary to further determine the two out-of-plane rotation angles of detector 401, namely the column rotation angles around its central column. and the row rotation angle around its center The specific steps are as follows:

[0122] First, turn on the X-ray source 201 and fine-tune the direction of the correction phantom 100 to ensure that the incident cone-beam X-ray is perpendicular to the plane of the correction phantom 100. Acquire a projected image of the correction phantom 100 and measure the pixel coordinates of feature point a on the crossbar 104. ) and the pixel coordinates of feature point b ( Then, rotate the detector 401 plane around its normal vector by an angle η (this angle has been determined in the geometric correction), and acquire images again to obtain the pixel values ​​of the new coordinates of point a and point b after rotation. )and( Based on the geometric correspondence between the coordinates before and after rotation, the column rotation angle of the detector 401 plane around its central column can be calculated. .

[0123] Calculate the column rotation angle of the detector plane about its central column. Then, continue measuring the pixel values ​​of the initial coordinates of another pair of feature points c and d on the cross bar 104. )and( ), in correcting the normal vector rotation angle of the detector Sum and column rotation angle Then, the image was re-acquired, and the pixel values ​​of the coordinates of points c and d after rotation were obtained. )and( Based on the geometric transformation of this set of coordinates, the row rotation angle of the detector plane around its center row can be further calculated. .

[0124] Furthermore, before calibration, the calibration phantom 100 is fine-tuned so that the pixel coordinates of the center point of the crossbar 104 are aligned with the coordinates of the center incident point. When the crossbar 104 overlaps, a clear projection of the crossbar 104 onto the detector 401 can be obtained. The projection length of the transverse line segment of the crossbar 104 is measured. When the correction phantom 100 is rotated until the transverse projection length reaches its maximum value, it indicates that the cone beam is now perpendicularly incident on the plane of the correction phantom. This position is defined as the initial zero-degree direction of the correction phantom rotation.

[0125] Correcting the column rotation angle of the detector plane around its central column The process is as follows:

[0126] After fine-tuning the calibration phantom 100 to the initial position described above, a projection image of it is captured using a micro-CT system. The pixel values ​​of the projection points of feature points a and b in the crossbar 104 on detector 401 can be measured as M ( ) and N ( ), and the ideal projection points of feature points a and b on the cross steel bar 104 are point M and point B, respectively. If the detector plane rotates about its normal vector After the angle is adjusted, the pixel values ​​of the projected coordinates of points a and b will be transformed into ( )and( The coordinate transformation relationship is shown in equations (8) and (9):

[0127] (8)

[0128] (9)

[0129] Keeping the position of the calibration phantom 100 unchanged; since the X-axis displacement stage of the three-axis moving stage 301 is equipped with an absolute grating ruler, and the mounting position of the detector 401 on the optical platform G is known, the distance from the calibration phantom 100 to the detector 401 can be accurately measured. Meanwhile, the distance from the radiation source 201 to the detector 401 has been obtained from equation (2). Therefore, the distance from the radiation source 201 to the correction phantom 100 can be calculated. (At this point, the column rotation angle needs to be considered) With row rotation angle (Impact). According to Figure 8 The geometric relationship shown indicates the column rotation angle of the detector plane about its central column. The calculation formula is shown in equation (10):

[0130] (10)

[0131] Correcting the row rotation angle of the detector plane around the center row The process is as follows:

[0132] like Figure 9 As shown, by continuing to photograph the calibration phantom 100 using the X-ray source 201, the pixel values ​​of the projection points of feature points c and d in the crossbar 104 on the detector 401 can be measured as P( ) and Q ( ), and the ideal projection points of feature point c and point d on the cross steel bar 104 are respectively point c and point d. and points Correct the normal vector rotation angle of the detector. Afterwards, the pixel values ​​of points c and d are transformed into ( )and( The transformation relationship is shown by equations (11) and (12):

[0133] (11)

[0134] (12)

[0135] Finally, the column rotation angle of the detector plane around its central column was corrected. After that, at this time ( ) point sum ( The pixel values ​​of the point coordinates are further transformed into ( )and( The transformation formulas are shown in equations (13) and (14); similarly, the following can be calculated. Furthermore, the length of the corrected line segment cd is calculated using the existing formula for the length or distance between two points.

[0136] (13)

[0137] (14)

[0138] according to Figure 6 The geometric relationships shown can be used to derive the row rotation angle of the detector plane around the center row. The calculation formulas are shown in equations (15) and (16):

[0139] (15)

[0140] (16)

[0141] After completing the geometric correction of the cone-beam CT system, a series of geometric correction parameters can be obtained, specifically including: the distance R from the X-ray source 201 to the detector 401. FD The distance R from the radiation source 201 to the correction phantom 100 FI Coordinates of the central incident point The normal vector rotation angle of detector 401 Detector 401 column rotation angle and the row rotation angle of detector 401 The geometric parameters obtained above can be used to correct the geometric position and focus drift issues of the cone-beam CT system.

[0142] Implementation Examples

[0143] Based on the above-mentioned methods for calculating geometric parameters and testing the drift of the X-ray source focus, this invention provides an example.

[0144] In this example, the calibration phantom 100 is placed within the micro-CT imaging system according to the above method. First, dark-field calibration is performed on detector 401. Then, the X-ray source 201 is turned on, and dead pixel and gain calibrations are performed on detector 401. After calibration, the X-ray source 201 is turned on, maintaining the centroid of the crossbar 104 at the center of the incident point coordinates. nearby.

[0145] After rotating the calibration phantom 100 one revolution, the motion trajectories of steel balls 1 and 2 are fitted into ellipses, as shown below. Figure 10 As shown in the diagram; for the red ellipse, the pixel coordinates of the steel ball closest to 1600 on the V-axis are selected from all the pixel values ​​of the steel ball coordinates, and then the selected steel ball coordinates are connected to form an ellipse. For the blue ellipse, the pixel coordinates of the steel ball closest to 1364 on the V-axis are selected from all the pixel values ​​of the steel ball coordinates, and then the selected steel ball coordinates are connected to form an ellipse. Specifically, the red trajectory is the motion trajectory of steel ball number 1, and the corresponding equation is shown in equation (17) (for ease of understanding, the variables in the UV coordinate system are...). , with variables in common equations (Calculations can be performed on an alternative basis)

[0146] (17)

[0147] The equation of the major axis of the ellipse of steel ball No. 1 is calculated as shown in equation (18):

[0148] (18)

[0149] The intersection of the major axis and the ellipse (223.0677, 1386.483) and the intersection point (1882.31, 1382.326); Continuing the calculation, the equation of the minor axis of the ellipse of steel ball No. 1 is shown in equation (19):

[0150] (19)

[0151] The intersection of the minor axis and the ellipse (1052.588, 1344.243) and the intersection point (1052.79, 1424.567); Next, calculate steel ball number 2 (e.g. Figure 10 The equation for the motion trajectory of the blue trajectory is shown in equation (20):

[0152] (20)

[0153] The equation of the major axis of the ellipse of steel ball No. 2 is calculated as shown in equation (21):

[0154] (twenty one)

[0155] The intersection of the major axis and the ellipse (225.7, 1624.22) and the intersection point (1879.99, 1620.08);

[0156] The equation of the minor axis of the ellipse of steel ball No. 2 is obtained by further calculation as shown in equation (22):

[0157] (twenty two)

[0158] The intersection of the minor axis and the ellipse (1052.65, 1544.17) and the intersection point : (1053.04, 1700.13).

[0159] In this embodiment, based on the calculated elliptical equations and major and minor axis equations of steel balls 1 and 2, the distance from the radiation source 201 to the detector 401 can be calculated using the aforementioned method. The distance from the X-ray source 201 to the center of rotation of the calibration phantom 100 Coordinates of the center incident point ( , and the normal rotation angle of the detector. The calculation results are shown in Table 1. Then, based on the aforementioned two-dimensional geometric correction method, the rotation angle of the detector array can be further calculated. and detector rotation angle The calculation results are shown in Table 1, where the pixel size of the CT projection image is 2048×2048.

[0160] After geometric correction, the obtained correction parameters are input into the reconstruction algorithm for 3D reconstruction. The sliced ​​images of the reconstructed phantom are shown below. Figure 11 As shown, the image geometry is accurate, and artifacts such as stripes and rings caused by system geometric mismatch have been largely eliminated, indicating that the geometric correction has achieved the expected results.

[0161] Furthermore, a major advantage of the method proposed in this invention is that it can simultaneously achieve dynamic monitoring of the focal position drift of the X-ray source in a single scan without disturbing the correction phantom. Using the method described, the position of the X-ray source 201 at each scanning angle was calculated. direction and The amount of focus drift in the direction, and its drift trajectory as follows Figure 12 As shown. Analysis reveals that the focus is on The directional drift is more significant, while... The directional drift is almost negligible, and the maximum drift does not exceed the size range of the focal point itself. This work not only verifies the effectiveness of geometric correction but also provides accurate data for introducing focal drift compensation into subsequent reconstruction algorithms.

[0162] Table 1 Geometric parameter values ​​of cone-beam CT system after geometric correction

[0163]

[0164] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the methods and apparatus without departing from the spirit and scope of the present invention, and such modifications and substitutions should also be considered to fall within the protection scope of the present invention.

Claims

1. A device for correcting cone beam CT system geometry and focal spot drift, characterized by: include, The calibration phantom (100) includes a transparent pressure plate (101) and a plurality of positioning steel balls (102) disposed between the two transparent pressure plates (101), positioning steel bars (103) and cross steel bars (104) distributed along the vertical center line of the transparent pressure plate (101). The positioning steel balls (102) form a square area on the plate surface of the transparent pressure plate (101), and on the two vertical sides (A, B) of the square area, a number of positioning steel balls (102) are distributed at equal intervals on one side and a number of positioning steel balls (102) are distributed at non-equal intervals on the other side. The positioning steel strips (103) are distributed in the middle of the two horizontal sides (C, D) of the square area; The cross bar (104) is located in the middle of the square area, and its cross intersection point is offset from the center point of the square area.

2. The correction device for cone beam CT system geometry and focal spot drift according to claim 1, characterized in that: The positioning steel ball (102), positioning steel bar (103), and cross steel bar (104) are all made of materials with a large radiation attenuation coefficient.

3. The correction device for cone beam CT system geometry and focal spot drift according to claim 2, characterized in that: It also includes, The radiation source unit (200) includes a radiation source (201) and a support frame (202) for fixing the radiation source (201). The moving unit (300) includes a three-axis moving stage (301) and a rotating stage (302) disposed in the height direction of the three-axis moving stage (301). The correction model (100) is fixed on the rotating stage (302) by a transition plate. The detection unit (400) includes a detector (401) and a support platform (402) for fixing the detector (401). The calibration phantom (100) and the detector (401) are located sequentially in the light extension direction of the radiation source (201).

4. A method for correcting and measuring the geometric position and focal drift of a cone-beam CT system, characterized in that: Includes the following steps: Construct a correction device for the geometric position and focus drift of the cone-beam CT system as described in claim 3; Initialize the calibration device and measure the initial physical parameters; Start the calibration device and perform the geometric calibration operation procedure; After the geometric correction procedure is completed, the focus offset correction procedure is performed. After the focus offset correction operation is completed, the deflection angle measurement operation of the detector (401) is performed.

5. The method for correcting and measuring the geometric position and focal drift of a cone-beam CT system according to claim 4, characterized in that: The initialization and correction device, and the measurement of initial physical parameters, include: Place the X-ray source (201), the rotating stage (302), and the detector (401) in ideal positions, keeping the plane of the detector (401) perpendicular to the X-ray emitted by the X-ray source (201), and the normal direction of the correction phantom (100) perpendicular to the normal direction of the rotating stage (302). The distance from the X-ray source (201) to the calibration phantom (100) and the distance from the X-ray source (201) to the detector (401) are measured. A UV plane coordinate system is established in the detector (401) plane with horizontal and vertical coordinates, and the incident point of the central ray beam generated by the ray source (201) to the detector (401) plane is recorded as the coordinate of the central incident point. At the calibration phantom (100), an XYZ three-dimensional coordinate system is established with the incident direction of the cone beam, the direction perpendicular to the cone beam, and the rotation axis of the rotary table (302). Establish a plane at the location of the radiation source (201). - Planar coordinate system, and - The coordinate system plane is parallel to the UV coordinate system plane; The cone-beam ray is kept passing through the intersection of the crossbars (104) inside the correction phantom (100), such that the pixel coordinates of this intersection on the detector (401) plane are close to the coordinates of the center incident point. .

6. The method for correcting and measuring the geometric position and focal drift of a cone-beam CT system according to claim 5, characterized in that: The geometric correction operation procedure for the correction device includes: Dark field correction is performed on detector (401); Turn on the X-ray source (201) and continue to perform dead pixel correction and gain correction on the detector (401); The cone beam emitted by the X-ray source (201) is directed toward the correction phantom (100) along the X-axis direction. The correction phantom (100) is rotated one revolution, and its projection images at various angles are collected. After the data collection is completed, the pixel coordinates of the center coordinates of the two selected positioning steel balls (102) on the calibration phantom (100) are calculated. The motion trajectory of the pixel coordinates of the center of each positioning steel ball (102) is fitted to obtain the corresponding two elliptical equations. The equations of the major and minor axes of the ellipse, the coordinates of the two intersection points of the minor axis and the ellipse, and the coordinates of the intersection point of the major axis and the ellipse are calculated. Based on the spatial geometric relationship between the calibration phantom (100) and the detector (401), the distance from the X-ray source (201) to the detector (401), the distance from the X-ray source (201) to the calibration phantom (100), and the coordinates of the center incident point of the central X-ray beam emitted by the X-ray source (201) on the plane of the detector (401) are calculated. And the normal vector rotation angle of the detector (401) is used for the geometric correction of the correction device.

7. The method for correcting and measuring the geometric position and focal drift of a cone-beam CT system according to claim 6, characterized in that: The process for performing focus offset correction includes: The two-dimensional projection dataset of the correction phantom (100) collected during the geometric correction operation is used as the actual projection image and saved as slow scan data; Without changing the scanning parameters and the position of the correction phantom (100), perform another CT scan with a small number of viewing angles and a short duration to obtain a set of reference projection data, which is then saved as fast scan data. Compare the actual projected image with the reference projected image under the corresponding viewpoint, and calculate the drift of the centroid of the cross bar (104) in the U and V directions by adaptive feature region matching, and then calculate the drift of the focal point in the xy plane coordinate system. The focus shift at other viewpoints is calculated using bicubic spline interpolation, thereby obtaining the focus shift of the actual projected image at all viewpoints. Finally, the focus is fitted to determine the focal point at... direction and The curve showing how the directional drift changes over time.

8. The method for correcting and measuring the geometric position and focal drift of a cone-beam CT system according to claim 7, characterized in that: The procedure for measuring the deflection angle of the detector (401) includes: Turn on the X-ray source (201), fine-tune the direction of the calibration phantom (100) to ensure that the cone-beam X-ray incident on the calibration phantom (100) is perpendicular to the plane of the calibration phantom (100), and measure the coordinates of point a of the cross bar (104) inside the calibration phantom (100). ) and the coordinates of point b ( After the detector (401) plane is rotated by a preset angle around its plane normal, the coordinates of points a and b are respectively ( )and( Then, based on geometric relationships, calculate the column rotation angle of the detector (401) plane around its central column. .

9. The method for correcting and measuring the geometric position and focal drift of a cone-beam CT system according to claim 8, characterized in that: The procedure for measuring the deflection angle of the detector (401) also includes, The orientation of the correction phantom (100) is finely adjusted again to ensure that the cone-beam rays incident on it are perpendicular to the plane of the correction phantom (100). The coordinates of point c of the cross steel bar (104) inside the correction phantom (100) are then measured. ) and the coordinates of point d ( ), Correcting the normal vector rotation angle of the detector (401) plane Sum and column rotation angle Afterwards, the coordinates of points c and d are respectively ( ) and( Then, based on geometric relationships, calculate the row rotation angle of the detector (401) plane around its center. .

Images