A method and apparatus for calibrating the geometric parameters of a C-arm CT system.

Abstract

This invention discloses a method and apparatus for calibrating geometric parameters in a C-arm CT system, comprising: Step 1, acquiring projection data of a phantom in circular trajectory cone-beam and C-arm scanning modes; Step 2, solving for the intrinsic parameters of the system that do not change with angle using a circular trajectory geometric parameter calibration method; Step 3, determining the projection addresses of marker points on the phantom onto the detector; Step 4, establishing an optimization model for the predicted projection addresses of marker points at a single scanning angle in the phantom coordinate system and the mean square error of extracting the projection addresses of marker points; Step 5, solving the optimization model to obtain extrinsic parameters that change with angle, and using the estimated intrinsic and extrinsic parameters for reconstruction. This invention is based on the widely used helical phantom, eliminating the need for additional phantom fabrication and calibration, and solves for geometric parameters separately, thus improving the stability of geometric parameter calibration and ensuring the accuracy and precision of the solution.

Description

Technical Field

[0001] This invention relates to the field of X-ray CT (Computed Tomography) imaging technology, and in particular to a method and apparatus for calibrating geometric parameters in a C-arm CT system. Background Technology

[0002] C-arm CT offers multiple scanning modes, allowing interventional physicians to perform intraoperative 3D imaging of patients in the operating room, and its application prospects are broad. A typical C-arm CT device mainly includes: a C-gantry, an X-ray source, X-ray detection equipment (such as an image intensifier and flat panel detector for image acquisition), mechanical motion and control devices, and an image processing workstation. The X-ray source and detector are located at opposite ends of the C-gantry. Typically, the mechanical motion device provides various movement modes, such as vertical lifting, horizontal extension, full-angle rotation, and track sliding. Moreover, different movement modes can be coordinated to achieve multiple scanning modes, meeting the scanning imaging needs of different patients and greatly facilitating intraoperative and interventional imaging. However, this open design can cause mechanical instability, resulting in C-arm wobbling during scanning, which in turn causes geometric artifacts in the reconstructed images, leading to low image resolution. When geometric artifacts are severe, they can affect the physician's interpretation.

[0003] Currently, geometric parameter calibration methods for this type of problem mainly include: analytical calibration methods based on phantoms, iterative calibration methods based on phantoms, image-based iterative calibration methods, and calibration methods that use cameras or sensors for real-time monitoring to obtain the relative position information of various components of the CT system. Analytical solution algorithms often ignore some parameters or have high requirements for phantom design; image-based iterative algorithms are usually computationally intensive and have requirements for image features; while calibration methods using external equipment have high requirements for camera placement and sensor sensitivity. Optimization algorithms based on phantoms are widely used because they can solve all geometric parameters and phantoms are simple to fabricate. Although helical phantoms have good 3D spatial information and can be used to optimize and solve geometric parameters angle by angle, the high correlation between geometric parameters means that the choice of initial values ​​and the accuracy of phantom markers ultimately affect the accuracy of the parameters. Summary of the Invention

[0004] The purpose of this invention is to provide a method and apparatus for calibrating geometric parameters in a C-arm CT system to overcome or at least mitigate at least one of the aforementioned defects of the prior art.

[0005] To achieve the above objectives, the present invention provides a method for calibrating geometric parameters in a C-arm CT system, comprising:

[0006] Step 1: Place the turntable in the middle of the C-frame and place the spiral mold vertically on the turntable. The mold rotates with the turntable while the C-arm remains stationary. This obtains the projection data of the mold under the circular trajectory cone beam scanning method. Place the stage in the middle of the C-frame and place the mold horizontally on the stage. Use the C-arm to scan while keeping the mold stationary to obtain the projection data of the mold under the C-arm scanning method.

[0007] Step 2: Based on the projection data of the phantom obtained from the circular trajectory cone beam scanning in Step 1, the intrinsic parameters of the system that do not change with the angle are solved using the circular trajectory geometric parameter calibration method; among which, the intrinsic parameters include the vertical distance F from the focal point of the X-ray source to the detector, and the perpendicular coordinates (V) of the focal point of the X-ray source on the detector. S W S );

[0008] Step 3: Based on the projection data of the C-arm scanning phantom and the total number of marker points in Step 1, determine the projection address of the i-th marker point r on the detector.

[0009] Step 4: Establish the predicted projection address of the i-th marker point r under a single scanning angle in the phantom coordinate system. And extract the projection address of the i-th marker point r An optimized model for the mean square error;

[0010] Step 5: Solve the optimization model to obtain the extrinsic parameters that vary with the angle, and use the estimated intrinsic and extrinsic parameters for reconstruction; wherein, the extrinsic parameters include the focal coordinates of the ray source (s). x s y s z The in-plane rotation angle α, pitch angle β, and tilt angle γ of the detector.

[0011] Furthermore, the phantom includes a cylindrical support body. A predetermined number of N and predetermined positions of marker points r are set in the wall between the inner and outer walls of the support body. The marker points r are arranged in a spiral shape from one end of the support body to the other end. The centers of the marker points r can be connected in sequence to form a spiral line. The central axis of the spiral line in step 1 is basically consistent with the rotation axis of the turntable. The projected image is the projection image of X-rays penetrating the phantom on the flat panel detector.

[0012] Furthermore, the marker point r is a spherical structure with a preset diameter.

[0013] Furthermore, step 2 specifically includes:

[0014] Step 21: By minimizing the mean square error between the projected address of the marker point R at multiple angles and the extracted projected address, the optimization model for the geometric parameter calibration of the circular trajectory is established as follows (1): the vector of the detector in the horizontal direction is set as V = (V X V Y V Z ) T The vertical vector of the detector is set as W = (W X W Y W Z ) T V X V Y V Z W X W Y W Z The components of the detector's horizontal and vertical direction vectors in the circular trajectory coordinate system on the X, Y, and Z axes, respectively:

[0015]

[0016] Where, ξ k ξ 1 These are the geometric parameters of the circular trajectory cone-beam scanning system at the k-th and 1-th scanning angles, respectively. SDD is the intersection point O of the extension of the line from the focal point of the X-ray source to the rotation axis and the detector. S The distance between the detector and the focal point of the X-ray source, A, B, and Γ are the rotation angles of the detector around the three coordinate axes, respectively. O W O () is the intersection point O S The coordinates on the detector, M is the total number of scanning angles, N represents the total number of marker points R, and E(ξ) 1 Let θ be the objective function. k This represents the angle value of the k-th scan angle. These are the projection addresses of the k-th scanning angle and the ith marker point R in the horizontal and vertical directions parallel to the detector row, respectively. Let R be the predicted k-th scanning angle and the projected addresses of the ith marker point R in the horizontal and vertical directions parallel to the detector row, respectively, as described by the following formula (2). Let R represent the coordinates of the i-th marker point R in the X, Y, and Z directions respectively in the circular trajectory coordinate system under the first scanning angle, as described by the following formula (3):

[0017]

[0018] Step 22: Calculate the system intrinsic parameters using the following formula (4). The intrinsic parameters include the vertical distance F from the focal point of the X-ray source to the detector and the perpendicular coordinate (V) of the focal point of the X-ray source on the detector.S W S S represents the focal point of the radiation source.

[0019]

[0020] Where U = V × W, U represents the direction vector of the normal direction of the detector plane, U X This represents the X-axis component of U in the circular trajectory coordinate system.

[0021] Furthermore, the optimization model for the mean squared error in step 4 is described by the following equation (5):

[0022]

[0023] Where η represents the extrinsic parameter of the system to be solved. The projected address of the i-th marker point r is calculated based on the estimated parameter η;

[0024] The projected address of any point r in the phantom coordinate system is expressed as the following equation (6):

[0025]

[0026] In the formula, r = (r x r y r z ), (v r w r ) represents the projection address of the marker point r onto the detector, and U, V, and W are vectors representing the normal direction of the plane where the detector is located, the horizontal direction parallel to the row of detectors, and the vertical direction parallel to the column of detectors, respectively.

[0027] The present invention also provides a calibration device for geometric parameters in a C-arm CT system, comprising:

[0028] The system intrinsic parameter calculation unit is used to obtain the projection data of the helical phantom in circular trajectory cone-beam scanning mode by rotating the phantom with the turntable while keeping the C-arm stationary, when the helical phantom is placed vertically on the turntable in the middle of the C-frame. Then, the system intrinsic parameters that do not change with the angle are solved using the circular trajectory geometric parameter calibration method. Among them, the intrinsic parameters include the vertical distance F from the focal point of the X-ray source to the detector, and the perpendicular coordinates (V) of the focal point of the X-ray source on the detector. S W S );

[0029] The projection image acquisition unit is used to acquire the projection data of the model by scanning with a C-arm while keeping the model stationary on a platform in the middle of a C-frame.

[0030] The optimization model building unit is used to determine the projection address of the i-th marker r on the detector based on the projection data of the C-arm scanning phantom and the total number of marker points. And establish the predicted projection address of the i-th marker point r under a single scanning angle in the phantom coordinate system. And extract the projection address of the i-th marker point r An optimized model for the mean square error;

[0031] The intrinsic and extrinsic parameter reconstruction unit is used to solve the optimization model, obtain the extrinsic parameters that vary with the angle, and use the estimated intrinsic and extrinsic parameters for reconstruction; wherein, the extrinsic parameters include the focal coordinates of the ray source (s). x s y s z The in-plane rotation angle α, pitch angle β, and tilt angle γ of the detector.

[0032] Furthermore, the phantom includes a cylindrical support body. Within the wall between the inner and outer walls of the support body, there are a preset number N of marker points r marked in a preset position and in a predetermined order. The marker points r are arranged in a spiral shape from one end of the support body to the other end. The centers of the marker points r can be connected in sequence to form a spiral line. The central axis of the spiral line in the system intrinsic parameter calculation unit is basically consistent with the rotation axis of the turntable. The projected image is the projection image of X-rays penetrating the phantom on the flat panel detector.

[0033] Furthermore, the marker point r is a spherical structure with a preset diameter.

[0034] Furthermore, the system intrinsic parameter calculation unit specifically includes:

[0035] Step 21: By minimizing the mean square error between the projected address of the marker point R at multiple angles and the extracted projected address, the optimization model for the geometric parameter calibration of the circular trajectory is established as follows (1): the vector of the detector in the horizontal direction is set as V = (V X V Y V Z ) T The vertical vector of the detector is set as W = (W X W Y W Z ) T V X V Y V Z W X W Y W Z The components of the detector's horizontal and vertical direction vectors in the circular trajectory coordinate system on the X, Y, and Z axes, respectively:

[0036]

[0037] Where, ξ k ξ 1 These are the geometric parameters of the circular trajectory cone-beam scanning system at the k-th and 1-th scanning angles, respectively. SDD is the intersection point O of the extension of the line from the focal point of the X-ray source to the rotation axis and the detector. S The distance between the detector and the focal point of the X-ray source, A, B, and Γ are the rotation angles of the detector around the three coordinate axes, respectively. O W O () is the intersection point O S The coordinates on the detector, M is the total number of scanning angles, N represents the total number of marker points R, and E(ξ) 1 Let θ be the objective function. k This represents the angle value of the k-th scan angle. These are the projection addresses of the k-th scanning angle and the ith marker point R in the horizontal and vertical directions parallel to the detector row, respectively. Let R be the predicted k-th scanning angle and the projected addresses of the ith marker point R in the horizontal and vertical directions parallel to the detector row, respectively, as described by the following formula (2). Let R represent the coordinates of the i-th marker point R in the X, Y, and Z directions respectively in the circular trajectory coordinate system under the first scanning angle, as described by the following formula (3):

[0038]

[0039] Step 22: Calculate the system intrinsic parameters using the following formula (4). The intrinsic parameters include the vertical distance F from the focal point of the X-ray source to the detector and the perpendicular coordinate (V) of the focal point of the X-ray source on the detector. S W S S represents the focal point of the radiation source.

[0040]

[0041] Where U = V × W, U represents the direction vector of the normal direction of the detector plane.

[0042] Furthermore, in step 5, the optimization model for a single scanning angle in the phantom coordinate system is established as follows (5):

[0043]

[0044] Where η represents the system extrinsic parameter to be solved, f is the perpendicular distance from the focal point of the X-ray source to the detector, (v s w s Let be the perpendicular coordinates of the focal point of the X-ray source onto the detector. Let r be the projected coordinates of the extracted i-th marker point. The projected address of the i-th marker point r calculated based on parameter η;

[0045] The projected address of any point r in the phantom coordinate system is expressed as the following equation (6):

[0046]

[0047] In the formula, r = (r x r y r z ), (v r w r ) represents the projection address of the marker point r onto the detector, and U, V, and W are vectors representing the normal direction of the plane where the detector is located, the horizontal direction parallel to the row of detectors, and the vertical direction parallel to the column of detectors, respectively.

[0048] The present invention has the following advantages due to the adoption of the above technical solutions:

[0049] 1. This invention separates the geometric parameters, allowing for accurate estimation of intrinsic parameters and good decorrelation of extrinsic parameters, making them easy to estimate using the least squares method. This method improves the robustness of parameter solving, and its practicality has been well verified in actual C-arm CT systems.

[0050] 2. This invention is based on a widely used spiral phantom. Compared with another commonly used double-circle calibration phantom, this phantom has a better spatial distribution and is easier to adjust. By designing a suitable phantom, the error in parameter solving caused by overlapping marker points can be avoided. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of a spiral mold.

[0052] Figure 2 This is a flowchart illustrating the calibration of geometric parameters in a C-arm CT system according to an embodiment of the present invention.

[0053] Figure 3 A coordinate system diagram illustrating the geometrical relationships between the X-ray source, turntable, and detector.

[0054] Figure 4a This is a schematic diagram of the projection of the model at a scanning angle.

[0055] Figure 4b A schematic diagram of the maximum density projection of a phantom projection for half a circle of ideal circular trajectory conical beam scanning.

[0056] Figure 4c This is a schematic diagram of the maximum density projection of the phantom projection during a half-circle scan of the C-arm.

[0057] Figure 5aThe tomographic map is a reconstruction of an unlabeled phantom.

[0058] Figure 5b To calibrate the maximum density projection of the reconstructed image.

[0059] Figure 5c This is a tomographic map of the reconstructed phantom after calibration.

[0060] Figure 5d It is the maximum density projection of the reconstructed image after calibration.

[0061] Figure 6a and Figure 6b Two tomographic images reconstructed for an unlabeled sample.

[0062] Figure 6c and Figure 6d Two tomographic images reconstructed from the calibrated sample. Detailed Implementation

[0063] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0064] In this invention, the C-arm is a rigid structure, and the geometric parameters are divided into two parts based on whether they change with the scanning angle. A high-precision circular trajectory cone-beam scanning mode is constructed using hardware-aided equipment, and the intrinsic parameters are solved using a circular trajectory geometric parameter calibration method. Then, an optimization model is established by minimizing the mean square error between the actual and predicted projection addresses of the marker points, and the optimization model is solved angle by angle using the information from the phantom marker points to calibrate the extrinsic parameters that change with the angle.

[0065] The method for calibrating geometric parameters in a C-arm CT system provided in this embodiment of the invention includes:

[0066] Step 1: Place the high-precision turntable in the center of the C-frame and vertically place the helical phantom on the turntable. The phantom rotates with the turntable while the C-arm remains stationary, thereby obtaining the projection image of the helical phantom on the flat panel detector under the circular trajectory conical beam scanning mode. Alternatively, place the stage in the center of the C-frame and horizontally place the phantom on the stage. Scan with the C-arm while keeping the phantom stationary. The C-arm scans in the vertical plane to obtain the projection image of the phantom under the C-arm scanning mode.

[0067] The structure of the phantom is as follows: Figure 1As shown, the phantom includes a cylindrical support 1. Within the wall 13 between the inner wall 11 and the outer wall 12 of the support 1, there are a predetermined number N of marker points R, each marked in a predetermined order. The marker points R are arranged spirally from one end of the support 1 to the other. Connecting the centers of the marker points R sequentially forms a spiral. The central axis of the spiral in step 1 is substantially aligned with the rotation axis of the turntable. The projected image is the projection image of X-rays penetrating the phantom onto a flat panel detector. The predetermined positions are determined based on the structure of the phantom to facilitate sequential marking.

[0068] In this embodiment, the material of the support 1 is generally chosen to have a high radiation absorption rate and a sturdy structure that is not easily deformed, such as the plexiglass used in this embodiment. The marker point R is a spherical structure with a preset diameter, not exceeding 2mm. The diameter and the total number of marker points only need to ensure that their projections do not overlap, but they cannot be too small, as this would make it difficult to segment and extract the centroid of the projection. The spherical structure can be a steel ball or a small metal sphere; it has high absorption rate, is easy to manufacture, and facilitates image processing related work on it and its projection.

[0069] When placing the helical phantom, ensure that the central axis of the helix is ​​approximately aligned with the rotation axis of the C-arm CT system, i.e., place it horizontally on the stage. To fully utilize the information from the marker points, the phantom should not extend beyond the field of view. Scan the phantom with the C-arm while keeping it stationary.

[0070] The projection of the phantom is shown in Figure 4. For easy comparison, Figure 4b The maximum density projection of the half-circle projection of the scanning phantom under an ideal scanning trajectory is given. Figure 4c It is the maximum density projection of half a circle of the C-arm scanning phantom. Figure 4b The trajectory is a semi-circular ellipse, while Figure 4c Irregular fluctuations then occurred.

[0071] Before calibrating the geometric parameters of the circular trajectory, it is necessary to establish a geometric system for the circular trajectory cone-beam scanning. This geometric system includes, for example, the following: Figure 3 The coordinate system and selected geometric parameters are used for characterization. In this embodiment, as shown... Figure 3 As shown, O is the foot of the perpendicular from the ray source to the rotation axis, and also the origin of the circular trajectory coordinate system. S This is the intersection point of the extension of the vertical distance SOD from the X-ray source to the rotation axis and the detector. This embodiment uses seven geometric parameters to characterize the circular trajectory cone-beam scanning system. These parameters include the vertical distance SOD from the X-ray source's focal point to the rotation axis, the distance SOD between the intersection point OS of the extension of the vertical distance SOD from the X-ray source's focal point to the rotation axis and the detector, and the rotation angles A, B, and Γ of the detector around the three coordinate axes. O W O() is the intersection point O S The coordinates of the detector in the horizontal direction.

[0072] Step 2: Based on the projection data of the phantom obtained from the circular trajectory cone beam scanning in Step 1, such projection data includes, for example: the vertical distance SOD from the focal point of the X-ray source to the rotation axis, and the intersection point O of the extension line of the vertical distance SOD from the focal point of the X-ray source to the rotation axis and the detector. S The distance SDD between the X-ray source and the focal point, the scanning angle of the projection data, the detector resolution, and the detector element size are used to solve for the intrinsic parameters of the system that do not change with the angle using the circular trajectory geometric parameter calibration method. Among these intrinsic parameters in the C-arm coordinate system, the perpendicular distance f from the X-ray source's focal point to the detector, and the perpendicular coordinate (v) of the X-ray source's focal point on the detector are used. s w s In the circular trajectory coordinate system, the intrinsic parameters f and (v) are... s w s ) can be described as F, (V s W s In essence, both refer to the same parameter, only the symbols used differ in different coordinate systems. Similarly, the symbol R for the marker point is used in the circular trajectory coordinate system, while r is used in the C-arm coordinate system. However, the number of marker points N, the number of scanning angles M, the nth ball i, and the nth scanning angle k use the same symbols in both coordinate systems.

[0073] As a preferred embodiment of step 2, it specifically includes:

[0074] Step 21: By minimizing the mean square error between the projected address of the marker point R at multiple angles and the extracted projected address, the optimization model for the geometric parameter calibration of the circular trajectory is established as follows (1): the detector's horizontal direction vector is set to V = (V X V Y V Y ) T The vertical direction vector of the detector is set to W = (W X W Y W Z ) T V X V Y V Y W represents the components of the detector's horizontal direction vector along the X, Y, and Z axes in the circular trajectory coordinate system. X W Y W Z The components of the vertical direction vector of the detector in the circular trajectory coordinate system on the X, Y, and Z axes are as follows:

[0075]

[0076] Where, ξ k ξ 1 These are the geometric parameters of the circular trajectory cone-beam scanning system at the k-th and 1-th scanning angles, respectively. SDD is the intersection point O of the extension of the line from the focal point of the X-ray source to the rotation axis and the detector. S The distance between the detector and the focal point of the X-ray source, A, B, and Γ are the rotation angles of the detector around the three coordinate axes, respectively. O W O () is the intersection point O S The coordinates on the detector, M is the total number of scanning angles, N represents the total number of marker points R, and E(ξ) 1 Let θ be the objective function. k This represents the angle value of the k-th scan angle. These are the projection addresses of the k-th scanning angle and the ith marker point R in the horizontal and vertical directions parallel to the detector row, respectively. Let R be the predicted k-th scanning angle and the projected addresses of the ith marker point R in the horizontal and vertical directions parallel to the detector row, respectively, as described by the following formula (2). Let R represent the coordinates of the i-th marker point R in the circular trajectory coordinate system under the first scanning angle in the X, Y, and Z directions, respectively, as described by the following formula (3);

[0077]

[0078] The spatial coordinates of the marker point at different angles can be obtained from the spatial coordinates of the initial scan angle and the rotation angle, where the rotation angle is known.

[0079] Step 22: Calculate the system intrinsic parameters using the following formula (4). The intrinsic parameters include the vertical distance F from the focal point of the X-ray source to the detector and the perpendicular coordinate (V) of the focal point of the X-ray source on the detector. S W S S represents the focal point of the radiation source.

[0080]

[0081] Where U = V × W, U represents the direction vector of the normal direction of the detector plane.

[0082] In one embodiment, the optimization model for the mean squared error in step 4 is described by the following equation (5):

[0083]

[0084] Where η represents the system extrinsic parameter to be solved, f is the perpendicular distance from the focal point of the X-ray source to the detector, (vs w s Let be the perpendicular coordinates of the focal point of the X-ray source onto the detector. Let r be the projected coordinates of the extracted i-th marker point. The projected address of the i-th marker point r is calculated based on the estimated parameter η;

[0085] The projected address of any point r in the phantom coordinate system is expressed as the following equation (6):

[0086]

[0087] In the formula, r = (r x r y r z ), (v r w r ) represents the projection address of the marker point r onto the detector, and u, v, and w represent the normal direction of the plane where the detector is located, the horizontal direction parallel to the row of detectors, and the vertical direction parallel to the column of detectors, respectively.

[0088] Step 3: Based on the projection data of the C-arm scanning phantom, the total number of marker points, and the scanning angle from Step 1, determine the projection address of the i-th marker point r on the detector. By scanning the projection data of the phantom with a C-arm and using existing technologies, the projection image of the phantom, detector resolution, and total number of scanning angles can be obtained.

[0089] The marker point refers to the centroid of the marker point on the phantom. The position of the centroid projection is obtained from the projection of the marker point on the phantom onto the two-dimensional elliptical surface of the detector using image processing methods. In addition to the method of extracting the marker point projection during the circular trajectory calibration process, this step can also be implemented using image processing methods such as threshold segmentation, top-hat transformation, and connected component marking.

[0090] Step 4: Establish the predicted projection address of the i-th marker point r under a single scanning angle in the phantom coordinate system. And extract the projection address of the i-th marker point r An optimized model for the mean square error.

[0091] In one embodiment, the optimization model for the mean squared error can be described by the following equation (5):

[0092]

[0093] Where η represents the system extrinsic parameter to be solved, f is the perpendicular distance from the focal point of the X-ray source to the detector, (v s w s Let be the perpendicular coordinates of the focal point of the X-ray source onto the detector. Let r be the projected coordinates of the extracted i-th marker point. The projected address of the i-th marker point r is calculated based on the estimated parameter η;

[0094] The projected address of any point r in the phantom coordinate system is expressed as the following equation (6):

[0095]

[0096] In the formula, r = (r x r y r z ), (v r w r ) represents the projection address of the marker point r onto the detector, and u, v, and w are vectors representing the normal direction of the plane where the detector is located, the horizontal direction parallel to the row of detectors, and the vertical direction parallel to the column of detectors, respectively.

[0097] Step 5: Solve the optimization model to obtain the extrinsic parameters that vary with the angle, and use the estimated intrinsic and extrinsic parameters for reconstruction; wherein, the extrinsic parameters include the focal coordinates of the ray source (s). x s y s z The in-plane rotation angle α, pitch angle β, and tilt angle γ of the detector.

[0098] It should be noted that the method for solving this optimization model can be an existing method such as the Powell algorithm.

[0099] The reconstruction results are shown in Figures 5 and 6:

[0100] Figure 5 compares the phantom reconstruction image without geometric parameter calibration with the phantom reconstruction image calibrated using the method of the present invention. Figure 5a and Figure 5c This indicates a fault line that is being reconstructed. Figure 5b and Figure 5d This indicates that the reconstruction data is along the principal axis of the phantom ( Figure 1 The maximum density projection along the z-axis. It can be seen that... Figure 5a and 5b Both the hollow cylinder and the small steel ball showed severe deformation, making it difficult to glean any useful information from the image. Figure 5c and 5d The hollow cylinder and the outlines of the markers are clearly visible, and the markers are evenly arranged on the cylinder wall. Figure 5c Other artifacts, such as metallic artifacts and hardening artifacts, exist near the marker points, but geometric artifacts are significantly removed.

[0101] Figure 6a and 6b These are two tomographic images of trabecular bone reconstructed without geometric parameter calibration. Figure 6c and 6d This shows two tomographic images of the trabecular bone reconstructed using the two-step calibration method of this invention. (Observation) Figure 6a and 6b As can be seen, reconstructing the bone directly from the projection data without geometric parameter calibration resulted in significant bone deformation, and severe geometric artifacts blurred the internal structure of the bone, making it difficult to discern bone tissue. Figure 6c and 6d This allows for a clearer reconstruction of bone tissue, resulting in a significant improvement in image quality.

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Those skilled in the art should understand that modifications can be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for calibrating geometric parameters in a C-arm CT system, characterized in that, include: Step 1: Place the turntable in the middle of the C-frame and place the spiral mold vertically on the turntable. The mold rotates with the turntable while the C-arm remains stationary. This obtains the projection data of the mold under the circular trajectory cone beam scanning method. Place the stage in the middle of the C-frame and place the mold horizontally on the stage. Use the C-arm to scan while keeping the mold stationary to obtain the projection data of the mold under the C-arm scanning method. Step 2: Based on the projection data of the phantom obtained from the circular trajectory cone beam scanning in Step 1, the intrinsic parameters of the system that do not change with the angle are solved using the circular trajectory geometric parameter calibration method; among which, the intrinsic parameters include the vertical distance F from the focal point of the X-ray source to the detector, and the perpendicular coordinates (V) of the focal point of the X-ray source on the detector. S W S ); Step 3: Based on the projection data of the C-arm scanning phantom and the total number of marker points N in Step 1, determine the projection address of the i-th marker point r on the detector. Step 4: Establish the predicted projection address of the i-th marker point r under a single scanning angle in the phantom coordinate system. And extract the projection address of the i-th marker point r An optimized model for the mean square error; Step 5: Solve the optimization model to obtain the extrinsic parameters that vary with the angle, and use the estimated intrinsic and extrinsic parameters for reconstruction; wherein, the extrinsic parameters include the focal S coordinate of the ray source (s... x s y s z The in-plane rotation angle α, pitch angle β, and tilt angle γ of the detector.

2. The method for calibrating geometric parameters in a C-arm CT system according to claim 1, characterized in that, The phantom includes a cylindrical support (1). A predetermined number of marker points r are set in the wall (13) between the inner wall (11) and the outer wall (12) of the support (1) and are marked in sequence. The marker points r are arranged in a spiral from one end of the support (1) to the other end. The centers of the marker points r are connected in sequence to form a spiral. The central axis of the spiral in step 1 is basically consistent with the rotation axis of the turntable. The projected image is the projection image of X-rays penetrating the phantom on the flat panel detector.

3. The method for calibrating geometric parameters in a C-arm CT system according to claim 2, characterized in that, The marker point r is a spherical structure with a preset diameter.

4. The method for calibrating geometric parameters in a C-arm CT system according to any one of claims 1-3, characterized in that, Step 2 specifically includes: Step 21: By minimizing the mean square error between the projected address of the marker point at multiple angles and the extracted projected address, the optimization model for the geometric parameter calibration of the circular trajectory is established as follows (1): the vector of the detector in the horizontal direction is set as V = (V X V Y V Z ) T The vertical vector of the detector is set as W = (W X W Y W Z ) T V X V Y V Z W X W Y W Z The components of the detector's horizontal and vertical direction vectors in the circular trajectory coordinate system on the X, Y, and Z axes, respectively: Where, ξ k ξ 1 These are the geometric parameters of the circular trajectory cone-beam scanning system at the k-th and 1-th scanning angles, respectively. SDD is the intersection point O of the extension of the line from the focal point of the X-ray source to the rotation axis and the detector. S The distance between the detector and the focal point of the X-ray source, A, B, and Γ are the rotation angles of the detector around the three coordinate axes, respectively. O W O () is the intersection point O S The coordinates on the detector, M is the total number of scanning angles, E(ξ) 1 Let θ be the objective function. k This represents the angle value of the k-th scan angle. These are the projection addresses of the k-th scanning angle and the i-th marker point in the horizontal and vertical directions parallel to the detector row, respectively. Let be the predicted k-th scanning angle, the normal direction of the i-th marker point in the plane where the detector is located, the horizontal direction parallel to the detector row, and the vertical direction parallel to the detector column, respectively, and be described by the following formula (2). Let X, Y, and Z represent the coordinates of the i-th marker point in the circular trajectory coordinate system at the first scanning angle, respectively, and be described by the following formula (3): Step 22: Calculate the system intrinsic parameters using the following formula (4). The intrinsic parameters include the vertical distance F from the focal point of the X-ray source to the detector and the perpendicular coordinate (V) of the focal point of the X-ray source on the detector. S W S ): Where U = V × W, U represents the direction vector of the normal direction of the detector plane, U X This represents the X-axis component of U in the circular trajectory coordinate system.

5. The method for calibrating geometric parameters in a C-arm CT system according to any one of claims 1-3, characterized in that, The optimization model for the mean square error in step 4 is described by the following equation (5): Where η represents the extrinsic parameter of the system to be solved. The projected address of the i-th marker point r is calculated based on the estimated system extrinsic parameter η; The projected address of any marker point r in the phantom coordinate system is expressed as the following equation (6): In the formula, r = (r x r y r z ), (v r w r ) represents the projection address of the marker point r onto the detector, and U, V, and W are vectors representing the normal direction of the plane where the detector is located, the horizontal direction parallel to the row of detectors, and the vertical direction parallel to the column of detectors, respectively.

6. A calibration device for geometric parameters in a C-arm CT system, characterized in that, include: The system intrinsic parameter calculation unit is used to obtain the projection data of the helical phantom in circular trajectory cone-beam scanning mode by rotating the phantom with the turntable while keeping the C-arm stationary, when the helical phantom is placed vertically on the turntable in the middle of the C-frame. Then, the system intrinsic parameters that do not change with the angle are solved using the circular trajectory geometric parameter calibration method. Among them, the intrinsic parameters include the vertical distance F from the focal point of the X-ray source to the detector, and the perpendicular coordinates (V) of the focal point of the X-ray source on the detector. S W S ); The projection image acquisition unit is used to acquire the projection data of the model by scanning with a C-arm while keeping the model stationary on a platform in the middle of a C-frame. The optimization model building unit is used to determine the projection address of the i-th marker r on the detector based on the projection data of the C-arm scanning phantom and the total number of markers N. And establish the predicted projection address of the i-th marker point r under a single scanning angle in the phantom coordinate system. And extract the projection address of the i-th marker point r An optimization model for the mean square error, i = 1...N; The intrinsic and extrinsic parameter reconstruction unit is used to solve the optimization model, obtain the extrinsic parameters that vary with the angle, and use the estimated intrinsic and extrinsic parameters for reconstruction; wherein, the extrinsic parameters include the focal S-coordinate of the ray source (s... x s y s z The in-plane rotation angle α, pitch angle β, and tilt angle γ of the detector.

7. The calibration device for geometric parameters in a C-arm CT system according to claim 6, characterized in that, The phantom includes a cylindrical support (1). A predetermined number of marker points r are set in the wall (13) between the inner wall (11) and the outer wall (12) of the support (1) and are marked in sequence. The marker points r are arranged in a spiral from one end of the support (1) to the other end. The centers of the marker points r are connected in sequence to form a spiral. The central axis of the spiral in step 1 is basically consistent with the rotation axis of the turntable. The projected image is the projection image of X-rays penetrating the phantom on the flat panel detector.

8. The calibration device for geometric parameters in a C-arm CT system according to claim 7, characterized in that, The marker point r is a spherical structure with a preset diameter.

9. The calibration device for geometric parameters in a C-arm CT system according to any one of claims 6-8, characterized in that, The method of using circular trajectory geometric parameter calibration to solve for system intrinsic parameters that do not change with angle specifically includes: By minimizing the mean square error between the projected address of the marker point at multiple angles and the extracted projected address, the optimization model for the geometric parameter calibration of the circular trajectory is established as follows (1), and the vector of the detector in the horizontal direction is set as V = (V X V Y V Z ) T The vertical vector of the detector is set as W = (W X W Y W Z ) T V X V Y V Z W X W Y W Z The components of the detector's horizontal and vertical direction vectors in the circular trajectory coordinate system on the X, Y, and Z axes, respectively: Where, ξ k ξ 1 These are the geometric parameters of the circular trajectory cone-beam scanning system at the k-th and 1-th scanning angles, respectively. SDD is the intersection point O of the extension of the line from the focal point of the X-ray source to the rotation axis and the detector. S The distance between the detector and the focal point of the X-ray source, A, B, and Γ are the rotation angles of the detector around the three coordinate axes, respectively. O W O () is the intersection point O S The coordinates on the detector, M is the total number of scanning angles, E(ξ) 1 Let θ be the objective function. k This represents the angle value of the k-th scan angle. These represent the projected addresses of the k-th scanning angle, the normal direction of the i-th marker point in the plane of the detector, the horizontal direction parallel to the detector row, and the vertical direction parallel to the detector column, respectively. Let R be the predicted k-th scanning angle and the projected addresses of the i-th marker point R in the horizontal and vertical directions parallel to the detector row, respectively, as described by the following formula (2). Let X, Y, and Z represent the coordinates of the i-th marker point in the circular trajectory coordinate system at the first scanning angle, respectively, and be described by the following formula (3): The system intrinsic parameters are calculated using the following formula (4), which includes the vertical distance F from the focal point of the X-ray source to the detector and the perpendicular coordinate (V) of the focal point of the X-ray source on the detector. S W S ): Where U = V × W, U represents the direction vector of the normal direction of the detector plane, U X This represents the X-axis component of U in the circular trajectory coordinate system.

10. The calibration device for geometric parameters in a C-arm CT system according to any one of claims 6-8, characterized in that, The optimization model for a single scanning angle in the phantom coordinate system is established as follows (5): Where η represents the extrinsic parameter of the system to be solved. Let N be the projection address of the i-th marker point r calculated based on the system extrinsic parameter η, where N represents the total number of marker points and S represents the focal point of the ray source. The projected address of any marker point r in the phantom coordinate system is expressed as the following equation (6): In the formula, r = (r x r y r z ), (v r w r ) represents the projection address of the marker point r onto the detector, and U, V, and W are vectors representing the normal direction of the plane where the detector is located, the horizontal direction parallel to the row of detectors, and the vertical direction parallel to the column of detectors, respectively.

Images