CN108122203B - Geometric parameter correction method, device, equipment and system - Google Patents

Abstract

The present application provides a method, device, device and system for calibrating geometric parameters. The method includes: acquiring multiple two-dimensional projection images projected by a calibration phantom onto a detector at different gantry angles; The physical projection position coordinates of the center of the small ball in the 2D projection image; a set of global system geometric parameters are obtained by using the physical projection position coordinates; all 2D projection images are reconstructed according to the global system geometric parameters, and the 3D calibration phantom is obtained. Reconstruct the image; determine the virtual projection position coordinates of the center of the ball in the 3D reconstructed image on the acquired 2D projection image; calculate the residuals between the physical projection position coordinates and the virtual projection position coordinates; parameters are corrected. The present application uses the residual difference between the physical projection coordinates of the ball center and the virtual projection coordinates of the ball center to compensate the system geometric parameters of different gantry angles, and corrects the geometric parameter estimation errors.

Description

A method, device, equipment and system for calibrating geometric parameters

technical field

The present application relates to medical imaging technology, and in particular, to a method, device, device and system for calibrating geometric parameters.

Background technique

The Cone Beam Computed Tomography (CBCT) imaging system generally includes an X-ray radiation source, a detection plate and a rotating gantry. Among them, the X-ray radiation source and the detection plate are attached to the rotating gantry and rotate with the rotation of the rotating gantry. The principle of CBCT imaging is to rotate the gantry around the three-dimensional object to obtain the projection images of the object at multiple angles, and then apply the cone beam projection image reconstruction algorithm to reconstruct the three-dimensional image of the object.

When reconstructing an image, the CBCT imaging system needs to obtain accurate geometric system parameter information of the imaging system. The geometric parameters of the CBCT system refer to the geometric parameters that describe the angle and position between the X-ray light source, object and detector of the system, such as X -The distance from the ray radiation source to the detection plate, the distance from the X-ray radiation source to the rotation axis, the inclination and rotation angle of the detection plate, etc. Due to the gravity of the X-ray radiation source and the detection plate, these parameters change with the rotation of the rotating gantry, that is to say, the geometric system parameters of the imaging system are different under different rotating gantry angles. Inaccurate CBCT geometric system parameter information will cause geometric artifacts in reconstructed images and reduce the quality of reconstructed images. Therefore, the geometric parameter correction of CBCT system is of great significance to improve the quality of reconstructed images and prevent image artifacts.

At present, the common practice is to place a calibration phantom with several marked points between the X-ray radiation source and the detection plate, rotate the gantry to obtain 2D projection images of the calibration phantom at different angles, and then according to the 2D The geometric relationship between the marker points in the projection image and the geometric parameters of the imaging system is preset, and the geometric parameters of the imaging system are calculated. Among them, the specific implementation process is as follows:

The first solution is: the calibration phantom is a cylindrical phantom containing 24 metal spheres of the same size, these spheres form two rings coaxial with the cylindrical phantom, each ring consists of 12 evenly distributed It is composed of small balls, wherein the position error of the small balls is required to be no more than 25μm. After that, the calibration phantom is placed between the X-ray radiation source and the detection plate, so that the center of the calibration phantom is aligned with the isocenter, and the axis direction of the calibration phantom is aligned with the rotation axis of the rotating gantry. The rotating gantry was rotated at 1-degree intervals, and a 2D projection image of the calibration phantom was acquired at each gantry angle. For each projection image, do the following processing: first, determine the position coordinates of the projection of the ball in the projection image; then, in the defined projection coordinate system and detector coordinate system, use the position coordinates of the ball in the phantom coordinate system and The relationship between the position and coordinates of the ball on the projection map is given, and the analytical expression for directly calculating the geometric parameters of the system is given, and the geometric parameters of the system are calculated; finally, by fitting the coordinates of the X-ray radiation source at various angles into a circular trajectory , re-establish a projection system coordinate system, and recalculate the geometric parameters at all angles under this projection system coordinate system. However, in this scheme, the requirements for the relative positional accuracy of the small balls in the calibration phantom are relatively strict, and the relative positional deviation of the small balls affects the calculation accuracy of the final system geometric parameters.

The second scheme: It is usually assumed that the system geometric parameters of CBCT are a set of fixed global parameters at different gantry angles, and then the relationship between the position information of the marker points in the calibration phantom and the position information of its projected image is used to calculate this. Group global parameters. However, the assumptions of this method are usually not established in practice, and the reconstruction of 3D images with calculated global parameters will introduce geometric artifacts and reduce image quality.

Therefore, how to reduce the accuracy requirements of the correction phantom and accurately calculate the geometric parameters of the imaging system is a technical problem to be solved at present.

SUMMARY OF THE INVENTION

In view of this, the present application provides a method, device, device and system for calibrating geometric parameters, which can correct the estimation errors of geometric parameters under different gantry angles while reducing the accuracy requirements of the calibration phantom.

Specifically, the application is achieved through the following technical solutions:

A first aspect provides a method for calibrating geometric parameters, including:

Acquire multiple 2D projection images of the calibration phantom projected onto the detector at different gantry angles;

determining the physical projection position coordinates of the center of the ball in the two-dimensional projection image, and the ball is embedded in the calibration phantom;

Obtaining a set of global system geometric parameters using the physical projected position coordinates;

Reconstructing all two-dimensional projection images according to the global system geometric parameters to obtain a three-dimensional reconstructed image of the correction phantom;

determining the virtual projection position coordinates of the spherical center of the small ball on the acquired two-dimensional projection image in the three-dimensional reconstructed image;

obtaining the residual between the physical projection position coordinates and the virtual projection position coordinates;

The geometric parameters under different gantry angles are corrected by using the residuals.

A second aspect provides a device for calibrating geometric parameters, comprising:

an acquisition unit, used for acquiring multiple two-dimensional projection images of the correction phantom projected onto the detector under different gantry angles;

a first determining unit, configured to determine the physical projection position coordinates of the center of the small ball in the two-dimensional projection image, and the small ball is embedded in the calibration phantom;

a first obtaining unit, configured to obtain a set of global system geometric parameters by using the physical projection position coordinates;

a reconstruction unit, configured to reconstruct all two-dimensional projection images according to the global system geometric parameters to obtain a three-dimensional reconstructed image of the correction phantom;

a second determining unit, configured to determine the virtual projection position coordinates of the spherical center of the small ball in the acquired two-dimensional projection image in the three-dimensional reconstructed image;

a second obtaining unit, configured to obtain a residual between the physical projection position coordinates and the virtual projection position coordinates;

A correction unit, configured to use the residual to correct geometric parameters at different gantry angles.

A third aspect provides an image processing device, comprising: a processor; and a memory for storing instructions executable by the processor; wherein,

The processor is configured to:

acquiring multiple two-dimensional projection images of the calibration phantom projected onto the detector under different gantry angles;

determining the physical projection position coordinates of the center of the ball in the two-dimensional projection image, and the ball is embedded in the calibration phantom;

Obtaining a set of global system geometric parameters using the physical projected position coordinates;

Reconstructing all two-dimensional projection images according to the global system geometric parameters to obtain a three-dimensional reconstructed image of the correction phantom;

determining the virtual projection position coordinates of the spherical center of the small ball on the acquired two-dimensional projection image in the three-dimensional reconstructed image;

Obtain the residual of the physical projection position coordinates and the virtual projection position coordinates; use the residual to correct geometric parameters at different gantry angles.

A fourth aspect provides an imaging system, comprising: a rotating gantry, a radiation source and a detector attached to the rotating gantry, and further comprising:

A calibration phantom is used to be placed between the radiation source and the detector, wherein the calibration phantom includes a plurality of metal spheres, all the metal spheres approximately form a straight line, and the adjacent metal spheres are arranged between adjacent metal spheres. The interval is approximately equal;

an image processing device for acquiring a plurality of two-dimensional projection images projected by the correction phantom onto the detector at different gantry angles; determining the physical projection position coordinates of the center of the sphere in the two-dimensional projection images, The small ball is embedded in the correction phantom; a set of global system geometric parameters is obtained by using the physical projection position coordinates; all two-dimensional projection images are reconstructed according to the global system geometric parameters to obtain the correction phantom the three-dimensional reconstructed image; determine the virtual projection position coordinates of the ball center in the three-dimensional reconstructed image on the acquired two-dimensional projection image; obtain the residual between the physical projection position coordinates and the virtual projection position coordinates; The geometric parameters under different gantry angles are corrected by using the residuals.

In the embodiment of the present application, the two-dimensional projection images of the calibration phantom at different gantry angles are collected, and the position coordinates of the physical projection of the center of the small ball in the acquired two-dimensional projection image are determined, and then the small ball is used. The analytic relationship between the physical projection position coordinates of the heart and the geometric parameters of the CBCT imaging system calculates a set of global system geometric parameters. Then use this set of global geometric parameters to reconstruct all 2D projection images to obtain 3D reconstructed images of the corrected phantom; then, determine the position coordinates of the center of the ball in the 3D reconstructed image, and calculate the center of the ball in the 3D reconstructed image The virtual projection position coordinates projected to the acquired two-dimensional projection image under each acquisition rack angle. Finally, calculate the residual between the collected physical projection coordinates of the ball center and the calculated virtual projection coordinates of the ball center, and correct the geometric parameter estimation error at each gantry angle according to the residual. It can be seen that, in the embodiment of the present application, by analyzing the projection images of the calibration phantom at different gantry angles, and calculating a set of fixed global system geometric parameters, there is no need to mark the difference between the spheres in the calibration phantom. The precise relative position between the balls greatly facilitates the implementation of the algorithm and reduces the cost; at the same time, the residual difference between the collected physical projection coordinates of the ball and the calculated virtual projection coordinates of the ball is used to determine the difference between the different gantry angles. The system geometric parameters are compensated to correct the geometric parameter estimation errors caused by the assumption that the system geometric parameters are fixed.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not limiting of the present application.

Description of drawings

FIG. 1 is a schematic structural diagram of an imaging system provided by an embodiment of the present application;

2 is a schematic structural diagram of a calibration phantom provided by an embodiment of the present application;

3 is a flowchart of a method for calibrating geometric parameters provided by an embodiment of the present application;

4A is a schematic diagram of a geometric system parameter of a CBCT system provided by an embodiment of the present application;

FIG. 4B is another schematic diagram representing the geometric system parameters of the CBCT system provided by the embodiment of the present application;

5A is a schematic diagram of a circular motion trajectory of each small ball in the calibration phantom provided by this embodiment of the present application;

5B is a schematic diagram of the projection of the circular motion trajectory of each ball in the calibration phantom provided by the embodiment of the present application on the flat panel detector as an elliptical trajectory;

6 is a schematic diagram of a similar triangular relationship provided by an embodiment of the present application;

7 is a schematic diagram of the calculation principle of the distance from the ray source to the rotation axis provided by the embodiment of the present application;

Fig. 8 is a kind of hardware structure diagram of the equipment where the correction device of the geometric parameter provided by the application is located;

9 is a schematic structural diagram of a device for calibrating geometric parameters provided by an embodiment of the present application;

FIG. 10 is another schematic structural diagram of a device for calibrating geometric parameters provided by an embodiment of the present application.

Detailed ways

Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. Where the following description refers to the drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the illustrative examples below are not intended to represent all implementations consistent with this application. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present application as recited in the appended claims.

The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to limit the application. As used in this application and the appended claims, the singular forms "a," "the," and "the" are intended to include the plural forms as well, unless the context clearly dictates otherwise. It will also be understood that the term "and/or" as used herein refers to and includes any and all possible combinations of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in this application to describe various information, such information should not be limited by these terms. These terms are only used to distinguish the same type of information from each other. For example, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information without departing from the scope of the present application. Depending on the context, the word "if" as used herein can be interpreted as "at the time of" or "when" or "in response to determining."

For ease of understanding, before describing the embodiments of the present application, the application architecture involved in the present application is briefly introduced.

Please refer to FIG. 1 , which is a schematic structural diagram of an imaging system provided by an embodiment of the present application; the imaging system, for example, a cone beam computer reconstruction tomography (CBCT, Come beam computer tomography) imaging system, the imaging system Generally, it includes: an image processing device 10, a rotating gantry 11, a radiation source (such as an X-ray radiation source) 12 attached to the rotating gantry 11, a detector 13 and a calibration phantom 14, and the rotating gantry 11 can surround rotates about the axis of rotation 15; wherein,

The calibration phantom 14 is used to be placed between the radiation source 12 and the detector 13, wherein the calibration phantom 14 contains several metal spheres, all the metal spheres approximately form a straight line, and the interval between adjacent metal spheres can be equal or approximately equal.

It should be noted that, when placing the calibration phantom 14, try to make the calibration phantom 14 and the rotation axis of the rotating frame 11 parallel but not on the same axis to ensure that all projections of the metal balls in the calibration phantom 14 are on the detector. 13 on.

Wherein, a calibration phantom provided in this embodiment may be a stick-shaped object with a length of about 28 cm and a diameter of 6 mm. In the calibration phantom, 21 metal balls are embedded as an example, and the diameter of each metal ball is The distance between the 3mm metal ball and the adjacent metal balls can be 10mm, or approximately equal to 10mm, of course, it can be adaptively adjusted as needed. That is to say, the spacing requirements of adjacent metal balls do not have to be particularly precise. Specifically, as shown in FIG. 2, FIG. 2 is a schematic structural diagram of a calibration phantom provided by an embodiment of the present application. In FIG. 2, the distance between adjacent metal balls is taken as an example, and the specific application is not limited to this.

The image processing device 10 is connected to the detector and the calibration phantom respectively, and can be applied to various computers to obtain multiple two-dimensional images of the calibration phantom 14 projected onto the detector at different gantry angles. projection image; determine the physical projection position coordinates of the center of the ball in the two-dimensional projection image, the ball is embedded in the calibration phantom 14; use the physical projection position coordinates to obtain a set of global system geometric parameters Reconstructing all two-dimensional projection images according to the global system geometric parameters to obtain a three-dimensional reconstructed image of the correction phantom; determining the position of the spherical center of the sphere on the acquired two-dimensional projection image in the three-dimensional reconstructed image virtual projection position coordinates; obtain residuals between the physical projection position coordinates and the virtual projection position coordinates; use the residuals to correct geometric parameters at different gantry angles.

The process of image processing by the image processing device 10 can be seen in the embodiments corresponding to FIG. 3 to FIG. 8 below, which will not be repeated here.

Please refer to FIG. 3 , which is a flowchart of a method for calibrating geometric parameters provided by an embodiment of the present application. The method includes:

Step 301: Acquire multiple two-dimensional projection images of the calibration phantom projected onto the detector at different gantry angles;

The calibration phantom in this step is shown in FIG. 2 , which is not repeated here.

In this step, the correction phantom needs to be placed between the radiation source and the detector, so that the imaging system can acquire two-dimensional (ie 2D) projection images projected by the correction phantom onto the detector at different angles. It should be noted that when placing the calibration phantom, try to make the calibration phantom and the rotation axis of the rotating frame parallel but not on the axis, and ensure that the metal ball (hereinafter collectively referred to as the ball) in the calibration phantom is projected. All on the detector.

In this step, the imaging system rotates and scans a circle, and then the two-dimensional projection images of the correction phantom projected onto the detector under different gantry angles can be obtained. The specific process is well known to those skilled in the art. It is not repeated here.

Step 302: Determine the physical projection position coordinates of the center of the small ball in the two-dimensional projection image, and the small ball is embedded in the calibration phantom;

表示,m＝1,2,…M,n＝1,2,…,N，其中M是获取的二维投影图像的总张数，N是小球的个数。Among them, the physical projection position coordinates are used

Represents, m=1,2,...M,n=1,2,...,N, where M is the total number of acquired two-dimensional projection images, and N is the number of small balls.

In this step, a determination method is as follows:

First, Gaussian filtering is performed on the two-dimensional projection image to remove random noise to obtain a filtered image; the specific implementation process thereof is a well-known technology for those skilled in the art, and will not be repeated here.

Secondly, calculate the outline of the projection of the ball according to the filtered image;

In the specific implementation, the watershed segmentation algorithm can be used for the filtered image to initially calculate the outline of the projection of the small ball, and fill the outline with the pixel values of the background around the small ball to obtain a background image;

Then, subtract the background image from the filtered image to obtain a background-free image;

Finally, the Canny edge detection algorithm is used to detect the edge of the background-free image, and the edge of the ball projection is obtained, and the edge is fitted into a circle, and the center of the circle is the physical projection position coordinate of the ball.

Another determination method is: firstly use the trained convolutional neural network to segment the two-dimensional projection image, and then calculate the physical projection position coordinates of the center of the small ball in the segmented image. For the skilled person, the technology is already well-known and will not be repeated here.

Step 303: using the physical projection position coordinates to calculate a set of global system geometric parameters;

Among them, a set of global system geometric parameters can be represented by R, D, (u ₀ , v ₀ ), n, as shown in FIG. 4A and FIG. 4B , and FIG. 4A and FIG. 4B are representations of the CBCT system provided by this embodiment, respectively. A schematic diagram of the parameters of the geometric system;

As shown in Figure 4A, R is the vertical distance from the radiation source to the z-axis of the rotation axis; D is the distance from the z-axis of the vertical rotation axis of the radiation source to the detector; (u ₀ , v ₀ ) is the vertical rotation axis of the radiation source z-axis at The coordinates of the vertical projection point on the detector; as shown in Figure 4B, η is the rotation angle of the detector around the vertical projection point (u ₀ , v ₀ ).

Since the motion is mutual, the radiation source and the detection plate rotate around the axis of rotation once while the calibration phantom is stationary. It can be considered that the radiation source and the detection plate are stationary and the calibration phantom rotates around the axis of rotation once. All the analysis below All are carried out under this motion model. At this time, the motion of each ball in the calibration phantom is a circular motion trajectory, and the projection of the circular motion trajectory of each small ball on the flat panel detector is an elliptical trajectory, as shown in Figures 5A and 5B Show. 5A is a schematic diagram of a circular motion trajectory of each ball in the calibration phantom provided by this embodiment; The projection on the device is a schematic diagram of an elliptical trajectory.

The realization process of calculating a set of global system geometric parameters R, D, u ₀ , v ₀ , η is provided below.

In Fig. 5A and Fig. 5B, the intersections of the circular trajectory of the ball i with the x-axis and the y-axis are θ _i1 , θ _i2 , θ _i3 and θ _i4 , respectively, and their elliptical trajectories projected on the detection board are respectively A _i1 (u _i1 ,v _i1 ), A _i2 (u _i2 ,v _i2 ), A _i3 (u _i3 ,v _i3 ) and A _i4 (u _i4 ,v _i4 ), A _i1 , A _i2 , A _i3 and A _i4 respectively are the extreme points of the short axis and the extreme point of the long axis of the elliptical trajectory.

Then, combined with the schematic diagram of the similar triangular relationship shown in Figure 6, the following relationship can be obtained:

Among them, A ₀ is the projection point of the z-axis of the vertical rotation axis of the radiation source on the detector.

Adding and subtracting the above two formulas can be obtained:

Then there is:

In addition, from Figure 6, the following formula can also be clearly drawn:

So you can get:

It can be seen from the above formula that v ₀ has a linear relationship with D, and the parameters of this straight line can be obtained by calculating the extreme points of N small balls. The N following value pairs (Y _i ,X _i ) can be obtained,

Then, a straight line X=v ₀ +D·Y can be used to fit the above-mentioned pair of values (Y _i , X _i ), so that the geometric parameters v ₀ and D can be obtained.

Next, to calculate u ₀ and η, first calculate the center coordinates of the elliptical trajectory formed by each ball in the projection image. The center coordinates of the elliptical trajectory can be obtained from the abscissa and ordinate of the above four extreme points. The specific formula is as follows:

Then, u=k+tanη·v can be used to fit the center point (u _i0 , v _i0 ) of the ellipse trajectory obtained above. When k and tanη are obtained, u ₀ can be obtained by substituting the previously obtained v ₀ into u=k+tanη·v, and η can be obtained from the value of tan η.

A schematic diagram of the calculation principle of the distance R from the ray source to the rotation axis is shown in Fig. 7 . Among them, the subscripts 1 and 2 respectively represent two small balls in the calibration phantom, l is the known distance between the two small balls, h is the vertical distance between the circular trajectories of the two small balls, r ₁ and r ₂ is the radius of the circular trajectory of ball 1 and ball 2, respectively. So there is

Among them, α ₁ =θ ₁₀ -θ ₁₄ , α ₂ =θ ₂₀ -θ ₂₄ , θ ₁₀ and θ ₂₀ are the initial position angles of balls 1 and 2, and θ ₁₄ and θ ₂₄ are one pole of balls 1 and 2 value points.

From the previous steps, the following relation is known

Using the above proportional relationship and the known distance l between two small balls, R can be obtained. The specific formula is as follows:

So far, a set of global system geometric parameters R, D, u ₀ , v ₀ , η have been calculated.

Step 304: Reconstruct all 2D projection images according to the global system geometric parameters to obtain a 3D reconstructed image of the correction phantom;

In this step, according to the calculated set of global system geometric parameters, the FDK algorithm is used to reconstruct all two-dimensional projection images to obtain a 3D reconstructed image of the correction phantom. The process of reconstructing all two-dimensional projection images by using the FDK algorithm is a well-known technology for those skilled in the art, and details are not described herein again.

Step 305: Determine the virtual projection position coordinates of the center of the small ball in the 3D reconstructed image on the acquired 2D projection image;

A determination method is: first determine the three-dimensional position coordinates of the center of the small ball in the three-dimensional reconstructed image; the specific process is:

First, perform threshold filtering on the three-dimensional reconstructed image, and set the three-dimensional pixel points whose pixel value is lower than the predetermined threshold to 0;

Secondly, for each pixel (x ⁰ , y ⁰ , z ⁰ ) in the 3D reconstructed image, where 0 in x ⁰ represents the pixel before iterative reconstruction, if the values of each pixel and its adjacent 8 pixels are both If it is not 0, then iteratively calculate the center of gravity in the neighborhood of each pixel (x ⁰ ±k, y ⁰ ±k, z ⁰ ±k), where k is the size of the neighborhood range and is greater than the radius of the sphere, as follows shown in the formula

i=1,2,...,l

Among them, ⁱ in x ⁱ , y ⁱ , and zi represents the pixel of the i-th iteration. I(x, y, z) is the pixel value at (x, y, z) of the 3D reconstructed image after threshold filtering, and l is the number of iterations. And mark the three-dimensional position coordinates of the center of a small ball in the finally calculated three-dimensional reconstructed image as (x _n , y _n , z _n ), n=1,2,...,N, where N is the number of small balls number, this example can take 21 as an example.

Then, according to the global system geometric parameters R, D, u ₀ , v ₀ , η , forward projection is used to calculate the virtual projection position coordinates of the centers of several spheres in the 3D reconstructed image in the acquired 2D projection image.

表示,m＝1,2,…M,n＝1,2,…,N，其中M是获取的二维投影图像的总张数，N是小球的个数。Among them, the centers of several small balls are represented by (x _n , y _n , z _n ), n=1,2,...,N; the virtual projection position coordinates are represented by

Represents, m=1,2,...M,n=1,2,...,N, where M is the total number of acquired two-dimensional projection images, and N is the number of small balls.

Step 306: Calculate the residual difference between the physical projection position coordinates of the ball center in the two-dimensional projection image and the virtual projection position coordinates of the ball center in the three-dimensional reconstructed image in the two-dimensional projection image;

表示，残差用

表示，如下式所示Among them, the physical projection position coordinates are used

means that the residual is

represented, as shown in the following formula

The main reason for the generation of residuals is that in step 303 it is assumed that the system geometry is stationary during the rotation of the gantry, while in practice the system geometry changes as the gantry rotates, for example due to gravity The reason for this is that the position of the detector changes with the rotation of the gantry. From another perspective, the geometric model of the CBCT imaging system in step 303 is a simple approximate model of the actual imaging system, that is, the geometric parameters in this embodiment are only an approximate description of the geometric model of the system. Therefore, the residual is a systematic error that can be described by building a model.

Step 307: Use the residual to correct geometric parameters at different gantry angles.

In this step, curve fitting is first performed on the residual to obtain the fitted relationship curve between the residual and the gantry angle; Compensation is performed to correct the errors of geometric parameters under different frame angles.

The following is an example of calibrating the vertical projection point coordinates (u ₀ , v ₀ ) of the vertical rotation axis of the radiation source on the detector.

The residual is a systematic error that is a sine function of the gantry angle. The residuals satisfy the following formula

Δu(θ)=a _u cos(b _u θ+c _u )+d _u

Δv(θ)=a _v cos(b _v θ+c _v )+d _v

The relationship between the residuals and the gantry angle can be obtained by comparing

n=1,2,...,N is obtained by curve fitting. Use the above relationship to correct the vertical projection point coordinates (u ₀ , v ₀ ), and then obtain the vertical projection point coordinates of the random frame angle change:

u ₀ (θ)=u ₀ +Δu(θ)

v ₀ (θ)=v ₀ +Δv(θ)

The vertical projection point coordinates of the random gantry angle change are a correction to the static system geometric model in step 303, which compensates for the deviation of the system geometric parameters from the global geometric parameters during the gantry rotation process.

In the embodiment of the present application, two-dimensional projection images of the calibration phantom at different gantry angles are collected, and the physical projection position coordinates of the center of the small ball in the collected two-dimensional projection image are determined, and then the The physical projection position coordinates directly obtain a set of global system geometric parameters under a static simple geometric model. Then use this set of global geometric parameters to reconstruct all 2D projection images to obtain 3D reconstructed images of the corrected phantom; then, determine the position coordinates of the center of the sphere in the 3D reconstructed image, and calculate the sphere in the 3D reconstructed image. The virtual projection position coordinates of the center on the 2D projection image collected at each gantry angle. Finally, calculate the residuals between the physical projection position coordinates of the ball center in the collected 2D image and the virtual projection position coordinates of the ball center in the 3D reconstructed image on the 2D projection images collected at each gantry angle, and The geometric parameters at each gantry angle are corrected according to the residual. It can be seen that, in the embodiment of the present application, a set of fixed and invariable global system geometric parameters is calculated by using the projected images of the calibration phantom at different gantry angles using the analytical method, and there is no need to mark the space between the small balls in the calibration phantom. Accurate relative position, which greatly facilitates the implementation of the algorithm and reduces the cost; at the same time, the residual difference between the collected physical projection coordinates of the ball and the calculated virtual projection coordinates of the ball is used for the system of different gantry angles. The geometric parameters are compensated to correct the geometric parameter estimation errors caused by the assumption that the system geometric parameters are fixed.

Corresponding to the foregoing embodiments of the geometric parameter calibration method, the present application further provides an embodiment of a geometric parameter calibration device.

The embodiments of the geometric parameter correction device provided in this application can be applied to equipment. The apparatus embodiment may be implemented by software, or may be implemented by hardware or a combination of software and hardware. Taking software implementation as an example, a device in a logical sense is formed by reading the corresponding computer program instructions in the non-volatile memory into the memory for operation by the processor of the device where it is located. From the perspective of hardware, as shown in FIG. 8 , it is a hardware structure diagram of the device where the geometric parameter correction device provided by the present application is located, except for the processor, memory, network interface, and non-volatile shown in FIG. 8 . In addition to the memory, the device in which the apparatus in the embodiment is located generally may also include other hardware according to the actual function of the device, and details are not described herein again.

Please refer to FIG. 9 , which is a schematic structural diagram of an apparatus for calibrating geometric parameters provided by an embodiment of the present application. The apparatus includes: an acquisition unit 91 , a first determination unit 92 , a first acquisition unit 93 , a reconstruction unit 94 , a second Determining unit 95, second obtaining unit 96 and correcting unit 97, wherein,

an acquisition unit 91, configured to acquire a plurality of two-dimensional projection images projected by the correction phantom onto the detector under different gantry angles;

a first determining unit 92, configured to determine the physical projection position coordinates of the center of the ball in the two-dimensional projection image, and the ball is embedded in the calibration phantom;

a first obtaining unit 93, configured to calculate a set of global system geometric parameters by using the physical projection position coordinates;

A reconstruction unit 94, configured to reconstruct all the two-dimensional projection images according to the global system geometric parameters to obtain a three-dimensional reconstructed image of the correction phantom;

The second determining unit 95 is configured to determine the virtual projection position coordinates on the two-dimensional projection images of the spherical centers of the small balls collected at various gantry angles in the three-dimensional reconstructed image;

a second obtaining unit 96, configured to calculate the residual between the physical projection position coordinates and the virtual projection position coordinates;

The correction unit 97 is used for correcting and compensating geometric parameters under different gantry angles by using the residuals.

Optionally, in another embodiment, based on the foregoing embodiment, the first determining unit 91 includes: a filtering unit 911 , a first calculating unit 912 , a filling unit 913 , and a second calculating unit 914 , the edge detection unit 915 and the first fitting unit 916 , the schematic diagram of which is shown in FIG. 10 . in,

A filtering unit 911, configured to perform Gaussian filtering on the two-dimensional projection image to obtain a filtered image;

a first calculation unit 912, configured to calculate the outline of the projection of the ball according to the filtered image;

Filling unit 913, used to fill the outline with the pixel value of the background around the ball to obtain a background image;

The second computing unit 914 is used for subtracting the background image from the filtered image to obtain a background-free image;

Edge detection 915, for performing edge detection on the background-free image to obtain the edge of the projection of the small ball;

The first fitting unit 916 is configured to fit the edge into a circle, and the center of the circle is the physical projection position coordinate of the small ball.

Optionally, in another embodiment, on the basis of the foregoing embodiment, the first determining unit 91 includes: a dividing unit and a position coordinate calculating unit (not shown in the figure), wherein,

a segmentation unit for segmenting the two-dimensional projection image by using the trained convolutional neural network to obtain a ball of the two-dimensional projection image;

The position coordinate calculation unit is used to calculate the physical projection position coordinates of the center of the small ball in the segmented image.

Optionally, in another embodiment, based on the foregoing embodiment, the first obtaining unit 93 is specifically configured to use the physical projection position coordinates in a static state during the rotation of the gantry. A set of global system geometric parameters are analytically calculated under the system geometric model of The coordinate of the vertical projection point of the axis on the detector and the rotation angle of the detector around the vertical projection axis.

Optionally, in another embodiment, based on the foregoing embodiment, the second determination unit 95 includes: a determination subunit and a projection unit (not shown in the figure), wherein,

A determination subunit, used to determine the three-dimensional position coordinates of the center of the small ball in the three-dimensional reconstructed image;

The projection unit is configured to calculate, according to the three-dimensional position coordinates, the virtual projection position coordinates on the two-dimensional projection images of the spherical centers of the small balls in the three-dimensional reconstructed image collected at various gantry angles.

Optionally, in another embodiment, based on the foregoing embodiment, the second obtaining unit 96 includes: a second fitting unit and a compensation unit (not shown in the figure), wherein,

a second fitting unit, configured to perform curve fitting on the residual and different gantry angles to obtain the fitted relationship curve between the residual and the gantry angle;

The compensation unit is used for compensating the geometric parameters under different gantry angles by using the fitted relation curve, and correcting the errors of the geometric parameters under different gantry angles.

For details of the implementation process of the functions and functions of each unit in the above device, please refer to the implementation process of the corresponding steps in the above method, which will not be repeated here.

For the apparatus embodiments, since they basically correspond to the method embodiments, reference may be made to the partial descriptions of the method embodiments for related parts. The device embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the present application. Those of ordinary skill in the art can understand and implement it without creative effort.

Embodiments of the present application further provide an image processing device, including: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to:

acquiring multiple two-dimensional projection images of the calibration phantom projected onto the detector under different gantry angles;

determining the physical projection position coordinates of the center of the ball in the two-dimensional projection image, and the ball is embedded in the calibration phantom;

Obtaining a set of global system geometric parameters using the physical projected position coordinates;

Reconstructing all two-dimensional projection images according to the global system geometric parameters to obtain a three-dimensional reconstructed image of the correction phantom;

determining the virtual projection position coordinates of the spherical center of the small ball on the acquired two-dimensional projection image in the three-dimensional reconstructed image;

Obtain the residual of the physical projection position coordinates and the virtual projection position coordinates; use the residual to correct geometric parameters at different gantry angles.

In this embodiment, the implementation process of each step executed by the processor is detailed in the above, which is not repeated here.

In the embodiment of the present application, by analyzing the projection images of the calibration phantom at different gantry angles, and calculating a set of fixed and invariable global system geometric parameters, it is not necessary to mark the exact relative relationship between the spheres in the calibration phantom. position, which greatly facilitates the implementation of the algorithm and reduces the cost; at the same time, the system geometric parameters of different gantry angles are calculated by using the residual between the collected physical projection coordinates of the ball and the calculated virtual projection coordinates of the ball. Compensation, which corrects the geometric parameter estimation error caused by the assumption that the system geometric parameters are fixed.

Those skilled in the art can clearly understand that the technology in the embodiments of the present invention can be implemented by means of software plus a necessary general hardware platform. Based on such understanding, the technical solutions in the embodiments of the present invention may be embodied in the form of software products in essence or the parts that make contributions to the prior art, and the computer software products may be stored in a storage medium, such as ROM/RAM , magnetic disk, optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

The above descriptions are only preferred embodiments of the present application, and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the present application. within the scope of protection.

Claims (15)

1. a correction method of geometric parameter, is characterized in that, comprises: Acquire multiple 2D projection images of the calibration phantom projected onto the detector at different gantry angles; determining the physical projection position coordinates of the center of the ball in the two-dimensional projection image, and the ball is embedded in the calibration phantom; Obtaining a set of global system geometric parameters using the physical projected position coordinates; Reconstructing all two-dimensional projection images according to the global system geometric parameters to obtain a three-dimensional reconstructed image of the correction phantom; determining the virtual projection position coordinates of the spherical center of the small ball on the acquired two-dimensional projection image in the three-dimensional reconstructed image; obtaining the residual between the physical projection position coordinates and the virtual projection position coordinates; The geometric parameters under different gantry angles are corrected by using the residuals. 2. The method according to claim 1, wherein the determining the physical projection position coordinates of the center of the small ball in the two-dimensional projection image comprises: Gaussian filtering is performed on the two-dimensional projection image to obtain a filtered image; Calculate the outline of the projection of the ball according to the filtered image; Fill the outline with the pixel value of the background around the ball to obtain a background image; subtracting the background image from the filtered image to obtain a background-free image; Perform edge detection on the background-free image to obtain the edge of the ball projection; The edge is fitted into a circle, and the center of the circle is the coordinates of the physical projection position of the center of the small ball. 3. The method according to claim 1, wherein the determining the physical projection position coordinates of the center of the small ball in the two-dimensional projection image comprises: Use the trained convolutional neural network to perform image segmentation on the two-dimensional projection image; Calculate the physical projection position coordinates of the center of the small ball in the segmented image. 4. The method according to any one of claims 1 to 3, wherein obtaining a set of global system geometric parameters by using the physical projection position coordinates comprises: During the rotation of the gantry, a set of global system geometric parameters is analytically calculated under the static system geometric model by using the physical projection position coordinates. The global system geometric parameters include: the vertical distance from the radiation source to the rotation axis, the radiation source The distance from the vertical rotation axis of the source to the detector, the coordinates of the vertical projection point of the vertical rotation axis of the radiation source on the detector, and the rotation angle of the detector around the vertical projection axis. 5 . The method according to claim 4 , wherein the determining the virtual projection position coordinates of the center of the small ball in the three-dimensional reconstructed image on the acquired two-dimensional projection image comprises: 5 . Determine the three-dimensional position coordinates of the center of the small ball in the three-dimensional reconstructed image; The virtual projection position coordinates of the center of the sphere in the three-dimensional reconstructed image in the acquired two-dimensional projection image are calculated according to the global system geometric parameters. 6. The method according to claim 5, wherein using the residual to correct geometric parameters at different gantry angles comprises: Perform curve fitting on the residual to obtain the fitted relationship between the residual and the gantry angle; Using the fitted relationship curve, the geometric parameters under different gantry angles are compensated, and the errors of the geometric parameters under different gantry angles are corrected. 7. A correction device for geometric parameters, characterized in that, comprising: an acquisition unit, used for acquiring multiple two-dimensional projection images of the correction phantom projected onto the detector under different gantry angles; a first determining unit, configured to determine the physical projection position coordinates of the center of the small ball in the two-dimensional projection image, and the small ball is embedded in the calibration phantom; a first obtaining unit, configured to obtain a set of global system geometric parameters by using the physical projection position coordinates; a reconstruction unit, configured to reconstruct all two-dimensional projection images according to the global system geometric parameters to obtain a three-dimensional reconstructed image of the correction phantom; a second determining unit, configured to determine the virtual projection position coordinates of the spherical center of the small ball in the acquired two-dimensional projection image in the three-dimensional reconstructed image; a second obtaining unit, configured to obtain a residual between the physical projection position coordinates and the virtual projection position coordinates; A correction unit, configured to use the residual to correct geometric parameters at different gantry angles. 8. The apparatus according to claim 7, wherein the first determining unit comprises: a filtering unit, configured to perform Gaussian filtering on the two-dimensional projection image to obtain a filtered image; a first calculation unit, configured to calculate the outline of the projection of the ball according to the filtered image; a filling unit, used for filling the outline with the pixel value of the background around the ball to obtain a background image; a second computing unit for subtracting the background image from the filtered image to obtain a background-free image; an edge detection unit for performing edge detection on the background-free image to obtain the edge of the projection of the small ball; The first fitting unit is used for fitting the edge into a circle, and the center of the circle is the physical projection position coordinate of the center of the small ball. 9. The apparatus according to claim 7, wherein the first determining unit comprises: a segmentation unit for segmenting the two-dimensional projection image by using a trained convolutional neural network; The position coordinate calculation unit is used to calculate the physical projection position coordinates of the center of the small ball in the segmented image. 10. The device according to any one of claims 7 to 9, characterized in that The first obtaining unit is specifically configured to use the physical projection position coordinates to analytically calculate a set of global system geometric parameters under a static system geometric model during the rotation of the gantry, wherein the global system geometric parameters Including: the vertical distance from the radiation source to the rotation axis, the distance from the vertical rotation axis of the radiation source to the detector, the coordinates of the vertical projection point of the vertical rotation axis of the radiation source on the detector and the rotation angle of the detector around the vertical projection axis. 11. The apparatus according to claim 10, wherein the second determining unit comprises: A determination subunit, used to determine the three-dimensional position coordinates of the center of the small ball in the three-dimensional reconstructed image; The projection unit is configured to calculate, according to the three-dimensional position coordinates, the virtual projected position coordinates of the spherical center of the small ball in the acquired two-dimensional projected image in the three-dimensional reconstructed image. 12. The apparatus according to claim 11, wherein the second obtaining unit comprises: a second fitting unit, configured to perform curve fitting on the residual to obtain a relationship curve between the fitted residual and the gantry angle; The compensation unit is used for compensating the geometric parameters under different gantry angles by using the fitted relation curve, and correcting the errors of the geometric parameters under different gantry angles. 13. An image processing device, comprising: a processor; and a memory for storing executable instructions of the processor; wherein, The processor is configured to: Acquire multiple 2D projection images of the calibration phantom projected onto the detector at different gantry angles; determining the physical projection position coordinates of the center of the ball in the two-dimensional projection image, and the ball is embedded in the calibration phantom; Obtaining a set of global system geometric parameters using the physical projected position coordinates; Reconstructing all two-dimensional projection images according to the global system geometric parameters to obtain a three-dimensional reconstructed image of the correction phantom; determining the virtual projection position coordinates of the spherical center of the small ball on the acquired two-dimensional projection image in the three-dimensional reconstructed image; Obtain the residual of the physical projection position coordinates and the virtual projection position coordinates; use the residual to correct geometric parameters at different gantry angles. 14. An imaging system, comprising: a rotating gantry, a radiation source and a detector attached to the rotating gantry, further comprising: A calibration phantom is used to be placed between the radiation source and the detector, wherein the calibration phantom includes a plurality of metal spheres, all the metal spheres approximately form a straight line, and the adjacent metal spheres are arranged between adjacent metal spheres. The interval is approximately equal; an image processing device for acquiring a plurality of two-dimensional projection images projected by the correction phantom onto the detector at different gantry angles; determining the physical projection position coordinates of the center of the sphere in the two-dimensional projection images, The small ball is embedded in the correction phantom; a set of global system geometric parameters is obtained by using the physical projection position coordinates; all two-dimensional projection images are reconstructed according to the global system geometric parameters to obtain the correction phantom the three-dimensional reconstructed image; determine the virtual projection position coordinates of the ball center in the three-dimensional reconstructed image on the acquired two-dimensional projection image; obtain the residual between the physical projection position coordinates and the virtual projection position coordinates; The geometric parameters under different gantry angles are corrected by using the residuals. 15. The system of claim 14, wherein the calibration phantom is parallel to and not coaxial with the rotation axis of the rotating gantry.

Images