Automatic control method of medical linear accelerator and control system thereof

By acquiring images of medical linear accelerator labels and using depth coordinates, the position of the medical linear accelerator and the treatment bed is automatically calibrated, solving the problems of long manual calibration time and large errors, and improving treatment accuracy.

CN118925090AInactive Publication Date: 2024-11-12FUJIAN MEDICAL UNIV UNION HOSPITAL
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411050614.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2024-07-24
Filing Date
2024-08-01
Publication Date
2024-11-12
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The error between the treatment bed and the accelerator in existing medical linear accelerators affects the accuracy of treatment, and manual calibration is time-consuming and prone to large errors.

Method used

By acquiring and preprocessing the label image, the coordinate information in the XY and Z directions is obtained using the depth coordinate method, enabling automated calibration of the medical linear accelerator. This includes image channel processing, coordinate calculation and coordinate transformation, and using at least two cameras to determine the position of spatial points.

Benefits of technology

Automatic calibration between the medical linear accelerator and the treatment bed has been achieved, improving the accuracy of treatment and reducing the time and error of manual calibration.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118925090A_ABST
    Figure CN118925090A_ABST
Patent Text Reader

Abstract

The invention provides an automatic control method and system for a medical linear accelerator, and the method comprises the steps: 1, carrying out the image collection and preprocessing of a label image on a matched treatment bed of the medical linear accelerator, and obtaining the coordinate information of an X direction and a Y direction; step 2, based on the coordinate information of the X and Y directions in the step 1, using a depth coordinate method to obtain a coordinate of a Z direction; and step 3, realizing calibration of automatic control of the medical linear accelerator based on the identified label graph coordinates. The calibration device is used for achieving calibration of automatic control of the medical linear accelerator.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of medical equipment, and in particular relates to an automatic control method and a control system of a medical linear accelerator. Background Art

[0002] A medical accelerator is a particle accelerator device used in biomedicine for radiotherapy of tumors. A charged particle accelerator is an electromagnetic device that uses artificial methods to accelerate various types of charged particles to higher energy with the help of electric fields of different forms. It is often called a "particle accelerator" or "accelerator" for short. In order for charged particles to gain energy, there must be an accelerating electric field. Accelerators are divided into various types based on the different types of accelerated particles, the different forms of the accelerating electric field, and the different trajectories followed by the particle acceleration process. Internationally, the most commonly used in radiotherapy is the electron linear accelerator.

[0003] Since the medical linear accelerator is a precision instrument, when used for treatment, the error between the treatment bed and the medical linear accelerator will affect the accuracy of subsequent radiation to the target; quality control in the existing technology relies on manual measurement, but manual measurement has errors, which will affect the subsequent use results, and the time taken for manual measurement is longer than automatic calibration measurement. Summary of the invention

[0004] The present invention provides an automated control method and a control system for a medical linear accelerator, which are used to solve the problems mentioned in the above background technology.

[0005] The present invention is achieved through the following technical solutions:

[0006] An automated control method for a medical linear accelerator, the automated control method comprising the following steps:

[0007] Step 1: Capture and preprocess the label image on the medical linear accelerator treatment bed to obtain the coordinate information in the XY direction;

[0008] Step 2: Based on the coordinate information in the XY direction in step 1, the coordinate in the Z direction is obtained using the depth coordinate method;

[0009] Step 3: Based on the identified label map coordinates, the calibration of the automatic control of the medical linear accelerator is realized.

[0010] Furthermore, the step 1 is specifically that the image preprocessing includes the following steps:

[0011] Step 1.1: Process the R, G, and B channels of the label image separately and calculate the R(k) value of each pixel. The formula is as follows:

[0012]

[0013] Among them, R(k) is the intermediate result when calculating the parameters, I(k) and I(l) represent the brightness of two different pixels, d(k, l) represents the distance between the two pixels, Sα(t) is the slope function, α is the adjustable slope, and the formula is as follows:

[0014]

[0015] Step 1.2: Use the following formula to expand R(k) to [0, 1] to obtain the R, G, and B channels of the enhanced label image and obtain the global white balance. The formula is as follows:

[0016]

[0017] Among them, L(k) is the normalized median value;

[0018] Step 1.3: Solve the optimization problem and obtain the processed label image. The formula is as follows:

[0019]

[0020] Among them, ω represents the weight, A is the relative adjustment parameter, and the formula is as follows:

[0021] ω(k,l)=1 / ||(kl)||

[0022] A=max k L(k).

[0023] Furthermore, the coordinate information of the XY direction is obtained specifically as follows: in the binary image, the formula of the zero-order moment is:

[0024] M 00 =∑ I ∑ J V(i,j)

[0025] Where V(i,j) is the grayscale value of point (i,j), and its first-order definition is,

[0026] M 10 =∑ I ∑ J i·V(i,j)

[0027] M 01 =∑ I ∑ J j·V(i,j)

[0028] Where i and j are the x and y coordinates of each pixel respectively; M 10 The result is the sum of the x coordinates of all pixels in the white area, M 01The result is the sum of the y coordinates of all pixels in the white area. Using the first-order matrix, the centroid coordinates of the binary image are calculated. Specifically,

[0029]

[0030] Furthermore, the coordinates in the Z direction obtained by using the depth coordinate method are specifically that the imaging point p corresponds to the space point P one by one, and the mapping relationship is:

[0031]

[0032] Where (x, y) is the position coordinate of the spatial point P in the image physical coordinate system, (X c ,Y c ,Z c ) is the three-dimensional spatial position coordinate of the spatial point P in the camera coordinate system, and f is the internal focal length of the camera;

[0033] Represent the mapping relationship as a matrix form of a homogeneous equation

[0034]

[0035] The world coordinate system and the camera coordinate system can be converted through the following relationship

[0036]

[0037] (X w ,Y w ,Z w ) is the coordinate of the spatial point P in the world coordinate system, O = [0,0,0], R and t are obtained by calibration.

[0038] Furthermore, based on coordinate transformation and camera mapping, the center space point P of the label image is determined. At least two cameras are needed to determine the position. The tube optical centers of the left and right cameras are O1 and O2 respectively, and the imaging points of the space point P in the image coordinate pixel coordinate system of the left and right cameras are p1(x1, y1) and p2 respectively. r (x r ,y r ) are connected by O1p1 and O2p r The intersection of the two extended lines is the spatial point P, and the position of the spatial point P is determined by this.

[0039] Further, determine the disparity value x1-x r , the camera's baseline distance T and focal length f calculate the three-dimensional coordinates of the spatial point P in the camera coordinate system as,

[0040]

[0041] At this point, the position of the label image is determined, and the medical linear accelerator is controlled to achieve automatic calibration.

[0042] Furthermore, based on the optical center point O L The distance D to the space point P in the camera coordinate system is calculated as follows:

[0043] According to the above data error influence algorithm, the absolute error ΔD of distance D is calculated.

[0044] Where Δx, Δy and Δz are the absolute errors in the XYZ directions respectively;

[0045] Calculate the relative error δ of distance D;

[0046]

[0047] where δ x , δ y and δ z are the relative errors in XYZ directions respectively.

[0048] An automatic control system for a medical linear accelerator, the control system uses the automatic control method for a medical linear accelerator as described above, comprising:

[0049] Preprocessing module: collect and preprocess the label image on the treatment bed of the medical linear accelerator to obtain the coordinate information in the XY direction;

[0050] The calculation module uses the depth coordinate method to obtain the coordinates in the Z direction; based on the recognized label map coordinates, the calibration of the automatic control of the medical linear accelerator is realized.

[0051] The beneficial effects of the present invention are:

[0052] The invention is used to realize the calibration of automatic control of a medical linear accelerator.

[0053] The present invention provides automatic calibration between a medical linear accelerator and a treatment bed, which helps to make subsequent use more accurate and reduces the problems of long time and large errors that may occur in manual calibration. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a flow chart of the method of the present invention.

[0055] Figure 2 This is a schematic diagram of the measurement error caused by the position error between the medical linear accelerator and the treatment bed.

[0056] Figure 3 It is an error line graph of the X-axis of an embodiment of the present invention.

[0057] Figure 4 It is an error line graph of the Y-axis of an embodiment of the present invention.

[0058] Figure 5 It is an error line graph of the Z axis of an embodiment of the present invention.

[0059] Figure 6 It is an experimental diagram of the present invention. DETAILED DESCRIPTION

[0060] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] An automated control method for a medical linear accelerator, the automated control method comprising the following steps:

[0062] Step 1: Capture and preprocess the label image on the medical linear accelerator treatment bed to obtain the coordinate information in the XY direction;

[0063] Step 2: Based on the coordinate information in the XY direction in step 1, the coordinate in the Z direction is obtained using the depth coordinate method;

[0064] Step 3: Based on the identified label map coordinates, the calibration of the automatic control of the medical linear accelerator is realized.

[0065] Furthermore, the step 1 is specifically that the image preprocessing includes the following steps:

[0066] Step 1.1: Process the R, G, and B channels of the label image separately and calculate the R(k) value of each pixel. The formula is as follows:

[0067]

[0068] Among them, R(k) is the intermediate result when calculating the parameters, I(k) and I(l) represent the brightness of two different pixels, d(k, l) represents the distance between the two pixels, Sα(t) is the slope function, α is the adjustable slope, and the formula is as follows:

[0069]

[0070] Step 1.2: Use the following formula to expand R(k) to [0, 1] to obtain the R, G, and B channels of the enhanced label image and obtain the global white balance. The formula is as follows:

[0071]

[0072] Among them, L(k) is the normalized median value;

[0073] Step 1.3: Solve the optimization problem and obtain the processed label image. The formula is as follows:

[0074]

[0075] Among them, ω represents the weight, A is the relative adjustment parameter, and the formula is as follows:

[0076] ω(k,l)=1 / ||(kl)||

[0077] A=max k L(k).

[0078] Furthermore, the coordinate information of the XY direction is obtained specifically as follows: in the binary image, the formula of the zero-order moment is:

[0079] M 00 =∑ I ∑ J V(i,j)

[0080] Where V(i,j) is the grayscale value of point (i,j), and its first-order definition is,

[0081] M 10 =∑ I ∑ J i·V(i,j)

[0082] M 01 =∑ I ∑ J j·V(i,j)

[0083] Where i and j are the x and y coordinates of each pixel respectively; M 10 The result is the sum of the x coordinates of all pixels in the white area, M 01 The result is the sum of the y coordinates of all pixels in the white area. Using the first-order matrix, the centroid coordinates of the binary image are calculated. Specifically,

[0084]

[0085] Furthermore, the coordinates in the Z direction obtained by the depth coordinate method are specifically that the intersection point p of the line connecting the spatial point P and the optical center of the camera and the plane where the image coordinate system is located is the imaging position in the image coordinate system, and the imaging point p corresponds to the spatial point P one by one, and the mapping relationship is:

[0086]

[0087] Where (x, y) is the position coordinate of the spatial point P in the image physical coordinate system, (X c,Y c ,Z c ) is the three-dimensional spatial position coordinate of the spatial point P in the camera coordinate system, and f is the internal focal length of the camera;

[0088] Represent the mapping relationship as a matrix form of a homogeneous equation

[0089]

[0090] The world coordinate system and the camera coordinate system can be converted through the following relationship

[0091]

[0092] (X w ,Y w ,Z w ) is the coordinate of the spatial point P in the world coordinate system, O = [0,0,0], R and t are obtained by calibration.

[0093] The above calculates the distance of the pixel in the focal length direction. The collection of these distances is the depth map. The depth value of the depth map is the vertical distance from the point to the surface, that is, Z in the camera coordinate system c The coordinate value of the direction.

[0094] Furthermore, based on coordinate transformation and camera mapping, the center space point P of the label image is determined. At least two cameras are needed to determine the position. The tube optical centers of the left and right cameras are O1 and O2 respectively, and the imaging points of the space point P in the image coordinate pixel coordinate system of the left and right cameras are p1(x1, y1) and p2 respectively. r (x r ,y r ) are connected by O1p1 and O2p r The intersection of the two extended lines is the spatial point P, and the position of the spatial point P is determined by this.

[0095] Further, determine the disparity value x1-x r , the camera's baseline distance T and focal length f calculate the three-dimensional coordinates of the spatial point P in the camera coordinate system as,

[0096]

[0097] At this point, the position of the label image is determined, and the medical linear accelerator is controlled to achieve automatic calibration.

[0098] Furthermore, based on the optical center point O L The distance D to the space point P in the camera coordinate system is calculated as follows:

[0099] According to the above data error influence algorithm, the absolute error ΔD of distance D is calculated.

[0100] Where Δx, Δy and Δz are the absolute errors in the XYZ directions respectively;

[0101] Calculate the relative error δ of distance D;

[0102]

[0103] where δ x , δ y and δ z are the relative errors in XYZ directions respectively.

[0104] An automatic control system for a medical linear accelerator, the control system uses the automatic control method for a medical linear accelerator as described above, comprising:

[0105] Preprocessing module: collect and preprocess the label image on the treatment bed of the medical linear accelerator to obtain the coordinate information in the XY direction;

[0106] The calculation module uses the depth coordinate method to obtain the coordinates in the Z direction; based on the recognized label map coordinates, the calibration of the automatic control of the medical linear accelerator is realized.

[0107] The present invention Figure 3 Figure 4 Figure 5 , respectively, a comparison is made between manual verification by different verifiers and the method of the present application.

Claims

1. An automated control method for a medical linear accelerator, characterized in that: The automatic control method comprises the following steps: Step 1: Capture and preprocess the label image on the medical linear accelerator treatment bed to obtain the coordinate information in the XY direction; Step 2: Based on the coordinate information in the XY direction in step 1, the coordinate in the Z direction is obtained using the depth coordinate method; Step 3: Based on the identified label map coordinates, the calibration of the automatic control of the medical linear accelerator is realized.

2. The automated control method for a medical linear accelerator according to claim 1, characterized in that: Specifically, the step 1 includes the following steps: Step 1.1: Process the R, G, and B channels of the label image separately and calculate the R(k) value of each pixel. The formula is as follows: Among them, R(k) is the intermediate result when calculating the parameters, I(k) and I(l) represent the brightness of two different pixels, d(k, l) represents the distance between the two pixels, Sα(t) is the slope function, α is the adjustable slope, and the formula is as follows: Step 1.2: Use the following formula to expand R(k) to [0, 1] to obtain the R, G, and B channels of the enhanced label image and obtain the global white balance. The formula is as follows: Among them, L(k) is the normalized median value; Step 1.3: Solve the optimization problem and obtain the processed label image. The formula is as follows: Among them, ω represents the weight, A is the relative adjustment parameter, and the formula is as follows: ω(k,l)=1 / ||(kl)|| A=max k L(k)。 3. The automated control method for a medical linear accelerator according to claim 2, characterized in that: The coordinate information of the XY direction is obtained specifically as follows: in the binary image, the formula of the zero-order moment is: M 00 =∑ I ∑ J V(i,j) Where V(i,j) is the grayscale value of point (i,j), and its first-order definition is, M 10 =∑ I ∑ J i·V(i,j) M 01 =∑ I ∑ J j·V(i,j) Where i and j are the x and y coordinates of each pixel respectively; M 10 The result is the sum of the x coordinates of all pixels in the white area, M 01 The result is the sum of the y coordinates of all pixels in the white area. Using the first-order matrix, the centroid coordinates of the binary image are calculated. Specifically, 4. The automated control method for a medical linear accelerator according to claim 3, characterized in that: The coordinates in the Z direction obtained by the depth coordinate method are specifically that the image point p corresponds to the space point P one by one, and the mapping relationship is: Where (x, y) is the position coordinate of the spatial point P in the image physical coordinate system, (X c ,Y c ,Z c ) is the three-dimensional spatial position coordinate of the spatial point P in the camera coordinate system, and f is the internal focal length of the camera; Represent the mapping relationship as a matrix form of a homogeneous equation The world coordinate system and the camera coordinate system can be converted through the following relationship (X w ,Y w ,Z w ) is the coordinate of the spatial point P in the world coordinate system, O = [0,0,0], R and t are obtained by calibration.

5. The automated control method for a medical linear accelerator according to claim 4, characterized in that: Based on coordinate transformation and camera mapping, the center space point P of the label image is determined. At least two cameras are needed to determine the position. The tube optical centers of the left and right cameras are O1 and O2 respectively, and the imaging points of the space point P in the image coordinate pixel coordinate system of the left and right cameras are p1(x1, y1) and p r (x r ,y r ) are connected by O1p1 and O2p r The intersection of the two extended lines is the spatial point P, and the position of the spatial point P is determined by this.

6. The automated control method for a medical linear accelerator according to claim 5, characterized in that: Determine the disparity value x1-x r , the camera's baseline distance T and focal length f calculate the three-dimensional coordinates of the spatial point P in the camera coordinate system as, At this point, the position of the label image is determined, and the medical linear accelerator is controlled to achieve automatic calibration.

7. The automated control method for a medical linear accelerator according to claims 4-6, characterized in that: Based on the optical center point O L The distance D to the space point P in the camera coordinate system is calculated as follows: According to the above data error influence algorithm, the absolute error ΔD of distance D is calculated. Where Δx, Δy and Δz are the absolute errors in the XYZ directions respectively; Calculate the relative error δ of distance D; where δ x ,δ y and δ z are the relative errors in XYZ directions respectively.

8. An automated control system for a medical linear accelerator, characterized in that: The control system uses the automated control method for a medical linear accelerator as claimed in any one of claims 1 to 7, comprising: Preprocessing module: collect and preprocess the label image on the treatment bed of the medical linear accelerator to obtain the coordinate information in the XY direction; The calculation module uses the depth coordinate method to obtain the coordinates in the Z direction; based on the recognized label map coordinates, the calibration of the automatic control of the medical linear accelerator is realized.