Method for epid-based radiotherapy setup correction and dose verification

Abstract

This invention relates to a method for radiotherapy positioning correction and dose verification based on EPID (Electronic Field Imaging Device), belonging to the field of radiotherapy dose verification technology. Specifically, it involves an in vivo three-dimensional inverse dose monitoring and verification method using a reconstructed MVCT based on EPID after registration. The method first reconstructs the structural information of the phantom by acquiring two-dimensional field images, then uses a specific three-dimensional registration method to register the reconstructed images with the planned CT scan to obtain the phantom's positioning error, and finally uses EPID to monitor and verify the dose received by the phantom.

Description

Technical Field

[0001] This invention relates to a radiotherapy positioning correction and dose verification method based on EPID, belonging to the field of radiotherapy dose verification technology. Specifically, it implements an in vivo three-dimensional inverse dose monitoring and verification method after registration using a reconstructed MVCT based on EPID (Electronic Field Imaging Device). Background Technology

[0002] With the continuous development of radiotherapy technology, including the introduction of IMRT and VMAT techniques, higher demands are placed on the accuracy of treatment planning, thus requiring strict quality assurance. EPID is a highly efficient device capable of rapidly acquiring images for radiotherapy dose verification. Dose verification is divided into pre-treatment and in vivo methods. In vivo dose verification involves acquiring EPID radiation field images during actual treatment, combining them with patient CT data, reconstructing the dose distribution within the patient's body, and comparing it with the in vivo dose calculated by TPS.

[0003] When using EPD for dose verification, the patient's positioning must first be confirmed to reduce errors caused by positioning. Common methods have the following shortcomings: low contrast between the radiation field image and the DRR image, making accurate registration impossible; when using CBCT for registration, the X-ray dose values ​​acquired during CBCT acquisition are not included in the actual dose received by the patient, and the radiation source used in CBCT is different from the radiation source used during patient treatment, requiring additional radiation sources to be added to the accelerator configuration. Summary of the Invention

[0004] The technical problem solved by this invention is to overcome the shortcomings of existing technologies and propose a radiotherapy positioning correction and dose verification method based on EPID. This method first reconstructs the structural information of the phantom by acquiring two-dimensional radiation field images, then uses a specific three-dimensional registration method to register the reconstructed images with the planned CT scan to obtain the phantom's positioning error. Finally, EPID is used to monitor and verify the dose received by the phantom. This method uses the radiation source during treatment to acquire corresponding reconstructed and registered images. When calculating the dose absorbed by the phantom, positioning deviations are fully considered, and additional dose absorbed during image acquisition and registration is included in the total dose absorbed by each part of the phantom. After the fractionated treatment, the subsequent treatment plan can be corrected based on the total dose absorbed by the phantom. This method can improve the accuracy of three-dimensional dose monitoring and verification. This invention compares the calculated three-dimensional dose value with the value calculated by the radiotherapy planning system, thus verifying the accuracy of the radiotherapy planning system's calculation and execution.

[0005] The technical solution of this invention is:

[0006] A method for radiotherapy positioning correction and dose verification based on EPD (Electronic Potential Intervention), comprising the following steps:

[0007] Step 1: Acquire EPID empty field image and transmission image of the phantom. The CT image acquired during phantom localization is KVCT. Convert the acquired EPID empty field image and transmission image of the phantom into a projection image.

[0008] Step 2: Preprocess the projection image obtained in Step 1 to reduce complexity, and reconstruct the preprocessed projection image to obtain local MVCT;

[0009] Step 3: Use the registration algorithm to register the local MVCT obtained in Step 2 and the KVCT in Step 1 to achieve three-dimensional rigid registration of the local MVCT and KVCT, and obtain the positioning error;

[0010] Step 4: Correct the position of the phantom based on the positioning error obtained in Step 3, and calculate the dose value absorbed by the tumor area in the body during the local MVCT acquisition process.

[0011] Step 5: Calculate the accelerator's emission intensity value using the X-ray detector's transmission image inversion method;

[0012] Step Six: Calculate the actual dose absorption value of the tumor region in vivo based on the accelerator emission intensity value obtained in Step Five;

[0013] Step 7: Sum the dose value calculated in Step 4 and the actual dose absorption value calculated in Step 6 to obtain the actual dose value absorbed inside the phantom.

[0014] Step 8: Compare the actual dose value absorbed inside the phantom obtained in Step 7 with the total dose value designed in the treatment plan, and then revise the subsequent treatment plan.

[0015] In step one, the obtained EPID empty field image and the transmission image of the phantom are converted into a projection image using formula (1):

[0016] P = ln(I0 / I) (1)

[0017] Where, P = (p1, p2, ..., p i , ...p n ) T It is a projected image, p i Ii represents the projection value of the i-th ray, I0 is the intensity value of the EPD empty field image, and I is the intensity value of the transmitted image with the phantom; i = 1, 2, 3, ..., n, where n is the number of rays;

[0018] In step two, the preprocessing method is as follows: resample the projected image using bilinear interpolation and remove redundant invalid areas, and adjust the image size and pixel size;

[0019] In step two, the reconstruction method is as follows: the preprocessed image is reconstructed through ART iterative reconstruction, and then TV regularization is added to obtain the reconstruction result;

[0020] In step three, the registration method is as follows:

[0021] Let the local MVCT be the floating image R and the KVCT be the reference image F. The mutual information between the two is expressed by entropy as Equation (2), which is shown below:

[0022] MI(R,F)=-∑ r P R (r)log2p R (r)-∑ f P F (r)log2p F (f)+∑ r，f P RF (r, f)log2p RF (r, f)(2)

[0023] Among them, P R (r), P F (r) represent the gray-level probability distributions of the floating image R and the reference image F, respectively, P RF (r, f) represents the joint probability distribution of the two images;

[0024] Using local MVCT as the reference window and KVCT as the search area, the center point coordinates of KVCT set in the treatment plan are (x1, y1, z1). The reference window slides from the upper left corner to the lower right corner of the search area in a step of 1 pixel. After each movement, the MI values ​​of the floating image R and the reference image F at the corresponding positions are calculated. After finding the maximum MI value and the corresponding center point coordinates (x2, y2, z2), the coordinates at this time are the actual treatment center point. The positioning error is calculated based on the coordinates of the two center points.

[0025] In step four, the method for correcting the position of the phantom is as follows: if there is a positioning error between KVCT and MVCT, the phantom is moved until there is no positioning error.

[0026] In step four, the method for calculating the dose value absorbed by the tumor region in vivo during local MVCT acquisition is as follows: using KVCT and the transmission image of the phantom, the dose value generated by the transmission image in the phantom at each angle (assuming there are m angles in total) is calculated and the dose values ​​generated at each angle are superimposed. The specific dose value calculation of the transmission image of a single phantom is shown in formula (3):

[0027]

[0028] in, The convolution operation is represented by D1′(d), where D1′(d) represents the dose produced by primary radiation, D2′(d) represents the dose produced by small-scale scattering, and D3′(d) represents the dose produced by large-scale scattering. W j (x, y) represents the depositional core weights corresponding to each component, Ψ p The dose emitted by the radioactive source is then calculated using formula (3), and the results are applied to m angles and then superimposed together as FD.

[0029] In step five, the specific method for inversion is as follows:

[0030] The scattered radiation value is removed from the X-ray detector transmission image to obtain the intensity value of the primary radiation on the EPID plane. The intensity value of the primary radiation on the EPID plane is used to back-calculate the intensity value of the primary radiation in front of the incident phantom according to the exponential decay law and the inverse square law. Formula (4) realizes the back-calculation of the intensity value of the primary radiation in front of the incident phantom from the intensity value of the primary radiation on the EPID plane. Formula (4) is as follows:

[0031]

[0032] Where, ψ p (t′, r, d) represents the emission intensity of the primary ray, SID is the distance from the source to the EPID plane, d is the vertical distance from the accelerator source to the layer containing the calculation point within the phantom, t′ is the equivalent thickness from the EPID plane to the calculation point, and t is the equivalent thickness of the path traversed from the accelerator source to the calculation point on the EPID plane. Let be the intensity value of the primary ray in the EPD plane, and a(r) and b(r) be the attenuation coefficients of the primary ray for solid water.

[0033] In step six, the formula for calculating the actual dose absorption value is as follows:

[0034]

[0035] Among them, Ψ q The intensity value of the protoray in front of the incident phantom is obtained from step five. The convolution operation is represented by D1′(d), where D1′(d) represents the dose produced by primary radiation, D2′(d) represents the dose produced by small-scale scattering, and D3′(d) represents the dose produced by large-scale scattering. i (x, y) represents the deposition kernel weights corresponding to each component. The calculation result is denoted as FD1, which is the three-dimensional dose distribution within the phantom.

[0036] In step seven, the total dose value absorbed inside the phantom is obtained by adding the FD calculated in step four to the FD1 calculated in step six, denoted as TN.

[0037] In step eight, TN is compared with the dose value designed in the treatment plan and stored to provide corrections for the next treatment.

[0038] Beneficial effects

[0039] The method of this invention uses the same radiation source for acquiring and registering images as the radiation source used during treatment. Compared with CBCT, it can solve the registration error introduced when the isocenter of CBCT rotation deviates from the isocenter of the treatment source, and can also reduce the cost of machine configuration. In addition, by incorporating the radiation dose value of the registration process into the dose value absorbed by the phantom, the actual dose value absorbed by the phantom can be calculated more accurately, which can provide a basis for optimizing the subsequent treatment plan and improve the treatment effect. Attached Figure Description

[0040] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0041] To better illustrate the purpose and advantages of this invention, the invention will be described below in conjunction with the accompanying drawings and examples.

[0042] This experiment reconstructed a local MVCT (Multiple Visual Census CT) from transmission images of the phantom and empty field images. The local MVCT was then compared with the planned CT to determine the positioning error. Based on this error, the phantom's position was adjusted to ensure correct placement. The radiation dose in the transmission field during reconstruction was calculated using the transmission images. Furthermore, the radiation intensity of the radiation source was deduced from the transmission images and the ratio of primary to scattered radiation, allowing for the calculation of the actual absorbed dose inside the phantom. This calculated result was added to the radiation dose from the transmission images to obtain the total absorbed dose, which will inform subsequent treatments.

[0043] like Figure 1 As shown in the figure, the specific implementation scheme of the radiotherapy positioning correction and dose verification method based on EPD in this example is as follows:

[0044] Step 1: Acquire EPID empty field image and transmission image of the phantom. The CT image acquired during phantom localization is KVCT. Convert the acquired EPID empty field image and transmission image of the phantom into a projection image.

[0045] The implementation method for step one is as follows:

[0046] By setting the EPD source axis distance to 154cm, the dose rate to 800MU / min, the accelerator hop count to 1mu, and the field size to 20cm×20cm, and after the gantry angle was set to 20 angles in 18-degree intervals from 0 to 360 degrees, EPD empty field images and phantom transmission images were acquired respectively. The source intensity was obtained from the acquired empty field images, and then the acquired EPD empty field images and phantom transmission images were converted into projection images using formula (1), as shown in the following formula (1):

[0047] P = In(I0 / I) (1)

[0048] Where, P = (p1, p2, ... p) i ) T It is a projected image, p i I represents the projection value of the i-th ray, I0 is the intensity value of the EPD empty field image, and I is the intensity value of the transmission image with the phantom.

[0049] Step 2: Preprocess the projection image obtained in Step 1 to reduce complexity, and reconstruct the preprocessed projection image to obtain local MVCT;

[0050] The implementation method for step two is as follows:

[0051] The image was resampled using bilinear interpolation and redundant invalid regions were removed. The image size was adjusted from 768×1024 pixels (0.39×0.39mm) to 1×1mm (300×300mm), and then resampled again to obtain an image size of 100×100 pixels (3×3mm). The preprocessed image was then reconstructed using ART iterative methods, followed by TV regularization to obtain the final reconstruction result.

[0052] Step 3: Use the registration algorithm to register the local MVCT obtained in Step 2 and the KVCT in Step 1 to achieve three-dimensional rigid registration of the local MVCT and KVCT, and obtain the positioning error;

[0053] The implementation method for step three is as follows:

[0054] Let the local MVCT be the floating image R and the KVCT be the reference image F. The mutual information between the two is expressed by entropy as Equation (2), which is shown below:

[0055] MI(R,F)=-∑ r P R (r)log2p R (r)-∑ f P F (r)log2p F (f)+∑r，f P RF (r, f)log2p RF (r, f)(2)

[0056] Where P R (r), P F (r) represent the gray-level probability distributions of the floating image R and the reference image F, respectively, P RF (r, f) represents the joint probability distribution of the two images.

[0057] Using local MVCT as the reference window and KVCT as the search area, the center point coordinates of KVCT set in the treatment plan are (x1, y1, z1). The reference window slides from the upper left corner to the lower right corner of the search area in a step of 1 pixel. After each movement, the MI value of the corresponding position of the two images is calculated. After finding the maximum MI value and the corresponding center point coordinates (x2, y2, z2), the coordinates at this time are the actual treatment center point. The positioning error is calculated according to the coordinates of the two center points according to formula (3).

[0058] Formula (3) is shown below:

[0059]

[0060] Step 4: Correct the position of the phantom based on the positioning error obtained in Step 3, and calculate the dose value absorbed by the tumor area in the body during the local MVCT acquisition process.

[0061] The implementation method of step four is as follows: The method for calculating the dose value absorbed by the tumor area in the body during the local MVCT acquisition process is as follows: Using KVCT and the transmission image of the phantom, the dose value generated by the transmission image in the phantom at 20 angles is calculated and the dose values ​​generated at each angle are superimposed. The specific dose value of the transmission image of a single phantom is calculated as shown in formula (4):

[0062]

[0063] in, The convolution operation is represented by D1′(d), where D1′(d) represents the dose produced by primary radiation, D2′(d) represents the dose produced by small-scale scattering, and D3′(d) represents the dose produced by large-scale scattering. W j (x, y) represents the depositional core weights corresponding to each component, Ψ p The dose emitted by the radioactive source is then calculated using formula (4) and applied to 20 angles, and then superimposed together as FD.

[0064] Step 5: Based on Step 4, use the acquired X-ray detector transmission images to invert and calculate the accelerator's emission intensity value;

[0065] Step five is detailed as follows:

[0066] The scattered radiation value is removed from the X-ray detector transmission image to obtain the intensity value of the primary radiation on the EPID plane. The intensity value of the primary radiation on the EPID plane is then used to inversely deduce the intensity value of the primary radiation in front of the incident phantom based on the exponential decay law and the inverse square law. Simultaneously, formula (5) is used to inversely deduce the intensity value of the primary radiation in front of the incident phantom from the intensity value of the primary radiation on the EPID plane. Formula (5) is shown below:

[0067]

[0068] Where, ψ p (t′, r, d) represents the original X-ray intensity along the path of the X-ray, SID is the distance from the source to the EPID plane, d is the vertical distance from the accelerator source to the layer containing the calculation point within the phantom, t′ is the equivalent thickness from the EPID plane to the calculation point, and t is the equivalent thickness of the path from the accelerator source to the calculation point on the EPID plane. Let be the intensity value of the primary ray in the EPD plane, and a(r) and b(r) be the attenuation coefficients of the primary ray for solid water.

[0069] Step Six: Calculate the actual dose absorption value of the tumor region in vivo based on the accelerator emission intensity value obtained in Step Five;

[0070] Step Six: Detailed steps: The formula for calculating the actual dose absorption value is as follows:

[0071]

[0072] Among them, Ψ q The intensity value of the protoray in front of the incident phantom is obtained from step five. The convolution operation is represented by D1′(d), where D1′(d) represents the dose produced by primary radiation, D2′(d) represents the dose produced by small-scale scattering, and D3′(d) represents the dose produced by large-scale scattering. W i (x, y) represents the deposition kernel weights corresponding to each component. The calculation result is denoted as FD1, which is the three-dimensional dose distribution within the phantom.

[0073] Step 7: Sum the dose value calculated in Step 4 and the actual dose absorption value calculated in Step 6 to obtain the actual dose value absorbed inside the phantom.

[0074] Step 7 Specific implementation method: Add the FD calculated in Step 4 to the FD1 calculated in Step 6 to obtain the total dose value absorbed inside the phantom, denoted as TN;

[0075] Step 8: Compare the actual dose value absorbed inside the phantom obtained in Step 7 with the total dose value designed in the treatment plan, and then revise the subsequent treatment plan.

[0076] Step 8: Compare and store the TN dose value with the treatment plan design to provide corrections for the next treatment.

[0077] In summary, the above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for EPID-based radiotherapy setup correction and dose verification, characterized in that The steps of this method include: Step 1: Acquire EPID empty field image and transmission image of the phantom. The CT image acquired for phantom localization is KVCT. Convert the acquired EPID empty field image and transmission image of the phantom into a projection image. Step 2: Preprocess the projection image obtained in Step 1 to reduce complexity, and reconstruct the preprocessed projection image to obtain local MVCT; Step 3: Register the local MVCT obtained in Step 2 with the KVCT in Step 1 to obtain the setup error; Step 4: Correct the position of the phantom based on the positioning error obtained in Step 3, and calculate the dose value absorbed by the tumor area in the body during the local MVCT acquisition process. Step 5: Calculate the accelerator's emission intensity value using the X-ray detector's transmission image inversion method; Step Six: Calculate the actual dose absorption value of the tumor region in vivo based on the accelerator emission intensity value obtained in Step Five; Step 7: Sum the dose value calculated in Step 4 and the actual dose absorption value calculated in Step 6 to obtain the actual dose value absorbed inside the phantom. Step 8: Compare the actual dose value absorbed inside the phantom obtained in Step 7 with the total dose value designed in the treatment plan, and then revise the subsequent treatment plan. In step two, the preprocessing method is as follows: resample the projected image using bilinear interpolation and remove redundant invalid areas, and adjust the image size and pixel size; In step two, the reconstruction method is as follows: the preprocessed image is reconstructed through ART iterative reconstruction, and then TV regularization is added to obtain the reconstruction result; In step three, the registration method is as follows: Let the local MVCT be the floating image R and the KVCT be the reference image F. The mutual information between the two is expressed by entropy as Equation (2), which is shown below: （2） wherein, , respectively denote the gray level probability distributions of the floating image R and the reference image F, denotes the joint probability distribution of the two images; Using local MVCT as the reference window and KVCT as the search area, the coordinates of the KVCT treatment center point set in the treatment plan are: The reference window slides from the top left corner to the bottom right corner of the search area in 1-pixel increments. After each movement, the MI values ​​at corresponding positions in the floating image R and the reference image F are calculated. The maximum MI value and the corresponding center point coordinates are then found. Then, the coordinates at this point are the actual treatment center point, and the positioning error is calculated based on the coordinates of the two center points.

2. The method for radiotherapy positioning correction and dose verification based on EPD according to claim 1, characterized in that: In step one, the obtained EPID empty field image and the transmission image of the phantom are converted into a projection image using formula (1): （1） in, It is a projected image. Indicates the first The projection value of the ray, The intensity value of the empty field image in EPD. The intensity value of the transmitted image of the phantom is i; i = 1, 2, 3, ..., n, where n is the number of rays.

3. The method for radiotherapy positioning correction and dose verification based on EPD according to claim 1, characterized in that: In step four, the method for correcting the position of KVCT is as follows: if there is a positioning error in the registration of KVCT and MVCT, the phantom is moved until there is no positioning error.

4. The method for radiotherapy positioning correction and dose verification based on EPD according to claim 3, characterized in that: In step four, the method for calculating the dose value absorbed by the tumor region in vivo during local MVCT acquisition is as follows: using the transmission image of the phantom, assuming there are m angles, calculate the dose value generated by the transmission image in the phantom at each angle and superimpose the dose values ​​generated at each angle. The specific dose value calculation for a single transmission image is shown in formula (3): （3） in, This represents the convolution operation. This indicates the dose produced by primary radiation. This indicates the dose produced by small-scale scattering. This indicates the dose produced by wide-range scattering. This represents the deposition kernel weights corresponding to each component. The dose emitted by the radioactive source is then calculated using formula (3), applied to m angles, and summed together as follows: .

5. The method for radiotherapy positioning correction and dose verification based on EPD according to claim 4, characterized in that: In step five, the specific method for inversion is as follows: The scattered radiation value is removed from the X-ray detector transmission image to obtain the intensity value of the original radiation on the EPID plane. The intensity value of the original radiation on the EPID plane is used to back-calculate the intensity value of the original radiation in front of the incident phantom according to the exponential decay law and the inverse square law. Formula (4) realizes the back-calculation of the intensity value of the original radiation on the EPID plane to obtain the intensity value of the original radiation in front of the incident phantom. Formula (4) is as follows: （4） in, Here, SID represents the emission intensity of the primary ray, SID is the distance from the source to the EPDD plane, and d is the vertical distance from the accelerator source to the layer containing the calculation point within the phantom. The equivalent thickness from the EPDI plane to the calculation point, The equivalent thickness of the path from the accelerator source to the calculation point in the EPID plane. The intensity value of the original ray in the EPD plane. and is the attenuation coefficient of the primary radiation for solid water.

6. The method for radiotherapy positioning correction and dose verification based on EPD according to claim 5, characterized in that: In step six, the formula for calculating the actual dose absorption value is as follows: （5） in, The intensity value of the protoray in front of the incident phantom is obtained from step five. This represents the convolution operation. This indicates the dose produced by primary radiation. This indicates the dose produced by small-scale scattering. This indicates the dose produced by wide-range scattering. This represents the depositional nucleus weights corresponding to each component, and the calculation results are denoted as... This result represents the three-dimensional dose distribution within the phantom.

7. The method for radiotherapy positioning correction and dose verification based on EPD according to claim 6, characterized in that: In step seven, the calculations from step four are... Compared with the calculation in step six By adding them together, we can obtain the total dose value absorbed inside the phantom.

Images