Flux graph generation method and device and treatment plan generation method and device

Abstract

The invention provides a flux graph generation method and device and a treatment plan generation method and device, relates to the technical field of computers, and solves the technical problems that a convolutional neural network lacks understanding of detailed information of a CT image, and a flux graph predicted based on the convolutional neural network is too smooth and is greatly different from a real flux graph in related technologies. The method comprises: determining a dose distribution or an initial flux map of a target volume at each of a plurality of field angles, the target volume comprising a target region and an organ at risk; inputting the dose distribution or the initial flux diagram of the target volume at each radiation field angle into a first flux diagram generation model based on a denoising diffusion probability model to obtain a flux diagram corresponding to each radiation field angle; wherein the flux diagram corresponding to each radiation field angle is used for representing the intensity of each small beam in the target volume corresponding to each radiation field angle.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method for generating a flux map, a method for generating a treatment plan, and a device. Background Art

[0002] Currently, a server can input a computed tomography (CT) image into a convolutional neural network to predict the flux map of a certain field angle.

[0003] However, the convolutional neural network in the above method lacks the understanding of the detailed information of the CT image (such as the gradient of the CT image), so that the predicted flux map is too smooth and has a large difference from the actual flux map. Summary of the invention

[0004] The present application provides a flux map generation method, a treatment plan generation method and a device, which solve the technical problems in the related art that the convolutional neural network lacks understanding of the detailed information of the CT image, and the flux map predicted by the convolutional neural network is too smooth and has a large difference from the actual flux map.

[0005] In a first aspect, the present application provides a flux map generation method, comprising: determining a dose distribution or an initial flux map of a target volume at each of a plurality of field angles, the target volume including a target area and organs at risk; inputting the dose distribution or the initial flux map of the target volume at each of the field angles into a first flux map generation model based on a denoised diffusion probability model to obtain a flux map corresponding to each of the field angles; wherein the flux map corresponding to each of the field angles is used to characterize the intensity of each small beam within the target volume corresponding to each of the field angles.

[0006] Optionally, the above-mentioned determination of the dose distribution of the target volume at each of the multiple field angles specifically includes: acquiring the image and contour data of the target volume; inputting the image and contour data of the target volume into a field dose prediction model to obtain the predicted three-dimensional dose distribution of the target volume at each field angle; and determining the predicted three-dimensional dose distribution of the target volume at each field angle as the dose distribution of the target volume at each field angle.

[0007] Optionally, the above-mentioned determination of the dose distribution of the target volume at each field angle among multiple field angles specifically also includes: projecting the predicted three-dimensional dose distribution of the target volume at each field angle onto a projection plane orthogonal to each field angle to obtain the dose distribution of the target volume at each field angle.

[0008] Optionally, determining the initial flux map of the target volume at each of the multiple field angles specifically includes: acquiring image and contour data of the target volume; inputting the image and contour data of the target volume into a neural network model for flux map prediction to obtain the initial flux map of the target volume at each of the multiple field angles.

[0009] Optionally, the flux map generation method also includes: acquiring structural data of the target volume at each of the multiple field angles; inputting the structural data of the target volume at each of the field angles, combined with the dose distribution or the initial flux map, into a second flux map generation model based on a denoised diffusion probability model to obtain a flux map corresponding to each field angle.

[0010] In a second aspect, the present application provides a method for generating a treatment plan, comprising: acquiring an image and contour data of a target volume, the target volume including a target area and organs at risk; determining a dose distribution or an initial flux map of the target volume at each of a plurality of field angles based on the image and contour data of the target volume; inputting the dose distribution or the initial flux map of the target volume at each of the field angles into a first flux map generation model based on a denoised diffusion probability model to obtain a flux map corresponding to each of the field angles, wherein the flux map corresponding to each of the field angles is used to characterize the intensity of each small beam in the target volume corresponding to each of the field angles; and generating a treatment plan based on the flux map corresponding to each of the field angles.

[0011] Optionally, the above-mentioned determining the dose distribution of the target volume at each of a plurality of field angles based on the image and contour data of the target volume specifically includes: inputting the image and contour data of the target volume into a field dose prediction model to obtain a predicted three-dimensional dose distribution of the target volume at each field angle; and determining the predicted three-dimensional dose distribution of the target volume at each field angle as the dose distribution of the target volume at each field angle.

[0012] Optionally, the above-mentioned determining the dose distribution of the target volume at each of a plurality of field angles based on the image and contour data of the target volume specifically also includes: projecting the predicted three-dimensional dose distribution of the target volume at each field angle onto a projection plane orthogonal to each field angle to obtain the dose distribution of the target volume at each field angle.

[0013] Optionally, determining the initial flux map of the target volume at each of the multiple field angles based on the image and contour data of the target volume specifically includes: inputting the image and contour data of the target volume into a neural network model for flux map prediction to obtain the initial flux map of the target volume at each of the multiple field angles.

[0014] Optionally, the method for generating a treatment plan also includes: acquiring structural data of the target volume at each of the multiple field angles; inputting the structural data of the target volume at each of the field angles, combined with the dose distribution or the initial flux map, into a second flux map generation model based on a denoised diffusion probability model to obtain a flux map corresponding to each of the field angles; and generating a treatment plan based on the flux map corresponding to each of the field angles.

[0015] In a third aspect, the present application provides an electronic device comprising: a processor and a memory configured to store processor executable instructions; wherein the processor is configured to execute the instructions to implement any one of the optional flux map generation methods in the above-mentioned first aspect, or to implement any one of the optional treatment plan generation methods in the above-mentioned second aspect.

[0016] In a fourth aspect, the present application provides a computer-readable storage medium having instructions stored thereon. When the instructions in the computer-readable storage medium are executed by an electronic device, the electronic device is enabled to execute any one of the optional flux map generation methods in the first aspect above, or execute any one of the optional treatment plan generation methods in the second aspect above.

[0017] In a fifth aspect, the present application provides a computer program product, which includes computer instructions. When the computer instructions are executed on an electronic device, the electronic device executes any optional flux map generation method as described in the first aspect, or executes any optional treatment plan generation method as described in the second aspect.

[0018] The flux map generation method, treatment plan generation method and device provided in the present application, the electronic device can first determine the dose distribution or initial flux map of the target volume at each of the multiple field angles; then the electronic device can input the dose distribution or initial flux map of the target volume at each field angle into the first flux map generation model based on the denoising diffusion probability model, and obtain the flux map corresponding to each field angle, and the flux map corresponding to each field angle is used to characterize the intensity of each small beam in the target volume corresponding to each field angle. In the present application, since the denoising diffusion probability model is a deep learning model based on likelihood estimation, higher quality images can be generated based on the denoising diffusion probability model. In this way, the electronic device inputs the dose distribution or initial flux map of the target volume at each field angle into the first flux map generation model based on the denoising diffusion probability model, and can generate a flux map of higher quality corresponding to each field angle, which can improve the authenticity of the flux map generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0020] Figure 1 A schematic diagram of a flow chart of a method for generating a flux map provided in an embodiment of the present application;

[0021] Figure 2 A schematic diagram of the principle of determining a dose distribution and an initial flux map provided in an embodiment of the present application;

[0022] Figure 3 A schematic diagram of a flow chart of another method for generating a flux map provided in an embodiment of the present application;

[0023] Figure 4 A flowchart of a method for generating a treatment plan provided in an embodiment of the present application;

[0024] Figure 5 Another schematic diagram of the principle of determining the dose distribution and the initial flux map provided in the embodiment of the present application;

[0025] Figure 6 A flowchart of another method for generating a treatment plan provided in an embodiment of the present application;

[0026] Figure 7 A schematic diagram of the structure of a flux map generating device provided in an embodiment of the present application;

[0027] Figure 8 A schematic diagram of the structure of a treatment plan generation device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The flux map generation method, treatment plan generation method and device provided in the embodiments of the present application will be described in detail below in conjunction with the accompanying drawings.

[0029] The terms “first”, “second” and the like in the specification and drawings of the present application are used to distinguish different objects rather than to describe a specific order of the objects.

[0030] In addition, the terms "including" and "having" and any variations thereof mentioned in the description of the present application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0031] It should be noted that in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0032] The term "and / or" used in the present application includes using either or both of the two methods.

[0033] In the description of the present application, unless otherwise specified, “plurality” means two or more.

[0034] The following is an explanation of some concepts involved in a flux map generation method, a treatment plan generation method and a device provided in an embodiment of the present application.

[0035] Radiotherapy is one of the important means of treating tumors. It achieves the purpose of destroying tumors by giving tumors a sufficiently high lethal radiation dose. Since radiation cannot distinguish between tumor cells and normal tissues, normal tissues will also be irradiated to varying degrees by radiation when a high dose is given to the tumor. Clinically, a treatment plan with modulated radiation field intensity distribution is usually used to achieve the purpose of giving a high dose to the tumor without causing excessive doses to normal tissues.

[0036] Intensity modulated therapy is a commonly used technology in tumor radiotherapy. The intensity modulated therapy plan requires the planner to continuously adjust the dose volume constraint target and the corresponding weight to achieve the dose distribution required by the clinical situation. In order to obtain better dosimetry targets, the planner needs excellent plan design skills and more optimization time to achieve the desired goals. The quality of the plan depends largely on the planner's experience and plan optimization time.

[0037] Based on the description in the background technology, since in the related technology, the convolutional neural network lacks the understanding of the detailed information of the CT image (such as the gradient of the CT image), the flux map predicted by inputting the CT image into the convolutional neural network is too smooth, and the difference from the flux map of the real image is large. Based on this, the embodiment of the present application provides a flux map generation method, a treatment plan generation method and device. Since the denoising diffusion probability model is a deep learning model based on likelihood estimation, higher quality images can be generated based on the denoising diffusion probability model. In this way, the electronic device inputs the dose distribution or initial flux map of the target volume at each field angle into the first flux map generation model based on the denoising diffusion probability model, and can generate a high-quality flux map corresponding to each field angle, which can improve the authenticity of the flux map generation.

[0038] like Figure 1 As shown, the flux map generation method provided in the embodiment of the present application may include S101-S102.

[0039] S101. The electronic device determines a dose distribution or an initial flux map of a target volume at each of a plurality of field angles.

[0040] The target volume includes the target area and organs at risk.

[0041] It should be understood that the dose distribution of the target volume at each field angle can characterize the dose of radiation received by the target volume at each field angle, and the initial flux map of the target volume at each field angle can characterize the intensity of each beamlet in the target volume corresponding to each field angle.

[0042] In the embodiment of the present application, a field angle can also be understood as a field direction.

[0043] Optionally, the electronic device may equally divide 360° according to preset angles (eg, 40° or 45°) to obtain the above-mentioned multiple field angles, or may arbitrarily provide some angles (eg, 20°, 80°) as the multiple field angles.

[0044] It should be noted that one form of the dose distribution in the embodiment of the present application may be image information of the dose distribution, such as a dose distribution image. The following description of the present application scheme will take the dose distribution image as an example. Another form of the dose distribution may also be parameter information (or data information) of the dose distribution, such as text or code that can characterize the dose distribution. The embodiment of the present application does not specifically limit the form of the dose distribution.

[0045] S102: The electronic device inputs the dose distribution or initial flux map of the target volume at each field angle into a first flux map generation model based on a denoising diffusion probability model to obtain a flux map corresponding to each field angle.

[0046] The flux map corresponding to each field angle is used to characterize the intensity of each beamlet in the target volume corresponding to each field angle.

[0047] In an optional implementation, the first flux map generation model may be obtained by training with dose distribution as a training sample, and the input of the first flux map generation model is the dose distribution of the target volume at each field angle. That is, the electronic device inputs the dose distribution of the target volume at each field angle into the first flux map generation model to obtain the flux map corresponding to each field angle.

[0048] In another optional implementation, the first flux map generation model may be obtained by training with flux maps as training samples, and the input of the first flux map generation model is the flux map of the target volume at each field angle. That is, the electronic device may input the initial flux map of the target volume at each field angle into the first flux map generation model to obtain the flux map corresponding to each field angle.

[0049] It can be understood that the denoised diffusion probability model is a deep learning model based on likelihood estimation. Compared with the convolutional neural model (or convolutional neural network) in the prior art, the electronic device inputs the dose distribution or initial flux map of the target volume at each field angle into the first flux map generation model based on the denoised diffusion probability model, and can generate higher quality images, that is, the flux map corresponding to each field angle obtained is a flux map of higher quality.

[0050] It is worth noting that, based on the denoising diffusion probability theory, the present application can decompose the flux map generation task (or prediction task) into a forward noise diffusion stage and a reverse denoising stage. In the forward noise diffusion stage, the dose distribution image or the initial flux map is gradually increased with Gaussian noise over a period of time T; in the reverse denoising stage, a high-quality flux map is gradually restored.

[0051] The technical solution provided by the above embodiment can at least bring the following beneficial effects: From S101-S102, it can be known that the electronic device can first determine the dose distribution or initial flux map of the target volume at each field angle among multiple field angles; then the electronic device can input the dose distribution or initial flux map of the target volume at each field angle into the first flux map generation model based on the denoising diffusion probability model to obtain the flux map corresponding to each field angle, and the flux map corresponding to each field angle is used to characterize the intensity of each small beam in the target volume corresponding to each field angle. In the embodiment of the present application, since the denoising diffusion probability model is a deep learning model based on likelihood estimation, a higher quality image can be generated based on the denoising diffusion probability model. In this way, the electronic device inputs the dose distribution or initial flux map of the target volume at each field angle into the first flux map generation model based on the denoising diffusion probability model, and can generate a flux map of higher quality corresponding to each field angle, which can improve the authenticity of the flux map generation.

[0052] like Figure 2 As shown, in one implementation of the embodiment of the present application, the electronic device determines the dose distribution of the target volume at each of a plurality of field angles, which may specifically include S1011-S1013.

[0053] S1011. The electronic device obtains the image and contour data of the target volume.

[0054] In an optional implementation, the image of the target volume may be a CT image, MRI image, or PET image of a patient. After the CT image, MRI image, or PET image is imported into a treatment planning system (TPS), the target area and organs at risk of the patient may be delineated, that is, the target volume may be determined.

[0055] It is understandable that the CT image, MRI image or PET image may also be imported into the contouring system to outline the target area and organs at risk of the patient, so as to determine the target volume.

[0056] It is worth noting that the contour data includes the contour of the target area and the contour of the organ at risk. The contour of the target area can represent the position of the target area in the target volume, and the contour of the organ at risk can represent the position of the organ at risk in the target volume.

[0057] S1012. The electronic device inputs the image and contour data of the target volume into a radiation field dose prediction model to obtain a predicted three-dimensional dose distribution of the target volume at each radiation field angle.

[0058] It should be understood that the field dose prediction model is a pre-trained neural network model for predicting the three-dimensional dose distribution of a volume (eg, a target volume) at each field angle.

[0059] S1013. The electronic device determines the predicted three-dimensional dose distribution of the target volume at each field angle as the dose distribution of the target volume at each field angle.

[0060] It is worth noting that the predicted three-dimensional dose distribution of the target volume at each field angle predicted by the electronic device based on the field dose prediction model is the three-dimensional spatial dose of the target volume at each field angle, and the three-dimensional spatial dose of the target volume at each field angle can characterize the spatial information of the dose of the target volume at each field angle. In the embodiment of the present application, the electronic device inputs the image and contour data of the target volume into the field dose prediction model, and can accurately and effectively predict the three-dimensional dose distribution (or three-dimensional spatial dose) of the target volume at each field angle, that is, it can obtain the spatial information that can characterize the dose of the target volume at each field angle. In addition, since there is a corresponding relationship between the dose distribution of the target volume at each field angle determined by the electronic device and the flux map corresponding to each field angle, the flux map corresponding to each field angle can be accurately and effectively determined based on the three-dimensional dose distribution (or three-dimensional spatial dose) of the target volume at each field angle.

[0061] Continue as Figure 2As shown, the electronic device provided in the embodiment of the present application determines the dose distribution of the target volume at each of the multiple field angles, and specifically may further include S1014.

[0062] S1014. The electronic device projects the predicted three-dimensional dose distribution of the target volume at each field angle onto a projection plane orthogonal to each field angle to obtain the dose distribution of the target volume at each field angle.

[0063] It should be understood that the electronic device can first determine the projection plane orthogonal to each field angle. After that, the electronic device can use a ray tracing algorithm to project the predicted three-dimensional dose distribution (or three-dimensional spatial dose) of the target volume at each field angle to a projection plane orthogonal to each field angle, that is, the two-dimensional dose projection of the target volume at each field angle can be obtained. In an embodiment of the present application, the electronic device can compress the three-dimensional dose distribution (or three-dimensional spatial dose) used to characterize the spatial information of the dose of the target volume at each field angle into a two-dimensional dose projection that can characterize the two-dimensional structure of the target volume at each field angle, that is, the two-dimensional dose projection of the target volume at each field angle can accurately and effectively reflect the amount of radiation (or size) that the target volume (including the target area and the organs at risk) can receive. In this way, the electronic device determines the two-dimensional dose projection of the target volume at each field angle as the dose distribution of the target volume at each field angle, and can accurately and effectively determine the flux map corresponding to each field angle.

[0064] Continue as Figure 2 As shown, the electronic device determines an initial flux map of the target volume at each of a plurality of field angles, which may specifically include S1011 and S1015.

[0065] S1011. The electronic device obtains the image and contour data of the target volume.

[0066] It should be understood that S1011 has been explained in the above embodiments and will not be repeated here.

[0067] S1015. The electronic device inputs the image and contour data of the target volume into a neural network model for flux map prediction to obtain an initial flux map of the target volume at each of a plurality of field angles.

[0068] It is worth noting that the neural network model is used to predict the initial flux map of a volume (such as a target volume) at each field angle. The neural network model can be understood as a trained network model with high prediction accuracy (or precision).

[0069] Specifically, the electronic device can train a network model with low prediction accuracy (or precision) in an initial state based on the images of each of the multiple volumes, the contour data of each of the multiple volumes, and the flux maps of each of the multiple volumes at each field angle to obtain the above-mentioned neural network model for flux map prediction.

[0070] Optionally, the neural network model can be a convolutional neural network (CNN).

[0071] In an embodiment of the present application, after the electronic device acquires the image and contour data of the target volume, it inputs the image and contour data of the target volume into a neural network model for flux map prediction, so as to conveniently and quickly obtain the initial flux map of the target volume at each of multiple field angles, thereby improving the generation rate of the flux map corresponding to each field angle.

[0072] Combination Figure 1 ,like Figure 3 As shown, the flux map generation method provided in the embodiment of the present application may also include S103-S104.

[0073] S103: The electronic device obtains structural data of the target volume at each of a plurality of field angles.

[0074] The structural data is used to characterize the tissue morphology of the target area and organs at risk included in the target volume.

[0075] S104, the electronic device inputs the structural data of the target volume at each field angle, combined with the dose distribution or the initial flux map, into a second flux map generation model based on a denoising diffusion probability model to obtain a flux map corresponding to each field angle.

[0076] Specifically, the electronic device may input the structural data of the target volume at each field angle and the dose distribution of the target volume at each field angle into the second flux map generation model to obtain the flux map corresponding to each field angle. Alternatively, the electronic device may input the structural data of the target volume at each field angle and the initial flux map of the target volume at each field angle into the second flux map generation model to obtain the flux map corresponding to each field angle.

[0077] It should be understood that the above-mentioned second flux map generation model can be obtained by training with structural data and dose distribution as training samples, and in this case, the input of the second flux map generation model is the structural data of the target volume at each field angle and the dose distribution of the target volume at each field angle. The second flux map generation model can also be obtained by training with structural data and initial flux map as training samples, and in this case, the input of the second flux map generation model is the structural data of the target volume at each field angle and the initial flux map of the target volume at each field angle.

[0078] In the embodiment of the present application, since the structural data of the target volume at each field angle can characterize the tissue morphology of the target area and the organ at risk included in the target volume, the electronic device inputs the structural data of the target volume at each field angle and the dose distribution (or initial flux map) of the target volume at each field angle into the second flux map generation model based on the denoising diffusion probability model, and can obtain a flux map with image quality equivalent to that of a real flux map, which can improve the authenticity of the flux map.

[0079] In the embodiment of the present application, after obtaining (or acquiring) the flux map corresponding to each field angle, the electronic device can generate a treatment plan based on the flux map corresponding to each field angle.

[0080] Based on this, Figure 4 As shown, the method for generating a treatment plan provided in the embodiment of the present application may include S201-S204.

[0081] S201. The electronic device obtains an image and contour data of a target volume.

[0082] The target volume includes the target area and organs at risk.

[0083] S202: The electronic device determines the dose distribution or initial flux map of the target volume at each of a plurality of field angles according to the image and contour data of the target volume.

[0084] S203 , the electronic device inputs the dose distribution or initial flux map of the target volume at each field angle into a first flux map generation model based on a denoising diffusion probability model to obtain a flux map corresponding to each field angle.

[0085] The flux map corresponding to each field angle is used to characterize the intensity of each small beam in the target volume corresponding to each field angle.

[0086] It should be noted that the explanations in S201 and S203 are the same or similar to the descriptions in S1011 and S102 above, and will not be repeated here.

[0087] S204: The electronic device generates a treatment plan according to the flux map corresponding to each radiation field angle.

[0088] It should be understood that the treatment plan may include the position information (or coordinates) of the multi-leaf laser blades, the number of hops of each sub-field at each moment, and the weight corresponding to the number of hops of each sub-field at each moment.

[0089] It can be understood that the number of jumps of each sub-field at each moment is determined by the product of the weight corresponding to the number of jumps of each sub-field at each moment and the prescription dose, and the prescription dose is the dose that the target volume should receive given in advance by the doctor.

[0090] In an optional implementation, the electronic device may also evaluate the treatment plan after generating it. Specifically, the electronic device may generate relevant dose distribution and / or dose volume histogram (DVH) based on the information in the treatment plan. Then, it is determined whether the dose reflected by the dose distribution and / or DVH map is within the limit range, which can also be understood as determining whether the dose reflected by the dose distribution and / or DVH map meets the clinical standard. After the treatment plan is evaluated as qualified, the electronic device outputs the treatment plan and the radiotherapy device executes the treatment plan.

[0091] The technical solution provided by the above embodiment can at least bring the following beneficial effects: From S201-S204, it can be known that the electronic device can obtain the image and contour data of the target volume, and determine the dose distribution or initial flux map of the target volume at each field angle among multiple field angles according to the image and contour data of the target volume; then, the electronic device can input the dose distribution or initial flux map of the target volume at each field angle into the first flux map generation model based on the denoising diffusion model to obtain the flux map corresponding to each field angle; finally, the electronic device can generate a treatment plan according to the flux map corresponding to each field angle. In the embodiment of the present application, since the denoising diffusion probability model is a deep learning model based on likelihood estimation, a higher quality image can be generated based on the denoising diffusion probability model. In this way, the electronic device inputs the dose distribution or initial flux map of the target volume at each field angle into the first flux map generation model based on the denoising diffusion probability model, and can generate a high-quality flux map corresponding to each field angle. Then, the electronic device can generate a real and effective treatment plan according to the flux map corresponding to each field angle.

[0092] like Figure 5 As shown, the electronic device determines the dose distribution of the target volume at each of a plurality of field angles according to the image and contour data of the target volume, which may specifically include S2021-S2023.

[0093] S2021. The electronic device inputs the image and contour data of the target volume into the field dose prediction model to obtain the predicted three-dimensional dose distribution of the target volume at each field angle.

[0094] S2022. The electronic device determines the predicted three-dimensional dose distribution of the target volume at each field angle as the dose distribution of the target volume at each field angle.

[0095] It should be noted that the explanations in S2021-S2022 are the same or similar to the descriptions in the above-mentioned S1012-S1013 and will not be repeated here.

[0096] In the embodiment of the present application, since the target feature of the target volume can characterize the pixel value of each pixel point included in the target volume and the positional relationship between the target area included in the target volume and the organ at risk. In this way, the electronic device inputs the target feature of the target volume into the field dose prediction model, and can accurately and effectively predict the three-dimensional dose distribution (or three-dimensional spatial dose) of the target volume at each field angle, that is, it can obtain spatial information that can characterize the dose of the target volume at each field angle. In addition, since there is a corresponding relationship between the dose distribution of the target volume at each field angle determined by the electronic device and the flux map corresponding to each field angle, the flux map corresponding to each field angle can be accurately and effectively determined based on the three-dimensional dose distribution (or three-dimensional spatial dose) of the target volume at each field angle.

[0097] Continue as Figure 5 As shown, the electronic device determines the dose distribution of the target volume at each of the multiple field angles according to the image and contour data of the target volume, and specifically may further include S2023.

[0098] S2023. Project the predicted three-dimensional dose distribution of the target volume at each field angle onto a projection plane orthogonal to each field angle to obtain the dose distribution of the target volume at each field angle.

[0099] It should be noted that the explanation in S2023 is the same or similar to the description in the above S1014 and will not be repeated here.

[0100] In an embodiment of the present application, the electronic device can compress the three-dimensional dose distribution (or three-dimensional spatial dose) used to characterize the three-dimensional spatial information of the target volume at each field angle into a two-dimensional dose projection that can characterize the two-dimensional structure of the target volume at each field angle, that is, the two-dimensional dose projection of the target volume at each field angle can accurately and effectively reflect the amount (or size) of radiation that the target volume (including the target area and the critical organ) can receive. In this way, the electronic device determines the two-dimensional dose projection of the target volume at each field angle as the dose distribution of the target volume at each field angle, and can accurately and effectively determine the flux map corresponding to each field angle.

[0101] Continue as Figure 5 As shown, the electronic device determines an initial flux map of the target volume at each of a plurality of field angles according to the image and contour data of the target volume, which may specifically include S2024.

[0102] S2024. The electronic device inputs the image and contour data of the target volume into a neural network model for flux map prediction to obtain an initial flux map of the target volume at each of a plurality of field angles.

[0103] It should be noted that the explanation in S2024 is the same or similar to the description in the above S1015 and will not be repeated here.

[0104] In an embodiment of the present application, after the electronic device acquires the image and contour data of the target volume, it inputs the image and contour data of the target volume into a neural network model for flux map prediction, so as to conveniently and quickly obtain the initial flux map of the target volume at each of multiple field angles, thereby improving the generation rate of the flux map corresponding to each field angle.

[0105] Combination Figure 4 ,like Figure 6 As shown, the method for generating a treatment plan provided in the embodiment of the present application may further include S205-S207.

[0106] S205. The electronic device obtains structural data of the target volume at each of a plurality of field angles.

[0107] In combination with the description of the above embodiments, it should be understood that the structural data is used to characterize the tissue morphology of the target area and the organ at risk included in the target volume.

[0108] S206. The electronic device inputs the structural data of the target volume at each field angle, combined with the dose distribution or the initial flux map, into a second flux map generation model based on a denoising diffusion probability model to obtain a flux map corresponding to each field angle.

[0109] S207: The electronic device generates a treatment plan according to the flux map corresponding to each radiation field angle.

[0110] It should be noted that the explanations in S205-S207 are the same or similar to the descriptions in the above-mentioned S103-S104 and S204, and will not be repeated here.

[0111] In the embodiment of the present application, since the structural data of the target volume at each field angle can characterize the tissue morphology of the target area and the organ at risk included in the target volume. In this way, the electronic device inputs the structural data of the target volume at each field angle and the dose distribution (or initial flux map) of the target volume at each field angle into the second flux map generation model based on the denoising diffusion probability model, and can obtain a flux map with image quality equivalent to the real flux map, which can improve the authenticity of the flux map. Afterwards, the electronic device can generate a real and effective treatment plan based on the flux map corresponding to each field angle.

[0112] The following uses 9-field intensity modulated radiation therapy (IMRT) for cervical cancer as an example to illustrate the flux map generation process and the treatment plan evaluation process provided by the embodiment of the present application. The flux map generation process may include steps A to J, and the treatment plan evaluation process may include steps K to N.

[0113] Step A: Import the patient's CT image into TPS, outline the target area and the contours of the organs at risk, and approve the structure and contour definition of the target area and the organs at risk after the doctor has checked and confirmed that they are correct.

[0114] Step B: Select treatment equipment and treatment technology.

[0115] Exemplarily, the treatment equipment may include a linear accelerator and a gamma knife, etc., and the treatment technology may be static intensity modulation or dynamic intensity modulation.

[0116] Step C: Define the required prescription dose, number of fractions and target volume for the given plan.

[0117] The prescribed dose required for the given plan is used to characterize the amount of radiation that the patient needs to receive during the treatment process, the number of fractions is used to characterize how many times of radiotherapy are required in total, and the planned target area definition can characterize the target area selected for treatment this time.

[0118] Step D: Set the isocenter and field direction.

[0119] Step E: Perform three-dimensional reconstruction according to the target area and the outline of the organ at risk defined above to obtain the corresponding three-dimensional structure matrix.

[0120] Step F: normalize the CT image, construct a 3D grayscale matrix, and superimpose it with the three-dimensional structure matrix obtained in step E in the channel dimension.

[0121] Step G: Input the features obtained in step F into the field dose prediction network to predict the three-dimensional dose distribution at each field angle.

[0122] Step H: Based on the three-dimensional dose distribution at each field angle obtained in step G and the corresponding field angle, a ray tracing algorithm is used to project the three-dimensional dose distribution onto a projection plane orthogonal to the field angle to form a two-dimensional dose projection feature matrix.

[0123] Step I: input the dose projection feature matrix calculated in step H into the flux map prediction network (i.e., the first flux map generation model mentioned above) to perform flux map prediction.

[0124] Step J, iterate the above process to predict the flux map corresponding to each field angle.

[0125] Step K: Generate a treatment plan based on the flux map corresponding to each field angle obtained in step J.

[0126] The treatment plan includes the position information of the multi-leaf laser blades, the number of jumps of each sub-field at each moment, and the weight corresponding to the number of jumps of each sub-field at each moment.

[0127] Step L: Generate dose distribution and DVH map according to treatment plan information.

[0128] Step M: determine whether the dose distribution and the dose reflected by the DVH map in step L are within the limit range (or whether they meet the clinical standards).

[0129] Step N: When the dose reflected by the dose distribution and the DVH diagram is within the limit range, it means that the treatment plan generated in step K is qualified, and the treatment plan can be output and executed by the radiotherapy equipment.

[0130] The embodiment of the present application can divide the functional modules of the electronic device etc. according to the above method example. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the embodiment of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0131] In the case of an integrated unit, Figure 7 FIG. 2 shows a possible structural diagram of the flux map generating device involved in the above embodiment. Figure 7As shown, the flux map generating device 10 may include: a processing module 101 and a communication module 102. The processing module 101 may be used to control and manage the actions of the flux map generating device 10. The communication module 102 may be used to support the communication between the flux map generating device 10 and other entities. Figure 7 As shown, the flux map generating device 10 may further include a storage module 103 for storing program codes and data of the flux map generating device 10 .

[0132] The processing module 101 may be a processor or a controller. The communication module 102 may be a transceiver, a transceiver circuit or a communication interface, etc. The storage module 103 may be a memory.

[0133] When the processing module 101 is a processor, the communication module 102 is a transceiver, and the storage module 103 is a memory, the processor, the transceiver, and the memory may be connected via a bus. The bus may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc.

[0134] Figure 8 FIG. 1 is a schematic diagram showing a possible structure of a device for generating a treatment plan involved in the above-mentioned embodiment. Figure 8 As shown, the treatment plan generating device 20 may include: a processing module 201 and a communication module 202. The processing module 201 may be used to control and manage the actions of the treatment plan generating device 20. The communication module 202 may be used to support the communication between the treatment plan generating device 20 and other entities. Figure 8 As shown, the treatment plan generating device 20 may further include a storage module 203 for storing program codes and data of the treatment plan generating device 20 .

[0135] The processing module 201 may be a processor or a controller. The communication module 202 may be a transceiver, a transceiver circuit or a communication interface, etc. The storage module 203 may be a memory.

[0136] When the processing module 201 is a processor, the communication module 202 is a transceiver, and the storage module 203 is a memory, the processor, the transceiver, and the memory may be connected via a bus. The bus may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc.

[0137] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0138] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0139] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0140] 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 distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0141] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using a software program, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loading and executing a computer program instruction on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from a website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (Digital Subscriber Line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium may be any available medium that a computer can access or may contain one or more servers, data centers and other data storage devices that can be integrated with a medium. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a DVD), or a semiconductor medium (eg, a solid state disk (SSD)).

[0142] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for generating a flux map, characterized in that: The method comprises: Determining a dose distribution or an initial flux map of a target volume at each of a plurality of field angles, wherein the target volume includes a target area and an organ at risk; Inputting the dose distribution or initial flux map of the target volume at each field angle into a first flux map generation model based on a denoised diffusion probability model to obtain a flux map corresponding to each field angle; The flux map corresponding to each field angle is used to characterize the intensity of each small beam in the target volume corresponding to each field angle.

2. The flux map generation method according to claim 1, characterized in that: The step of determining the dose distribution of the target volume at each of a plurality of field angles comprises: Acquiring image and contour data of the target volume; Inputting the image and contour data of the target volume into a radiation field dose prediction model to obtain a predicted three-dimensional dose distribution of the target volume at each radiation field angle; The predicted three-dimensional dose distribution of the target volume at each field angle is determined as the dose distribution of the target volume at each field angle.

3. The method for generating a flux map according to claim 2, characterized in that: The step of determining the dose distribution of the target volume at each of the plurality of field angles further includes: The predicted three-dimensional dose distribution of the target volume at each field angle is projected onto a projection plane orthogonal to each field angle to obtain the dose distribution of the target volume at each field angle.

4. The method for generating a flux map according to claim 1, characterized in that: Determining an initial flux map of the target volume at each of the plurality of field angles includes: Acquiring image and contour data of the target volume; The image and contour data of the target volume are input into a neural network model for flux map prediction to obtain an initial flux map of the target volume at each of a plurality of field angles.

5. The flux map generation method according to any one of claims 1 to 4, characterized in that: The method further comprises: Acquire structural data of the target volume at each of the multiple field angles; The structural data of the target volume at each field angle is combined with the dose distribution or the initial flux map and input into a second flux map generation model based on a denoising diffusion probability model to obtain a flux map corresponding to each field angle.

6. A method for generating a treatment plan, characterized in that: The method comprises: Acquiring images and contour data of a target volume, wherein the target volume includes a target area and organs at risk; Determining a dose distribution or an initial flux map of the target volume at each of a plurality of field angles according to the image and contour data of the target volume; Inputting the dose distribution or initial flux map of the target volume at each field angle into a first flux map generation model based on a denoised diffusion probability model to obtain a flux map corresponding to each field angle, wherein the flux map corresponding to each field angle is used to characterize the intensity of each small beam in the target volume corresponding to each field angle; A treatment plan is generated according to the flux map corresponding to each field angle.

7. The method for generating a treatment plan according to claim 6, characterized in that: Determining the dose distribution of the target volume at each of a plurality of field angles according to the image and contour data of the target volume comprises: Inputting the image and contour data of the target volume into a field dose prediction model to obtain a predicted three-dimensional dose distribution of the target volume at each field angle; The predicted three-dimensional dose distribution of the target volume at each field angle is determined as the dose distribution of the target volume at each field angle.

8. The method for generating a treatment plan according to claim 7, characterized in that: The step of determining the dose distribution of the target volume at each of a plurality of field angles according to the image and contour data of the target volume further comprises: The predicted three-dimensional dose distribution of the target volume at each field angle is projected onto a projection plane orthogonal to each field angle to obtain the dose distribution of the target volume at each field angle.

9. The method for generating a treatment plan according to claim 6, characterized in that: Determining an initial flux map of the target volume at each of the plurality of field angles according to the image and contour data of the target volume includes: The image and contour data of the target volume are input into a neural network model for flux map prediction to obtain an initial flux map of the target volume at each of a plurality of field angles.

10. The method for generating a treatment plan according to any one of claims 6 to 9, characterized in that: The method further comprises: Acquire structural data of the target volume at each of the multiple field angles; Inputting the structural data of the target volume at each field angle into a second flux map generation model based on a denoised diffusion probability model in combination with the dose distribution or the initial flux map to obtain a flux map corresponding to each field angle; A treatment plan is generated according to the flux map corresponding to each field angle.

11. An electronic device, characterized in that: The electronic device comprises: processor; a memory configured to store instructions executable by the processor; The processor is configured to execute the instructions to implement the flux map generation method according to any one of claims 1 to 5, or to implement the treatment plan generation method according to any one of claims 6 to 10.

12. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions in the computer-readable storage medium are executed by an electronic device, the electronic device is enabled to execute the flux map generating method according to any one of claims 1 to 5, or execute the treatment plan generating method according to any one of claims 6 to 10.

13. A computer program product, characterized in that The computer program product comprises computer instructions, and when the computer instructions are executed on an electronic device, the electronic device executes the flux map generating method according to any one of claims 1 to 5, or executes the treatment plan generating method according to any one of claims 6 to 10.

Images