Radiotherapy plan generation method and device, and storage medium

Abstract

The invention provides a radiotherapy plan generation method and device and a storage medium, and relates to the technical field of medical treatment, in particular to the technical field of radiotherapy. According to the specific implementation scheme, the method comprises the steps of predicting reference dose distribution and a reference flux graph of a to-be-radiotherapy object; taking the reference dose distribution and the reference flux diagram as objective functions, and executing reverse optimization to obtain an optimized flux diagram; and generating a radiotherapy plan based on the flux graph. Therefore, reverse optimization based on the flux diagram is realized, and the reverse optimization process also refers to the reference dose distribution and the reference flux diagram, so that the optimized flux diagram is more suitable for an actual application scene, and the accuracy of a radiotherapy plan obtained according to the optimized flux diagram is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of medical technology, in particular to the field of radiotherapy technology, and specifically to a method, device and storage medium for generating a radiotherapy plan. Background Art

[0002] In the field of medical technology, radiotherapy (abbreviated as radiotherapy) is one of the important means of treating tumors. It achieves the purpose of destroying tumors by giving tumors a sufficiently high radiation dose. In order to ensure that normal tissues are not exposed to excessive doses while giving tumors a high dose, a corresponding radiotherapy plan is usually formulated for the patient based on the medical data of the patient.

[0003] At present, a method based on flux map prediction is usually adopted to formulate a radiotherapy plan for the subject to be radiotherapy, that is, based on the medical data of the subject to be radiotherapy, the flux map of the subject to be radiotherapy is predicted, and then the radiotherapy plan for the subject to be radiotherapy is generated based on the flux map of the subject to be radiotherapy.

[0004] However, in the above technical solution, the accuracy of flux map prediction is low, which may affect the usability of radiotherapy planning. Summary of the invention

[0005] The present disclosure provides a method, device and storage medium for generating a radiotherapy plan, which are used to improve the accuracy of flux map prediction, thereby improving the accuracy of the radiotherapy plan.

[0006] According to one aspect of the present disclosure, a method for generating a radiotherapy plan is provided, the method comprising:

[0007] Predicting reference dose distribution and reference flux map of subjects to be radiotherapy;

[0008] Using the reference dose distribution and the reference flux map as objective functions, performing inverse optimization to obtain an optimized flux map;

[0009] Based on the optimized flux map, a radiotherapy plan is generated.

[0010] In some embodiments, predicting the reference dose distribution of the subject to be radiotherapy includes:

[0011] Acquiring first radiotherapy information of the subject to be radiotherapy; the first radiotherapy information includes radiotherapy subject information and radiotherapy equipment information;

[0012] The first radiotherapy information is input into a dose distribution prediction model, and the first radiotherapy information is processed by the dose distribution prediction model to obtain the reference dose distribution.

[0013] In some embodiments, predicting a reference flux map of the subject to be radiotherapy includes:

[0014] Acquiring second radiotherapy information of the subject to be radiotherapy; the second radiotherapy information includes radiotherapy subject information, radiotherapy equipment information and target area structure information;

[0015] The second radiotherapy information is input into the first flux map prediction model, and the second radiotherapy information is processed by the first flux map prediction model to obtain the reference flux map; or,

[0016] Acquiring third radiotherapy information of the object to be radiotherapy; the third radiotherapy information is the target area structure information;

[0017] The reference dose distribution and the third radiotherapy information are input into a second flux map prediction model, and the reference dose distribution and the third radiotherapy information are processed by the second flux map prediction model to obtain the reference flux map.

[0018] In some embodiments, before inputting the first radiotherapy information into a dose distribution prediction model and processing the first radiotherapy information by the dose distribution prediction model to obtain the reference dose distribution, the method further includes:

[0019] A dose distribution prediction model matching the case type of the object to be radiotherapy is selected from a plurality of candidate dose distribution prediction models; the dose distribution prediction model matching the case type of the object to be radiotherapy is trained based on first radiotherapy information of the case type.

[0020] In some embodiments, the reference dose distribution and the reference flux map are used as objective functions to perform inverse optimization to obtain an optimized flux map, including:

[0021] Using the reference flux map as the objective function, performing inverse optimization to obtain a first flux map; constructing a reference dose volume histogram based on the reference dose volume histogram; using the reference dose volume histogram as the objective function, performing inverse optimization on the first flux map to obtain the optimized flux map; or,

[0022] According to the reference dose distribution, a reference dose volume histogram is constructed; the reference flux map and the reference dose volume histogram are used as objective functions, and inverse optimization is performed to obtain the optimized flux map.

[0023] In some embodiments, the reference flux map is used as an objective function to perform inverse optimization to obtain a first flux map, including:

[0024] In any iteration process, the reference flux map is used as the objective function to adjust the flux map adjusted in the previous iteration process;

[0025] In response to the current iteration process satisfying the first preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the first flux map.

[0026] In some embodiments, the reference dose volume histogram is used as an objective function, and inverse optimization is performed on the first flux map to obtain the optimized flux map, including:

[0027] In any iteration process, the reference dose volume histogram is used as the objective function to adjust the flux map adjusted in the previous iteration process;

[0028] In response to the current iteration process satisfying the second preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the optimized flux map.

[0029] In some embodiments, the reference flux map and the reference dose volume histogram are used as objective functions to perform inverse optimization to obtain the optimized flux map, including:

[0030] In any iteration process, the reference flux map and the reference dose volume histogram are used as objective functions to adjust the flux map adjusted in the previous iteration process;

[0031] In response to the current iteration process satisfying the third preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the optimized flux map.

[0032] According to another aspect of the present disclosure, there is provided an electronic device, comprising:

[0033] at least one processor; and

[0034] a memory communicatively connected to the at least one processor; wherein,

[0035] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for generating a radiotherapy plan provided by the present disclosure.

[0036] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the electronic device to execute the method for generating a radiotherapy plan provided by the present disclosure.

[0037] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, which, when executed by a processor, implements the method for generating a radiotherapy plan provided by the present disclosure.

[0038] The technical solution provided by the present disclosure predicts the reference dose distribution and reference flux map of the object to be radiotherapy, and uses the predicted reference dose distribution and reference flux map as the objective function of plan optimization to perform inverse optimization, which not only realizes the inverse optimization based on the flux map, but also refers to the reference dose distribution and reference flux map in the inverse optimization process, so that the optimized flux map is more in line with the actual application scenario. In addition, the present disclosure also introduces artificial intelligence technology into the plan optimization process. Through artificial intelligence technology, optimization experience can be learned based on historical plans, and then the radiotherapy plan can be adjusted based on the optimization results by using inverse optimization. In this way, a method for generating a radiotherapy plan based on artificial intelligence guidance is provided, which not only realizes the rapid optimization of the radiotherapy plan, but also helps to improve the accuracy of the radiotherapy plan obtained according to the optimization results, greatly improves the efficiency of clinical planning, and reduces the workload and difficulty of medical personnel in completing the radiotherapy plan.

[0039] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0041] Figure 1 is a schematic diagram of an implementation environment of a method for generating a radiotherapy plan according to an embodiment of the present disclosure;

[0042] Figure 2 is a flow chart of a method for generating a radiotherapy plan according to an embodiment of the present disclosure;

[0043] Figure 3 is a flow chart of a method for generating a radiotherapy plan according to an embodiment of the present disclosure;

[0044] Figure 4 is a schematic diagram of a plan optimization process shown in an embodiment of the present disclosure;

[0045] Figure 5 is a structural block diagram of a radiotherapy plan generation device shown in an embodiment of the present disclosure;

[0046] Figure 6 The block diagram of an electronic device used to implement the method for generating a radiotherapy plan according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0047] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0048] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0049] First, the application scenarios involved in the embodiments of the present disclosure are described. The method for generating a radiotherapy plan provided in the embodiments of the present disclosure can be applied in the field of medical technology, and specifically can be applied in the scenario of radiotherapy (abbreviated as radiotherapy).

[0050] In the field of medical technology, radiotherapy is one of the important means of treating tumors. It achieves the purpose of eliminating tumors by giving tumors a sufficiently high radiation dose. However, since radiation cannot distinguish between tumor cells and normal tissues, normal tissues will also be irradiated to varying degrees while giving tumors a high dose. It should be noted that the intensity distribution of the radiation emitted by the radiotherapy equipment used in radiotherapy is uniform, while the tumor is a three-dimensional structure, and some parts of the tumor are thick and some parts are thin. If you want each part of the tumor to receive the same dose, you need to adjust the intensity distribution of the radiation in the irradiation field, that is, the intensity of the radiation is smaller for the thin part and higher for the thick part. In this way, the use of a scheme with adjusted intensity distribution can achieve the purpose of giving tumors a high dose without causing excessive doses to normal tissues.

[0051] At present, when adjusting the intensity distribution of radiation within the irradiation field, the intensity modulated radiation therapy (IMRT) technology is usually used. Among them, the intensity modulated radiation therapy technology is currently a commonly used technology in the radiotherapy of tumors. Specifically, it adjusts the intensity of the radiation to make the dose of the target area (that is, the tumor site that needs to be treated) equal, or adjusts the intensity of the dose within the irradiation volume according to the treatment needs to ensure that the surrounding normal tissues are irradiated with a smaller dose of radiation.

[0052] It should be noted that IMRT requires medical personnel (such as physicians, dosimetrists, physicists, etc.) to formulate a corresponding radiotherapy plan for the patient based on the patient's medical data (such as tumor shape or tumor location, etc.). Therefore, in order to ensure that a high dose is given to the tumor while normal tissue is not exposed to excessive doses, a corresponding radiotherapy plan is usually formulated based on the patient's medical data.

[0053] In the related art, a method based on flux map prediction is usually adopted to formulate a radiotherapy plan for the subject to be radiotherapy, that is, based on the medical data of the subject to be radiotherapy, the flux map of the subject to be radiotherapy is predicted, and then the radiotherapy plan for the subject to be radiotherapy is generated based on the flux map of the subject to be radiotherapy. However, in this technical solution, the accuracy of the flux map prediction is low, which may affect the availability of the radiotherapy plan.

[0054] Based on this, the embodiment of the present disclosure provides a method for generating a radiotherapy plan, which predicts the reference dose distribution and reference flux map of the object to be radiotherapy, and uses the predicted reference dose distribution and reference flux map as the objective function of plan optimization to perform inverse optimization, which not only realizes the inverse optimization based on the flux map, but also refers to the reference dose distribution and reference flux map in the inverse optimization process, so that the optimized flux map is more in line with the actual application scenario. In addition, the present disclosure also introduces artificial intelligence technology into the plan optimization process, through which optimization experience can be learned based on historical plans, and then the radiotherapy plan can be adjusted based on the optimization results by using inverse optimization. In this way, a method for generating a radiotherapy plan based on artificial intelligence guidance is provided, which not only realizes the rapid optimization of the radiotherapy plan, but also helps to improve the accuracy of the radiotherapy plan obtained according to the optimization results, greatly improves the efficiency of clinical planning, and reduces the workload and difficulty of medical personnel in completing the radiotherapy plan.

[0055] Figure 1 is a schematic diagram of an implementation environment of a method for generating a radiotherapy plan shown in an embodiment of the present disclosure. Figure 1 The implementation environment includes an image scanning device 101, a radiotherapy planning device 102 and a radiotherapy device 103.

[0056] The image scanning device 101 is a device for scanning and displaying the tumor site and surrounding normal tissue of the subject to be radiotherapy. In some embodiments, the image scanning device 101 can be at least one of a computed tomography (CT) device, an emission computed tomography (ECT) device, a magnetic resonance imaging (MRI) device, a positron emission tomography (PET) device, and an ultrasound examination device.

[0057] In the disclosed embodiment, the image scanning device 101 is used to obtain a medical scan image of the object to be radiotherapy, and upload the medical scan image of the object to be radiotherapy to the radiotherapy planning device 102, so that the radiotherapy planning device 102 performs a subsequent radiotherapy plan generation process based on the medical scan image of the object to be radiotherapy.

[0058] The radiotherapy planning device 102 is a device for generating a radiotherapy plan. In some embodiments, the radiotherapy planning device 102 may be at least one of a smart phone, a smart watch, a desktop computer, a laptop computer, a virtual reality terminal, an augmented reality terminal, a wireless terminal, and a laptop computer. Further, in some embodiments, the radiotherapy planning device 102 may run a radiotherapy planning system (TPS), which provides a function of generating a radiotherapy plan.

[0059] In the disclosed embodiment, the radiotherapy planning device 102 is used to predict the reference dose distribution and reference flux map of the object to be radiotherapy; the reference dose distribution and the reference flux map are used as objective functions, and inverse optimization is performed to obtain an optimized flux map; based on the optimized flux map, a radiotherapy plan is generated.

[0060] The radiotherapy device 103 is a device used for radiotherapy. In some embodiments, the radiotherapy device 103 can be at least one of a gamma knife, a linear accelerator, a neutron knife, an X-ray therapy machine, and the like.

[0061] In the disclosed embodiment, the radiotherapy device 103 is used to receive the radiotherapy plan from the radiotherapy planning device 102, and perform radiotherapy on the subject to be treated according to the radiotherapy plan.

[0062] In some embodiments, the implementation environment further includes a server 104. In some embodiments, the server 104 is used to provide background communication services for the above-mentioned image scanning device 101, radiotherapy planning device 102 and radiotherapy device 103.

[0063] Among them, the server 104 can be an independent physical server, or a server cluster or distributed file system composed of multiple physical servers, or at least one of the cloud servers that provide basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content distribution networks, and big data or artificial intelligence platforms, etc., which is not limited in the embodiments of the present disclosure. In some embodiments, the number of the above-mentioned servers 104 can be more or less, which is not limited in the embodiments of the present disclosure. Of course, the server 104 can also include other functions to provide more comprehensive and diversified services.

[0064] The following is based on Figure 1 The implementation environment shown is used to introduce the method provided by the embodiment of the present disclosure.

[0065] Figure 2 FIG. 1 is a flow chart of a method for generating a radiotherapy plan shown in an embodiment of the present disclosure. In some embodiments, the method for generating a radiotherapy plan is executed by an electronic device. For example, the electronic device may be the above-mentioned Figure 1 The radiotherapy planning equipment shown. Figure 2 As shown, the method includes the following steps.

[0066] S201. Predicting a reference dose distribution and a reference flux map of a subject to be radiotherapy.

[0067] The reference dose distribution may be a three-dimensional dose distribution, which is used to characterize the dose distribution of different parts of the target area of ​​the subject to be radiotherapy. In the disclosed embodiment, the reference dose distribution is used as an optimization target for subsequent inverse optimization to achieve inverse optimization of the radiotherapy plan.

[0068] The reference flux map is used to characterize the intensity distribution of radiation emitted by the radiotherapy device at different positions (x, y) in the target area of ​​the subject to be radiotherapy. In the disclosed embodiment, the reference flux map is used as the optimization target of subsequent inverse optimization to achieve inverse optimization of the radiotherapy plan.

[0069] In the disclosed embodiment, the predicted reference dose distribution and reference flux map are used to perform subsequent inverse optimization, so that a flux map with higher accuracy can be obtained, thereby improving the accuracy of generating a radiotherapy plan.

[0070] S202 , taking the reference dose distribution and the reference flux map as objective functions, performing inverse optimization to obtain an optimized flux map.

[0071] Among them, inverse optimization refers to the process of optimizing the corresponding flux map by continuously adjusting the constraints and corresponding weights and other parameters to achieve the optimization goal. For example, taking the target value as the dose distribution as an example, inverse optimization is to continuously adjust the constraints and corresponding weights of the dose distribution to optimize the corresponding flux map.

[0072] In some embodiments, the above-mentioned inverse optimization process may be: taking the reference flux map as the objective function, performing inverse optimization to obtain a first flux map, constructing a reference dose volume histogram based on the reference dose distribution, taking the reference dose volume histogram as the objective function, performing inverse optimization on the first flux map, and obtaining the optimized flux map.

[0073] In the above embodiment, two stages of reverse optimization are implemented. In the first stage of reverse optimization, the reference flux map is used as the objective function to quickly optimize the current radiotherapy plan, so that the first flux map after reverse optimization is closer to the reference flux map. Then, the reference dose volume histogram constructed based on the reference dose distribution is used as the objective function, and reverse optimization is continued on the basis of the optimization results of the first stage to obtain the final flux map. In this way, a flux map with high accuracy can be obtained, so that the optimized radiotherapy plan can meet the established clinical requirements, thereby improving the accuracy of generating the radiotherapy plan.

[0074] Alternatively, in other embodiments, the above-mentioned inverse optimization process may be: constructing a reference dose volume histogram based on the reference dose distribution, taking the reference flux map and the reference dose volume histogram as objective functions, performing inverse optimization, and obtaining the optimized flux map.

[0075] In the above embodiment, the reference flux map and the reference dose volume histogram constructed based on the reference dose distribution are used as objective functions to quickly optimize the current radiotherapy plan, so that a flux map with higher accuracy can be obtained, so that the optimized radiotherapy plan can meet the established clinical requirements, thereby improving the accuracy of generating the radiotherapy plan.

[0076] S203: Generate a radiotherapy plan based on the optimized flux map.

[0077] Among them, the radiotherapy plan is a document used to control the radiation therapy equipment to deliver radiation.

[0078] The technical solution provided by the disclosed embodiment performs inverse optimization by predicting the reference dose distribution and reference flux map of the object to be radiotherapy, and using the predicted reference dose distribution and reference flux map as the objective function of plan optimization. This not only realizes the inverse optimization based on the flux map, but the inverse optimization process also refers to the reference dose distribution and reference flux map, so that the optimized flux map is more in line with the actual application scenario, which is conducive to improving the accuracy of the radiotherapy plan obtained according to the optimized flux map.

[0079] Above Figure 2 This is a simple embodiment of the present disclosure. The following describes a method for generating a radiotherapy plan provided by the present disclosure based on a specific embodiment. Figure 3 FIG. 1 is a flow chart of a method for generating a radiotherapy plan shown in an embodiment of the present disclosure. In some embodiments, the method for generating a radiotherapy plan is executed by an electronic device. For example, the electronic device may be the above-mentioned Figure 1 The radiotherapy planning equipment shown. Figure 3 As shown, with the radiotherapy planning device as the execution body, the method includes the following steps.

[0080] S301. The radiotherapy planning device predicts the reference dose distribution of the subject to be radiotherapy.

[0081] The reference dose distribution may be a three-dimensional dose distribution, which is used to characterize the dose distribution of different parts of the target area of ​​the subject to be radiotherapy. Exemplarily, the reference dose distribution may include the dose distribution corresponding to each irradiation field of the radiotherapy device.

[0082] For example, the reference dose distribution may be in the form of an image, a table, or a curve. For example, in the form of an image, the reference dose distribution may be a three-dimensional dose distribution diagram. The embodiment of the present disclosure does not limit the form of expression of the reference dose distribution.

[0083] In the disclosed embodiment, the reference dose distribution is used as an optimization target for subsequent inverse optimization to achieve inverse optimization of the radiotherapy plan.

[0084] In some embodiments, the prediction process of the above reference dose distribution can refer to S301A to S301C described below.

[0085] S301A: The radiotherapy planning device obtains first radiotherapy information of the subject to be radiotherapy.

[0086] The first radiotherapy information includes radiotherapy object information and radiotherapy equipment information. In the disclosed embodiment, the first radiotherapy information is information referenced when predicting dose distribution.

[0087] In some embodiments, the radiotherapy object information may be a medical scan image of the object to be radiotherapy, a delineated image of the target area or organ, etc. The medical scan image is used to present the tumor site and surrounding normal tissue of the object to be radiotherapy in the form of an image. Exemplarily, the medical scan image may include at least one of a CT scan image, a nuclear magnetic resonance scan image, or an ultrasonic scan image.

[0088] In some embodiments, the radiotherapy equipment information may be the radiation source type, field configuration parameters, etc. of the radiotherapy equipment. The radiation source type may include the type, shape, position, direction, etc. of the radiation source. The field configuration parameters may include the field angle, field shape, field weight, number of beams, etc.

[0089] In some embodiments, the first radiotherapy information of the subject to be radiotherapy may be uploaded to the radiotherapy planning device by a medical staff. For example, the corresponding process may be: the radiotherapy planning device receives the upload operation of the first radiotherapy information by the medical staff, and obtains the first radiotherapy information of the subject to be radiotherapy in response to the upload operation of the first radiotherapy information.

[0090] The uploading operation of the first radiotherapy information may be at least one of an input operation, a selection operation or other types of operations. The embodiment of the present disclosure does not limit the specific implementation of the uploading operation.

[0091] Exemplarily, taking the input operation as an example, the operation panel of the radiotherapy planning device may display a first radiotherapy information input box, and the medical staff may enter the first radiotherapy information in the first radiotherapy information input box, thus completing the uploading of the first radiotherapy information. Alternatively, exemplarily, taking the selection operation as an example, the operation panel of the radiotherapy planning device may display multiple options for the first radiotherapy information, and the medical staff may select the corresponding first radiotherapy information from the multiple options for the first radiotherapy information, thus completing the uploading of the first radiotherapy information of the subject to be treated.

[0092] It is worth noting that, in other embodiments, the radiotherapy planning device may also adopt other implementations to obtain the first radiotherapy information. The embodiment of the present disclosure does not limit the process of obtaining the first radiotherapy information.

[0093] Regarding the medical scan image of the object to be radiotherapy included in the first radiotherapy information, the process of the radiotherapy planning device acquiring the medical scan image of the object to be radiotherapy may refer to the following (1-1) or (1-2).

[0094] (1-1) In a possible implementation, after scanning the object to be radiotherapy and generating a medical scan image of the object to be radiotherapy, the image scanning device sends the medical scan image of the object to be radiotherapy to the radiotherapy planning device. Accordingly, the radiotherapy planning device receives the medical scan image from the image scanning device, that is, obtains the medical scan image of the object to be radiotherapy.

[0095] (1-2) In another possible implementation, after scanning the object to be radiotherapy and generating the medical scan image of the object to be radiotherapy, the image scanning device outputs the medical scan image of the object to be radiotherapy. Exemplarily, the medical scan image outputted at this time may be in the form of a data packet file or in the form of a physical image. Then, the medical staff uploads the medical scan image outputted by the image scanning device to the radiotherapy planning device, and the radiotherapy planning device obtains the medical scan image of the object to be radiotherapy.

[0096] In the above embodiments, two implementation methods are provided for the radiotherapy planning device to acquire the medical scan image of the object to be radiotherapy, thereby improving the flexibility of acquiring the medical scan image.

[0097] It is worth noting that in other embodiments, the radiotherapy planning device may also adopt other implementations to obtain the medical scan image of the object to be radiotherapy. The disclosed embodiment does not limit the process of obtaining the medical scan image.

[0098] S301B. The radiotherapy planning device selects a dose distribution prediction model that matches the case type of the subject to be radiotherapy from a plurality of candidate dose distribution prediction models.

[0099] The dose distribution prediction model provides a dose distribution prediction function. In some embodiments, the dose distribution prediction model may be a three-dimensional dose distribution prediction model, which provides a three-dimensional dose distribution prediction function.

[0100] The multiple candidate dose distribution prediction models include dose distribution prediction models corresponding to different case types. The case type is a classification obtained by dividing the tumor according to normal tissue or a cell type similar thereto. Exemplarily, the case type may be gastric cancer, hepatocellular carcinoma, cholangiocarcinoma, adenocarcinoma, squamous cell carcinoma or osteosarcoma.

[0101] In the disclosed embodiment, the dose distribution prediction model that matches the case type of the object to be radiotherapy is trained based on the first radiotherapy information of the case type. It should be noted that before implementing this solution, the first radiotherapy information of the object to be radiotherapy of different case types (that is, the training data required for training the dose distribution prediction model) can be used to pre-train the dose distribution prediction models corresponding to the different case types, and the trained dose distribution prediction models corresponding to the different case types are maintained in the radiotherapy planning device.

[0102] In some embodiments, the initial model is iteratively trained using first radiotherapy information of different case types to obtain a dose distribution prediction model.

[0103] Accordingly, the above-mentioned iterative training process includes: during any iterative training process, the first radiotherapy information of the case type is input into the model obtained after the previous iterative training to obtain the output result of this iterative training, and based on the output result and the sample label of the case type, the model parameters are adjusted, and based on the adjusted model parameters, the next iterative training is performed.

[0104] The sample label is used to indicate the dose distribution of the case type. In some embodiments, the process of adjusting the model parameters is: based on the output result and the sample label of the case type, determining the model loss value of this iterative training, and adjusting the model parameters according to the model loss value.

[0105] The model loss value is used to represent the difference between the output result of the model and the sample label. In some embodiments, the model loss value may be a cross entropy loss value (Cross Entropy Loss) or a mean square error loss value (MeanSquare Error, MSE). In this way, by determining the model loss value, since the model loss value is used to represent the difference between the output result of the model and the sample label, the model parameters are adjusted according to the model loss value, which can improve the learning ability of the model, thereby training a model with better learning ability. It is worth noting that in another possible implementation method, other types of model loss values ​​can also be obtained to perform the above-mentioned process of adjusting the model parameters according to the model loss value. This is not limited to the embodiments of the present disclosure.

[0106] In one possible implementation, after adjusting the model parameters, it is also determined whether the model training has reached the target conditions. If the model training has not reached the target conditions, the next iterative training is performed based on the adjusted model parameters until the model training reaches the target conditions.

[0107] Exemplarily, the target condition satisfies at least one of the following conditions: the number of iterations of model training reaches the target number; or the model loss value is less than or equal to the target threshold. The target number is a pre-set number of training iterations, such as when the number of iterations reaches 100. The present disclosure embodiment does not limit the setting of the target number. The target threshold is a pre-set fixed threshold, such as when the model loss value is less than 0.0001. The present disclosure embodiment does not limit the setting of the target threshold.

[0108] In some embodiments, the process of selecting a dose distribution prediction model that matches the case type of the subject to be radiotherapy may refer to any one of the following (2-1) to (2-3).

[0109] (2-1) In a possible implementation, a medical staff performs a setting operation on the case type of the object to be radiotherapy in the operation panel of the radiotherapy planning device, and the radiotherapy planning device receives the setting operation on the case type of the object to be radiotherapy by the medical staff, and in response to the setting operation on the case type of the object to be radiotherapy, obtains the set case type of the object to be radiotherapy. Then, according to the case type of the object to be radiotherapy, a dose distribution prediction model matching the case type of the object to be radiotherapy is selected from a plurality of candidate dose distribution prediction models.

[0110] The setting operation for the case type of the subject to be treated with radiotherapy may be any one of an input operation, a selection operation or other types of operations. The embodiment of the present disclosure does not limit the specific implementation of the setting operation.

[0111] It is worth noting that in other embodiments, other implementations may be used to set the case type of the subject to be radiotherapy. The embodiment of the present disclosure does not limit the method for setting the case type.

[0112] (2-2) In another possible implementation, the operation panel of the radiotherapy planning device can display the identification of multiple candidate dose distribution prediction models, and the medical staff can perform an operation of selecting the identification of the dose distribution prediction model in the operation panel of the radiotherapy planning device. The radiotherapy planning device then receives the medical staff's operation of selecting the identification of the dose distribution prediction model, and in response to the operation of selecting the identification of the dose distribution prediction model, determines the dose distribution prediction model corresponding to the selected identification as the dose distribution prediction model that matches the case type of the object to be radiotherapy.

[0113] The identifier of the dose distribution prediction model may be any one of a name, a number or an identity document (ID), which is not limited in the embodiments of the present disclosure.

[0114] (2-3) In another possible implementation, after the radiotherapy planning device acquires the first radiotherapy information of the object to be radiotherapy based on S301 A, the device determines the case type of the object to be radiotherapy based on the first radiotherapy information of the object to be radiotherapy. Further, based on the case type of the object to be radiotherapy, a dose distribution prediction model matching the case type of the object to be radiotherapy is selected from a plurality of candidate dose distribution prediction models.

[0115] In the above embodiments, three implementation methods are provided for the radiotherapy planning device to select the dose distribution prediction model corresponding to the object to be radiotherapy, thereby improving the flexibility of selecting the dose distribution prediction model.

[0116] It is worth noting that in other embodiments, the radiotherapy planning device may also adopt other implementations to select a dose distribution prediction model that matches the case type of the object to be radiotherapy. The embodiments of the present disclosure are not limited to this.

[0117] The disclosed embodiment takes the example of selecting a dose distribution prediction model that matches the case type of the object to be radiotherapy based on S301B, and then predicting the reference dose distribution based on S301C using the selected dose distribution prediction model to illustrate the scheme. It should be noted that the above S301B is an optional step. In other embodiments, after the radiotherapy planning device obtains the first radiotherapy information of the object to be radiotherapy based on S301A, there is no need to execute S301B, and S301C is executed using the preset dose distribution prediction model.

[0118] S301C: The radiotherapy planning device inputs the first radiotherapy information into a dose distribution prediction model, processes the first radiotherapy information through the dose distribution prediction model, and obtains a reference dose distribution.

[0119] In some embodiments, the radiotherapy planning device calls a dose prediction engine, inputs the first radiotherapy information into a dose distribution prediction model, processes the first radiotherapy information through the dose distribution prediction model, and obtains the reference dose distribution.

[0120] In some embodiments, the radiotherapy planning device inputs the first radiotherapy information into a dose distribution prediction model, and extracts features of the first radiotherapy information through a feature extraction layer of the dose distribution prediction model to obtain basic features after feature extraction. The basic features are convolved through a convolution layer of the dose distribution prediction model to obtain target features after convolution. The target features are processed through a fully connected layer of the dose distribution prediction model to obtain the reference dose distribution.

[0121] In the above embodiment, the dose distribution prediction model is used to predict the reference dose distribution of the object to be radiotherapy, which can effectively improve the prediction accuracy of the reference dose distribution. Furthermore, the predicted reference dose distribution is subsequently used as the objective function of the plan optimization, which can further improve the accuracy of the plan optimization, thereby improving the accuracy of the generated radiotherapy plan.

[0122] It should be noted that the above S301A to S301C illustrate the scheme by taking the radiotherapy planning device using the dose distribution prediction model to predict the reference dose distribution as an example. In other embodiments, the radiotherapy planning device may also use other implementation methods to predict the reference dose distribution. For example, a Monte Carlo calculation model may be used to predict the reference dose distribution.

[0123] S302: The radiotherapy planning device predicts a reference flux map of the subject to be radiotherapy.

[0124] The reference flux map is used to characterize the intensity distribution of radiation emitted by the radiotherapy device at different positions (x, y) in the target area of ​​the subject to be radiotherapy. In some embodiments, the reference flux map may include a flux map corresponding to each irradiation field of the radiotherapy device.

[0125] In the disclosed embodiment, the reference flux map is used as an optimization target for subsequent inverse optimization to achieve inverse optimization of the radiotherapy plan.

[0126] In some embodiments, the prediction process of the reference flux diagram may refer to the contents shown in S302A to S302B below, or the contents shown in S302C to S302D below.

[0127] S302A: The radiotherapy planning device obtains second radiotherapy information of the subject to be radiotherapy.

[0128] The second radiotherapy information includes radiotherapy object information, radiotherapy equipment information and target area structure information. In the disclosed embodiment, the second radiotherapy information is the information referenced when predicting the flux map. In some embodiments, the target area structure information may be the target area contour, target area thickness, target area depth, etc.

[0129] It should be noted that the process of the radiotherapy planning device acquiring the second radiotherapy information of the object to be radiotherapy is described in S301A above and will not be described in detail.

[0130] S302B: The radiotherapy planning device inputs the second radiotherapy information into the first flux map prediction model, processes the second radiotherapy information through the first flux map prediction model, and obtains a reference flux map.

[0131] The first flux map prediction model provides a flux map prediction function.

[0132] In some embodiments, the radiotherapy planning device calls a flux map prediction engine, inputs the second radiotherapy information into a first flux map prediction model, processes the second radiotherapy information through the first flux map prediction model, and obtains a reference flux map.

[0133] In some embodiments, the radiotherapy planning device inputs the second radiotherapy information into the first flux map prediction model, and extracts features of the second radiotherapy information through the feature extraction layer of the first flux map prediction model to obtain basic features after feature extraction. The basic features are convolved through the convolution layer of the first flux map prediction model to obtain target features after convolution. The target features are processed through the fully connected layer of the first flux map prediction model to obtain the reference flux map.

[0134] In the above embodiment, the first flux map prediction model is used to predict the reference flux map of the object to be radiotherapy, which can effectively improve the prediction accuracy of the reference flux map. Then, the predicted reference flux map is subsequently used as the objective function of the plan optimization, which can further improve the accuracy of the plan optimization, thereby improving the accuracy of the generated radiotherapy plan.

[0135] The above S302A to S302B take the second radiotherapy information as input and use the first flux map prediction model to predict the reference flux map of the object to be radiotherapy as an example to illustrate the scheme. In other embodiments, the prediction process of the above reference flux map can also refer to the following S302C to S302D.

[0136] S302C: The radiotherapy planning device obtains the third radiotherapy information of the subject to be radiotherapy.

[0137] The third radiotherapy information is target area structure information. In the disclosed embodiment, the third radiotherapy information is information referenced when predicting the flux map in combination with the reference dose distribution.

[0138] It should be noted that the process of the radiotherapy planning device acquiring the third radiotherapy information of the subject to be radiotherapy is described in S301A above and will not be described in detail.

[0139] S302D: The radiotherapy planning device inputs the reference dose distribution and the third radiotherapy information into a second flux map prediction model, processes the reference dose distribution and the third radiotherapy information through the second flux map prediction model, and obtains the reference flux map.

[0140] The second flux map prediction model provides a function of predicting the flux map in combination with the reference dose distribution.

[0141] In some embodiments, the radiotherapy planning device calls a flux map prediction engine, inputs the reference dose distribution and the third radiotherapy information into a second flux map prediction model, processes the reference dose distribution and the third radiotherapy information through the second flux map prediction model, and obtains a reference flux map.

[0142] In some embodiments, the radiotherapy planning device inputs the reference dose distribution and the third radiotherapy information into the second flux map prediction model, and extracts features from the reference dose distribution and the third radiotherapy information through the feature extraction layer of the second flux map prediction model to obtain basic features after feature extraction. The basic features are convolved through the convolution layer of the second flux map prediction model to obtain target features after convolution. The target features are processed through the fully connected layer of the second flux map prediction model to obtain the reference flux map.

[0143] In the above embodiment, the second flux map prediction model is used in combination with the reference dose distribution and the third radiotherapy information to predict the reference flux map of the object to be radiotherapy, which can effectively improve the prediction accuracy of the reference flux map. Then, the predicted reference flux map is subsequently used as the objective function of the plan optimization, which can further improve the accuracy of the plan optimization, thereby improving the accuracy of the generated radiotherapy plan.

[0144] It should be noted that the above S302A to S302B and S302C to S302D all take the radiotherapy planning device using the flux map prediction model to predict the reference flux map as an example to illustrate the scheme. Among them, the training process of the flux map prediction model is similar to the training process of the dose distribution prediction model, which will not be repeated. In other embodiments, the radiotherapy planning device can also use other implementation methods to predict the reference flux map. Exemplarily, the reference flux map can be obtained by performing deconvolution processing based on the reference dose distribution of the object to be radiotherapy.

[0145] Through the above S301 to S302, the process of using the artificial intelligence model to predict the reference dose distribution and reference flux map of the object to be radiotherapy is described. Then, by using the predicted reference dose distribution and reference flux map to perform subsequent reverse optimization, a flux map with higher accuracy can be obtained, thereby improving the accuracy of generating a radiotherapy plan.

[0146] In the disclosed embodiment, there is provided a method based on artificial intelligence technology to realize automatic generation of radiotherapy plans, which combines the advantages of artificial intelligence technology and traditional optimization methods. Among them, based on artificial intelligence technology, it is possible to learn optimization experience based on historical plans, and then use traditional optimization methods to adjust the plan based on the optimization results, introduce manual planning experience into the plan optimization process, realize rapid optimization of the plan, and at the same time, provide medical personnel with a large space for plan adjustment.

[0147] In the related art, a radiotherapy plan may also be formulated based on a three-dimensional dose distribution prediction method or a dose and volume histogram (DVH) prediction method.

[0148] Among them, the method based on three-dimensional dose distribution prediction is mainly: predicting the three-dimensional dose distribution of the object to be radiotherapy through the medical data of the object to be radiotherapy, and then determining the dose-volume histogram of the object to be radiotherapy based on the predicted three-dimensional dose distribution, taking the dose-volume histogram as the initial state of plan optimization, and using traditional optimization algorithms to optimize the plan. The method based on dose-volume histogram prediction is mainly: predicting the dose-volume histogram of the object to be radiotherapy through the medical data of the object to be radiotherapy, taking the dose-volume histogram as the initial state of plan optimization, and using traditional optimization algorithms to optimize the plan.

[0149] Among the methods of formulating radiotherapy plans in related technologies, the method based on three-dimensional dose distribution prediction and the method based on dose-volume histogram prediction, converting the predicted three-dimensional dose distribution or dose-volume histogram into an executable radiotherapy plan still needs to rely on the medical staff's planning design skills, planning design experience and multiple iterative reverse optimization processes, which takes a considerable amount of time to complete the formulation of the radiotherapy plan.

[0150] It can be seen that the solution provided by the embodiments of the present disclosure can overcome the problem in the related art that the optimization time based on the traditional optimization method is long and the planning is time-consuming, and can also overcome the problem that the prediction results based on the flux map prediction method cannot be further optimized, which greatly improves the efficiency of clinical planning and reduces labor costs.

[0151] S303: The radiotherapy planning device uses the reference dose distribution and the reference flux map as objective functions, performs inverse optimization, and obtains an optimized flux map.

[0152] Among them, inverse optimization refers to the process of optimizing the corresponding flux map by continuously adjusting the constraints and corresponding weights and other parameters to achieve the optimization goal. For example, taking the target value as the dose distribution as an example, inverse optimization is to continuously adjust the constraints and corresponding weights of the dose distribution to optimize the corresponding flux map.

[0153] In some embodiments, the process of performing inverse optimization by the radiotherapy planning device in the above S303 can refer to the contents shown in the following S303A to S303C, or the contents shown in the following S302D to S302E.

[0154] S303A: The radiotherapy planning device uses the reference flux map as an objective function, performs inverse optimization, and obtains a first flux map.

[0155] The first flux map is used to refer to the flux map obtained after one stage of reverse optimization.

[0156] In some embodiments, during any iteration process, the radiotherapy planning device uses the reference flux map as the objective function to adjust the flux map adjusted in the previous iteration process. In response to the current iteration process satisfying the first preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the first flux map.

[0157] The first preset condition is used to determine whether the iteration process in S303A satisfies the condition for ending the iteration. It should be understood that if the iteration process satisfies the first preset condition, it means that the iteration process satisfies the condition for ending the iteration, and the iterative optimization is stopped; if the iteration process does not satisfy the first preset condition, it means that the iteration process does not satisfy the condition for ending the iteration, and the iterative optimization is continued.

[0158] In some embodiments, the first preset condition may include at least one of the following conditions (3-1) to (3-2).

[0159] (3-1) The first mean square error is less than the first preset threshold.

[0160] The first mean square error represents the difference between the adjusted flux map and the reference flux map. The first preset threshold is a preset error threshold, such as 0.1, 0.01 or other thresholds. The embodiment of the present disclosure does not limit the setting of the first preset threshold.

[0161] (3-2) The number of iterative optimizations is greater than or equal to the preset number of iterations.

[0162] The number of iterations refers to the number of iterations experienced during the optimization process. The preset number of iterations is a preset number of iterations, such as 100, 1000 or other values. The present disclosure does not limit the setting of the preset number of iterations.

[0163] In the above embodiment, the first preset condition is described by taking the conditions shown in (3-1) to (3-2) as an example. It should be noted that the radiotherapy planning device can determine whether the iterative process meets the conditions for the end of the iteration based on one or two of the conditions shown in (3-1) to (3-2). In this way, the types of the first preset conditions are enriched and the flexibility of the inverse optimization is improved.

[0164] S303B: The radiotherapy planning device constructs a reference dose volume histogram according to the reference dose distribution.

[0165] The reference dose volume histogram may be a dose volume histogram, which is used to represent the irradiation amount in the target area and the surrounding normal tissue volume. In some embodiments, the dose volume histogram represents the relationship between dose and volume, such as using the horizontal axis to represent the dose and the vertical axis to represent the volume (percentage volume). In this way, by referring to the dose volume histogram, it is possible to intuitively display how much volume exists in the target area and the normal tissue volume and how high the dose is irradiated.

[0166] In some embodiments, the radiotherapy planning device reconstructs a reference dose volume histogram based on the reference dose distribution using a DVH reconstruction algorithm.

[0167] S303C: The radiotherapy planning device uses the reference dose volume histogram as an objective function, performs inverse optimization on the first flux map, and obtains the optimized flux map.

[0168] In some embodiments, during any iteration process, the radiotherapy planning device uses the reference dose volume histogram as the objective function to adjust the flux map adjusted in the previous iteration process. In response to the current iteration process satisfying the second preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the optimized flux map.

[0169] The second preset condition is used to determine whether the iteration process in S303C satisfies the condition for ending the iteration. It should be understood that if the iteration process satisfies the second preset condition, it means that the iteration process satisfies the condition for ending the iteration, and the iterative optimization is stopped; if the iteration process does not satisfy the second preset condition, it means that the iteration process does not satisfy the condition for ending the iteration, and the iterative optimization is continued.

[0170] In some embodiments, the second preset condition includes at least one of the conditions shown in (4-1) to (4-3) below.

[0171] (4-1) The second mean square error is less than the second preset threshold.

[0172] The second mean square error represents the difference between the dose volume histogram corresponding to the adjusted flux map and the reference dose volume histogram. The second preset threshold is a preset error threshold, such as 0.1, 0.01 or other thresholds. The embodiment of the present disclosure does not limit the setting of the second preset threshold.

[0173] (4-2) The third mean square error is less than the third preset threshold.

[0174] The third mean square error is determined based on the first mean square error and the second mean square error. The third preset threshold is a preset error threshold, such as 0.1, 0.01 or other thresholds. The embodiment of the present disclosure does not limit the setting of the third preset threshold.

[0175] Exemplarily, the process of determining the third mean square error may be: performing a weighted summation on the first mean square error and the second mean square error, and determining a value obtained by the weighted summation as the third mean square error.

[0176] Also illustratively, the process of determining the third mean square error may be: performing weighted summation on the first mean square error and the second mean square error to obtain a value obtained by the weighted summation, determining an average value of the weighted summation values, and determining the average value as the third mean square error.

[0177] It is worth noting that in other embodiments, other implementations can also be used to determine the third mean square error, which is not limited in the embodiments of the present disclosure.

[0178] (4-3) The number of iterative optimizations is greater than or equal to the preset number of iterations.

[0179] In the above embodiment, the second preset condition is described by taking the conditions shown in (4-1) to (4-3) as an example. It should be noted that the radiotherapy planning device can determine whether the iterative process meets the condition for ending the iteration based on one, two or more of the conditions shown in (4-1) to (4-3). In this way, the types of the second preset condition are enriched and the flexibility of the inverse optimization is improved.

[0180] The above S303A to S303C realizes two-stage reverse optimization. In the reverse optimization process of the first stage, the reference flux map is used as the objective function to quickly optimize the current radiotherapy plan, so that the first flux map after reverse optimization is closer to the reference flux map. Then, the reference dose volume histogram constructed based on the reference dose distribution is used as the objective function, and the reverse optimization is continued on the basis of the optimization results of the first stage to obtain the final flux map. In this way, a flux map with high accuracy can be obtained, so that the optimized radiotherapy plan can meet the established clinical requirements, thereby improving the accuracy of generating the radiotherapy plan.

[0181] It should be noted that the above S303A to S303C take two-stage reverse optimization as an example to illustrate the process of the radiotherapy planning device performing reverse optimization. In other embodiments, the process of the radiotherapy planning device performing reverse optimization in the above S303 can also refer to the following S303D to S303E.

[0182] S303D. The radiotherapy planning device constructs a reference dose volume histogram based on the reference dose distribution.

[0183] S303E: The radiotherapy planning device uses the reference flux map and the reference dose volume histogram as objective functions, performs inverse optimization, and obtains the optimized flux map.

[0184] In some embodiments, during any iteration process, the radiotherapy planning device uses the reference flux map and the reference dose-volume histogram as objective functions, adjusts the flux map adjusted in the previous iteration process, stops the iterative optimization in response to the current iteration process satisfying the third preset condition, and determines the flux map adjusted in the current iteration process as the optimized flux map.

[0185] The third preset condition is used to determine whether the iteration process in S303E satisfies the condition for ending the iteration. It should be understood that if the iteration process satisfies the third preset condition, it means that the iteration process satisfies the condition for ending the iteration, and the iterative optimization is stopped; if the iteration process does not satisfy the third preset condition, it means that the iteration process does not satisfy the condition for ending the iteration, and the iterative optimization is continued.

[0186] In some embodiments, the third preset condition includes at least one of the conditions shown in (5-1) to (5-4) below.

[0187] (5-1) The first mean square error is less than the first preset threshold.

[0188] (5-2) The second mean square error is less than the second preset threshold.

[0189] (5-3) The third mean square error is less than the third preset threshold.

[0190] (5-4) The number of iterative optimizations is greater than or equal to the preset number of iterations.

[0191] In the above embodiment, the third preset condition is described by taking the conditions shown in (5-1) to (5-4) as an example. It should be noted that the radiotherapy planning device can determine whether the iterative process meets the condition for ending the iteration based on one, two or more of the conditions shown in (5-1) to (5-4). In this way, the types of the third preset condition are enriched and the flexibility of the inverse optimization is improved.

[0192] The above S303D to S303E use the reference flux map and the reference dose volume histogram constructed based on the reference dose distribution as the objective function to quickly optimize the current radiotherapy plan, and can obtain a flux map with higher accuracy, so that the optimized radiotherapy plan can meet the established clinical requirements, thereby improving the accuracy of the generated radiotherapy plan.

[0193] It should be noted that the above S303A to S303C and S303D to S303E respectively take two-stage reverse optimization and one-stage reverse optimization as examples to illustrate the process of performing reverse optimization of the radiotherapy planning device.

[0194] S304: The radiotherapy planning device generates a radiotherapy plan based on the optimized flux map.

[0195] Among them, the radiotherapy plan is a document used to control the radiation therapy equipment to deliver radiation.

[0196] In some embodiments, the radiotherapy planning device converts the optimized flux map into leaf control information of multi-leaf grating blades in the radiotherapy device, and obtains the radiotherapy plan based on the leaf control information of the multi-leaf grating blades.

[0197] The multi-leaf grating blades of the radiotherapy device are used to open and close in relative directions to adjust the radiation field shape of the radiation emitted by the radiotherapy device.

[0198] In some embodiments, the blade control information may include blade motion information and blade radiation information. The blade motion information is used to indicate the motion trajectory of each blade in the multi-leaf grating blade. Exemplarily, the blade motion information may include the motion position coordinates and motion speed of each blade. The blade radiation information is used to indicate the radiation condition of the multi-leaf grating blade. Exemplarily, the blade radiation information may include machine hop count (MU value), hop count weight, etc.

[0199] In some embodiments, the radiotherapy planning device uses a Leaf Sequencing algorithm to convert the flux map into leaf control information of the multi-leaf grating leaves in the radiotherapy device.

[0200] For example, Figure 4 is a schematic diagram of a plan optimization process shown in an embodiment of the present disclosure. Figure 4 , Figure 4 Based on the two possible implementations shown in S302 and the two possible implementations shown in S303 , four generation schemes of radiotherapy plans are shown.

[0201] exist Figure 4 In the scheme 1 shown, the flux map prediction model is used to predict the flux map of each irradiation field (i.e., the reference flux map), and the flux map of each irradiation field is used as the objective function of the inverse optimization to perform inverse optimization to obtain a flux map after primary optimization (i.e., the first flux map). The dose distribution prediction model is used to predict the three-dimensional dose distribution (i.e., the reference dose distribution), and the DVH (i.e., the reference dose volume histogram) is reconstructed based on the three-dimensional dose distribution. Furthermore, the reconstructed DVH is used as the objective function of the inverse optimization to perform inverse optimization to obtain a flux map after secondary optimization, and the flux map after secondary optimization is converted into a radiotherapy plan.

[0202] exist Figure 4 In the second scheme shown, the flux map of each irradiation field is predicted by the flux map prediction model. The three-dimensional dose distribution is predicted by the dose distribution prediction model, and the DVH is reconstructed according to the three-dimensional dose distribution. Then, the flux map of each irradiation field and the reconstructed DVH are used as the objective function of the inverse optimization, and the inverse optimization is performed to obtain the optimized flux map, which is converted into a radiotherapy plan.

[0203] exist Figure 4In the scheme 3 shown, the three-dimensional dose distribution is predicted by the dose distribution prediction model, and the DVH is reconstructed according to the three-dimensional dose distribution. The flux map of each irradiation field is predicted based on the predicted three-dimensional dose distribution by the flux map prediction model. Then, the flux map of each irradiation field is used as the objective function of the inverse optimization, and the inverse optimization is performed to obtain the flux map after the primary optimization. Then, the reconstructed DVH is used as the objective function of the inverse optimization, and the inverse optimization is performed to obtain the flux map after the secondary optimization, and the flux map after the secondary optimization is converted into a radiotherapy plan.

[0204] exist Figure 4 In the scheme 4 shown, the three-dimensional dose distribution is predicted by the dose distribution prediction model, and the DVH is reconstructed according to the three-dimensional dose distribution. The flux map of each irradiation field is predicted based on the predicted three-dimensional dose distribution by the flux map prediction model. Then, the flux map of each irradiation field and the reconstructed DVH are used as the objective function of the inverse optimization, and the inverse optimization is performed to obtain the optimized flux map, which is converted into a radiotherapy plan.

[0205] In some embodiments, the radiotherapy plan further includes the isocenter position of the radiotherapy device. Exemplarily, before executing S301, the isocenter position of the radiotherapy device is set on the radiotherapy planning device.

[0206] Accordingly, the process of the radiotherapy planning device obtaining the isocenter position of the radiotherapy device can be: in response to the information configuration operation of the radiotherapy device for the object to be radiotherapy, the configuration information of the radiotherapy device is obtained, and according to the type of radiotherapy device, the type of irradiation technology and the field configuration parameters, the isocenter position of the radiotherapy device is determined.

[0207] The configuration information includes the type of radiotherapy equipment, the type of irradiation technology, and the field configuration parameters. In some embodiments, the type of radiotherapy equipment may be intensity modulated radiation therapy (IMRT) or volumetric modulated arc therapy (VMAT). The type of irradiation technology may be the type of radiation, the energy of radiation, etc.

[0208] In some embodiments, the radiotherapy plan further includes a radiotherapy prescription dose and a number of radiotherapy treatments. Exemplarily, before executing S301, the radiotherapy prescription dose and the number of radiotherapy treatments are set on the radiotherapy planning device.

[0209] Accordingly, the process of the radiotherapy planning device acquiring the radiotherapy prescription dose and the number of radiotherapy treatments may be: in response to the configuration operation of the radiotherapy prescription dose and the number of radiotherapy treatments for the object to be radiotreated, acquiring the configured radiotherapy prescription dose and the number of radiotherapy treatments.

[0210] In some embodiments, when the radiotherapy plan meets clinical requirements, the radiotherapy plan is exported for subsequent radiotherapy.

[0211] The technical solution provided by the embodiment of the present disclosure predicts the reference dose distribution and reference flux map of the object to be radiotherapy, and uses the predicted reference dose distribution and reference flux map as the objective function of the plan optimization to perform inverse optimization, which not only realizes the inverse optimization based on the flux map, but also refers to the reference dose distribution and reference flux map in the inverse optimization process, so that the optimized flux map is more in line with the actual application scenario. In addition, the present disclosure also introduces artificial intelligence technology into the plan optimization process. Through artificial intelligence technology, optimization experience can be learned based on historical plans, and then the radiotherapy plan can be adjusted based on the optimization results by using inverse optimization. In this way, a method for generating a radiotherapy plan based on artificial intelligence guidance is provided, which not only realizes the rapid optimization of the radiotherapy plan, but also helps to improve the accuracy of the radiotherapy plan obtained according to the optimization results, greatly improves the efficiency of clinical planning, and reduces the workload and difficulty of medical personnel in completing the radiotherapy plan.

[0212] Figure 5 is a structural block diagram of a device for generating a radiotherapy plan shown in an embodiment of the present disclosure. Figure 5 , the device includes a prediction module 501, an optimization module 502 and a generation module 503. Wherein:

[0213] Prediction module 501, used to predict reference dose distribution and reference flux map of a subject to be radiotherapy;

[0214] An optimization module 502 is used to use the reference dose distribution and the reference flux map as objective functions, perform inverse optimization, and obtain an optimized flux map;

[0215] The generating module 503 is used to generate a radiotherapy plan based on the optimized flux map.

[0216] The technical solution provided by the disclosed embodiment performs inverse optimization by predicting the reference dose distribution and reference flux map of the object to be radiotherapy, and using the predicted reference dose distribution and reference flux map as the objective function of plan optimization. This not only realizes the inverse optimization based on the flux map, but the inverse optimization process also refers to the reference dose distribution and reference flux map, so that the optimized flux map is more in line with the actual application scenario, which is conducive to improving the accuracy of the radiotherapy plan obtained according to the optimized flux map.

[0217] In some embodiments, the prediction module 501 is specifically used to:

[0218] Acquiring first radiotherapy information of the subject to be radiotherapy; the first radiotherapy information includes radiotherapy subject information and radiotherapy equipment information;

[0219] The first radiotherapy information is input into a dose distribution prediction model, and the first radiotherapy information is processed by the dose distribution prediction model to obtain the reference dose distribution.

[0220] In some embodiments, the prediction module 501 is specifically used to:

[0221] Acquiring second radiotherapy information of the subject to be radiotherapy; the second radiotherapy information includes radiotherapy subject information, radiotherapy equipment information and target area structure information;

[0222] The second radiotherapy information is input into the first flux map prediction model, and the second radiotherapy information is processed by the first flux map prediction model to obtain the reference flux map; or,

[0223] Acquiring third radiotherapy information of the object to be radiotherapy; the third radiotherapy information is the target area structure information;

[0224] The reference dose distribution and the third radiotherapy information are input into a second flux map prediction model, and the reference dose distribution and the third radiotherapy information are processed by the second flux map prediction model to obtain the reference flux map.

[0225] In some embodiments, the device further comprises a selection module, configured to:

[0226] A dose distribution prediction model matching the case type of the object to be radiotherapy is selected from a plurality of candidate dose distribution prediction models; the dose distribution prediction model matching the case type of the object to be radiotherapy is trained based on first radiotherapy information of the case type.

[0227] In some embodiments, the optimization module 502 is specifically configured to:

[0228] Using the reference flux map as the objective function, performing inverse optimization to obtain a first flux map; constructing a reference dose volume histogram based on the reference dose volume histogram; using the reference dose volume histogram as the objective function, performing inverse optimization on the first flux map to obtain the optimized flux map; or,

[0229] According to the reference dose distribution, a reference dose volume histogram is constructed; the reference flux map and the reference dose volume histogram are used as objective functions, and inverse optimization is performed to obtain the optimized flux map.

[0230] In some embodiments, the optimization module 502 is specifically configured to:

[0231] In any iteration process, the reference flux map is used as the objective function to adjust the flux map adjusted in the previous iteration process;

[0232] In response to the current iteration process satisfying the first preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the first flux map.

[0233] In some embodiments, the optimization module 502 is specifically configured to:

[0234] In any iteration process, the reference dose volume histogram is used as the objective function to adjust the flux map adjusted in the previous iteration process;

[0235] In response to the current iteration process satisfying the second preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the optimized flux map.

[0236] In some embodiments, the optimization module 502 is specifically configured to:

[0237] In any iteration process, the reference flux map and the reference dose volume histogram are used as objective functions to adjust the flux map adjusted in the previous iteration process;

[0238] In response to the current iteration process satisfying the third preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the optimized flux map.

[0239] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, comprising at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for generating a radiotherapy plan provided by the present disclosure.

[0240] According to an embodiment of the present disclosure, the present disclosure further provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable an electronic device to execute the method for generating a radiotherapy plan provided by the present disclosure.

[0241] According to an embodiment of the present disclosure, the present disclosure further provides a computer program product, including a computer program, which, when executed by a processor, implements the method for generating a radiotherapy plan provided by the present disclosure.

[0242] In some embodiments, the electronic device may be the above Figure 1 The radiation therapy planning device shown in . Figure 6A schematic block diagram of an example electronic device 600 that can be used to implement an embodiment of the present disclosure is shown. The electronic device 600 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device 600 can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0243] like Figure 6 As shown, the electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 to a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0244] Multiple components in the electronic device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a disk, an optical disk, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the electronic device 600 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0245] The computing unit 601 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSP), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 601 performs the various methods and processes described above, such as the method for generating a radiotherapy plan. For example, in some embodiments, the method for generating a radiotherapy plan may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the method for generating a radiotherapy plan described above may be performed. Alternatively, in other embodiments, the computing unit 601 may be configured to execute the method for generating a radiotherapy plan in any other appropriate manner (for example, by means of firmware).

[0246] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard parts (ASSPs), system on chip systems (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0247] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0248] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0249] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user, such as a cathode ray tube (CRT) or a liquid crystal display (LCD) monitor; and a keyboard and a pointing device (e.g., a mouse or a trackball), through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and the input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0250] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: Local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.

[0251] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0252] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of this disclosure can be achieved, and this document is not limited here.

[0253] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for generating a radiotherapy plan, characterized in that: include: Predicting reference dose distribution and reference flux map of subjects to be radiotherapy; Taking the reference dose distribution and the reference flux map as objective functions, performing inverse optimization to obtain an optimized flux map; A radiotherapy plan is generated based on the optimized flux map.

2. The method according to claim 1, characterized in that Predicting the reference dose distribution of the subject to be radiotherapy comprises: Acquiring first radiotherapy information of the subject to be radiotherapy; the first radiotherapy information includes radiotherapy subject information and radiotherapy equipment information; The first radiotherapy information is input into a dose distribution prediction model, and the first radiotherapy information is processed by the dose distribution prediction model to obtain the reference dose distribution.

3. The method according to claim 2, characterized in that Predicting a reference flux map of the subject to be radiotherapy comprises: Acquiring second radiotherapy information of the subject to be radiotherapy; the second radiotherapy information includes the radiotherapy subject information, the radiotherapy equipment information and the target area structure information; inputting the second radiotherapy information into a first flux map prediction model, and processing the second radiotherapy information by using the first flux map prediction model to obtain the reference flux map; or, Acquiring third radiotherapy information of the object to be radiotherapy; the third radiotherapy information is the target area structure information; The reference dose distribution and the third radiotherapy information are input into a second flux map prediction model, and the reference dose distribution and the third radiotherapy information are processed by the second flux map prediction model to obtain the reference flux map.

4. The method according to claim 2, characterized in that: Before inputting the first radiotherapy information into a dose distribution prediction model and processing the first radiotherapy information by the dose distribution prediction model to obtain the reference dose distribution, the method further includes: A dose distribution prediction model matching the case type of the object to be radiotherapy is selected from a plurality of candidate dose distribution prediction models; the dose distribution prediction model matching the case type of the object to be radiotherapy is trained based on first radiotherapy information of the case type.

5. The method according to claim 1, characterized in that: The method of using the reference dose distribution and the reference flux map as objective functions and performing inverse optimization to obtain an optimized flux map comprises: Using the reference flux map as the objective function, performing inverse optimization to obtain a first flux map; constructing a reference dose volume histogram according to the reference dose distribution; using the reference dose volume histogram as the objective function, performing inverse optimization on the first flux map to obtain the optimized flux map; or, A reference dose volume histogram is constructed according to the reference dose distribution; the reference flux map and the reference dose volume histogram are used as objective functions, and inverse optimization is performed to obtain the optimized flux map.

6. The method according to claim 5, characterized in that The method of using the reference flux map as an objective function and performing inverse optimization to obtain a first flux map comprises: In any iteration process, the reference flux map is used as the objective function to adjust the flux map adjusted in the previous iteration process; In response to the current iteration process satisfying the first preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the first flux map.

7. The method according to claim 5, characterized in that The step of taking the reference dose volume histogram as an objective function and performing inverse optimization on the first flux map to obtain the optimized flux map comprises: In any iteration process, the reference dose volume histogram is used as an objective function to adjust the flux map adjusted in the previous iteration process; In response to the current iteration process satisfying the second preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the optimized flux map.

8. The method according to claim 5, characterized in that The step of taking the reference flux map and the reference dose volume histogram as objective functions and performing inverse optimization to obtain the optimized flux map comprises: In any iteration process, the reference flux map and the reference dose volume histogram are used as objective functions to adjust the flux map adjusted in the previous iteration process; In response to the current iteration process satisfying the third preset condition, the iterative optimization is stopped, and the flux map adjusted in the current iteration process is determined as the optimized flux map.

9. An electronic device, comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 8.

10. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause an electronic device to execute the method according to any one of claims 1 to 8.

11. A computer program product comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 8.

Images