Radiotherapy parameter generation method and device, equipment and storage medium

Abstract

The application relates to a radiotherapy parameter generation method, device and equipment and a storage medium. The method comprises the following steps: inputting a medical image into a preset neural network model to obtain an isodose line segmentation graph corresponding to the medical image; constructing a target function according to the isodose line segmentation graph and a preset probability threshold set; and solving the target function according to a preset constraint condition to obtain parameters corresponding to a radiotherapy plan. The method can improve the accuracy of the generated treatment plan.

Description

Technical Field

[0001] This application relates to the field of radiotherapy technology, and in particular to a method, apparatus, device and storage medium for generating radiotherapy parameters. Background Technology

[0002] With the development of radiotherapy technology, radiotherapy can use radiation to treat tumors locally. Therefore, accurate treatment planning is particularly important in radiotherapy.

[0003] In traditional techniques, doctors mainly predict dosage values ​​based on their own experience or obtain clinically required dosage values, and use the given dosage values ​​to obtain dosage distribution and generate a treatment plan.

[0004] However, traditional methods for generating treatment plans suffer from low accuracy. Summary of the Invention

[0005] Therefore, it is necessary to provide a method, apparatus, device, and storage medium for generating radiotherapy parameters that can improve the accuracy of the generated treatment plan, in response to the above-mentioned technical problems.

[0006] In a first aspect, this application provides a method for generating radiotherapy parameters, the method comprising:

[0007] The medical images are input into a preset neural network model to obtain the isodose line segmentation map corresponding to the medical images;

[0008] Based on the isodose line segmentation diagram and the preset set of probability thresholds, an objective function is constructed;

[0009] The objective function is solved according to the preset constraints to obtain the parameters corresponding to the radiotherapy plan.

[0010] In one embodiment, constructing the objective function based on the isodose line segmentation map and a preset set of probability thresholds includes:

[0011] The target index is determined based on the isodose line segmentation diagram.

[0012] The objective function is constructed using the target index and the set of probability thresholds.

[0013] In one embodiment, determining the target index based on the isodose line segmentation map includes:

[0014] The target index is determined from the isodose line segmentation map based on the type of region of interest in the medical image.

[0015] In one embodiment, constructing the objective function using the target indicator and the set of probability thresholds includes:

[0016] Perform a traversal operation; the traversal operation includes: selecting a target probability threshold from the probability threshold set, removing pixels in the target isodose line whose pixel value is less than the target probability threshold, obtaining the processed target isodose line, and generating a penalty function corresponding to the target probability threshold based on the processed target isodose line;

[0017] Return to perform the traversal operation until all probability thresholds in the probability threshold set have been traversed, and obtain the penalty function corresponding to each probability threshold;

[0018] The objective function is constructed based on the penalty function corresponding to each probability threshold.

[0019] In one embodiment, if the target indicator is a volume percentage, the step of constructing the target function using the target indicator and the set of probability thresholds includes:

[0020] Perform a traversal operation; the traversal operation includes: selecting a target probability threshold from the probability threshold set, obtaining the volume percentage and dose value of the pixel value of the region of interest in the medical image that is greater than the target probability threshold, and generating a penalty function corresponding to the target probability threshold based on the volume percentage and the dose value;

[0021] Return to perform the traversal operation until all probability thresholds in the probability threshold set have been traversed, and obtain the penalty function corresponding to each probability threshold;

[0022] The objective function is constructed based on the penalty function corresponding to each probability threshold.

[0023] In one embodiment, constructing the objective function based on the penalty function corresponding to each probability threshold includes:

[0024] The objective function is constructed by multiplying the penalty function corresponding to each probability threshold and the optimization weight corresponding to each probability threshold.

[0025] In one embodiment, the method further includes:

[0026] Multiple optimized radiotherapy plans are generated based on the parameters;

[0027] Based on the user-triggered selection command, the target radiotherapy plan is determined from the multiple optimized radiotherapy plans.

[0028] Secondly, this application also provides a radiotherapy parameter generation device, the device comprising:

[0029] The first acquisition module is used to input medical images into a preset neural network model to obtain isodose line segmentation maps corresponding to the medical images;

[0030] A construction module is used to construct an objective function based on the isodose line segmentation map and a preset set of probability thresholds;

[0031] The second acquisition module is used to solve the objective function according to preset constraints to obtain the parameters corresponding to the radiotherapy plan.

[0032] Thirdly, this application also provides a computer device, the computer device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0033] The medical images are input into a preset neural network model to obtain the isodose line segmentation map corresponding to the medical images;

[0034] Based on the isodose line segmentation diagram and the preset set of probability thresholds, an objective function is constructed;

[0035] The objective function is solved according to the preset constraints to obtain the parameters corresponding to the radiotherapy plan.

[0036] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0037] The medical images are input into a preset neural network model to obtain the isodose line segmentation map corresponding to the medical images;

[0038] Based on the isodose line segmentation diagram and the preset set of probability thresholds, an objective function is constructed;

[0039] The objective function is solved according to the preset constraints to obtain the parameters corresponding to the radiotherapy plan.

[0040] Fifthly, this application also provides a computer program product, which includes a computer program that, when executed by a processor, performs the following steps:

[0041] The medical images are input into a preset neural network model to obtain the isodose line segmentation map corresponding to the medical images;

[0042] Based on the isodose line segmentation diagram and the preset set of probability thresholds, an objective function is constructed;

[0043] The objective function is solved according to the preset constraints to obtain the parameters corresponding to the radiotherapy plan.

[0044] The aforementioned radiotherapy parameter generation method, apparatus, device, and storage medium, by inputting medical images into a preset neural network model, can obtain isodose line segmentation maps corresponding to the medical images through the neural network model. The pixel value of each pixel in the isodose line segmentation map represents the probability value of each pixel within the corresponding isodose line. Therefore, based on the isodose line segmentation map and a preset set of probability thresholds, an objective function can be constructed. Solving the constructed objective function according to preset constraints can yield a radiotherapy plan that better meets clinical requirements. In other words, solving the constructed objective function according to preset constraints improves the accuracy of the parameters corresponding to the obtained radiotherapy plan, thereby improving the accuracy of the generated treatment plan. Attached Figure Description

[0045] Figure 1 This is an application environment diagram of a radiotherapy parameter generation method in one embodiment;

[0046] Figure 2 This is a flowchart illustrating a method for generating radiotherapy parameters in one embodiment;

[0047] Figure 3 This is a schematic diagram of the distribution of the 5000 cGy isodose line in one embodiment;

[0048] Figure 4 This is a flowchart illustrating a method for generating radiotherapy parameters in another embodiment;

[0049] Figure 5 This is a schematic diagram illustrating the determination of volume percentage from isodose line segmentation based on key indicators of critical organs in one embodiment.

[0050] Figure 6 This is a flowchart illustrating a method for generating radiotherapy parameters in another embodiment;

[0051] Figure 7 This is a flowchart illustrating a method for generating radiotherapy parameters in another embodiment;

[0052] Figure 8 This is a schematic diagram of the interpolation method in one embodiment;

[0053] Figure 9 This is a flowchart illustrating a method for generating radiotherapy parameters in another embodiment;

[0054] Figure 10 This is a structural block diagram of a radiotherapy parameter generation device in one embodiment. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0056] The radiotherapy parameter generation method provided in this application embodiment can be applied to, for example, Figure 1 The computer device shown includes a processor and a memory connected via a system bus. The memory stores a computer program, and the processor executes the computer program to perform the steps described in the method embodiments below. Optionally, the computer device may further include a network interface, a display screen, and an input device. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with external terminals via a network connection. Optionally, the computer device may be a server, a personal computer, a personal digital assistant, or other terminal devices, such as tablet computers, mobile phones, etc., or it may be a cloud or remote server. This application embodiment does not limit the specific form of the computer device.

[0057] In one embodiment, such as Figure 2 As shown, a method for generating radiotherapy parameters is provided, which can be applied to... Figure 1 Taking a computer device as an example, the explanation includes the following steps:

[0058] S201, input the medical image into the preset neural network model to obtain the isodose line segmentation map corresponding to the medical image.

[0059] Isodose lines are an important feature of dose distribution. They reflect the regions above and below a certain dose value within a given dose distribution. Isodose lines can be used to delineate these regions. Theoretically, there are several isodose lines within a given dose distribution. The pixel value of each pixel in the isodose line segmentation image represents the probability that each pixel lies within its corresponding isodose line. For example, Figure 3 This is a schematic diagram of the distribution of the 5000 cGy isodose line. For example, X% of the prescribed amount can be taken as the dose value of the isodose line, where X = 10, 20, 50, 80, 90, etc.

[0060] In addition, each isodose line segmentation map consists of a 3D pixel image. The value of each pixel is in the range of 0 to 1, which represents the probability that the pixel is within the isodose line. 0 means that the dose at the pixel will definitely not exceed the corresponding dose value, 1 means that the dose at the pixel will definitely exceed the corresponding dose value, and other values ​​represent the probability that the dose at the pixel exceeds the corresponding dose value.

[0061] Optionally, in this embodiment, the medical image is a medical image including organs, and the preset neural network model can be any one of the following: Convolutional Neural Network (CNN), ResNet, DenseNet, and U-Net. Optionally, the loss function used when training the neural network model can be the mean square error function, mean absolute error, etc. Optionally, in this embodiment, when the computer device trains the preset neural network model, it can use the medical image including organs and the dose distribution corresponding to the generated treatment plan of the medical image to train the preset neural network model. It should be noted that in this embodiment, the isodose line segmentation map corresponding to the medical image can be a set of isodose line segmentation maps corresponding to several isodose lines. In some embodiments, different isodose lines can be output independently, that is, through the preset neural network model, a series of isodose line segmentation maps corresponding to different isodose lines are output.

[0062] S202, construct the objective function based on the isodose line segmentation diagram and the preset set of probability thresholds.

[0063] The preset probability threshold set can be formed by selecting a set of discretely distributed probability thresholds from a preset probability threshold range. For example, a set of uncertainty thresholds can be given as a uniformly distributed discrete value on [0,1]. Optionally, in this embodiment, the computer device can construct different penalty functions for isodose lines based on the probability values ​​in the isodose line segmentation map and different probability thresholds in the preset probability threshold set, and construct an objective function using the different penalty functions for isodose lines. Optionally, the computer device can process the isodose line segmentation map using the selected probability thresholds, removing pixels whose probability values ​​are less than the selected probability thresholds to obtain a processed isodose line segmentation map, and construct an objective function using the processed isodose line segmentation map. Optionally, the computer device can construct an objective function based on the isodose line segmentation map and the preset probability threshold set, based on any penalty function, such as a quadratic function, polynomial function, logarithmic function, Lp norm, etc.

[0064] S203, solve the objective function according to the preset constraints to obtain the parameters corresponding to the radiotherapy plan.

[0065] The preset constraint can be that the difference between the parameters corresponding to the radiotherapy plan obtained from solving the objective function twice consecutively is less than a preset threshold. It should be noted that the constructed objective function is a function related to the parameters corresponding to the radiotherapy plan. Optionally, in this embodiment, the computer device can adjust the parameters of the objective function to minimize its value, and then determine the parameters of the objective function at this point as the parameters corresponding to the radiotherapy plan.

[0066] It is understandable that the objective function can be one or more. If the objective function is one, the parameters corresponding to the obtained radiotherapy plan will also be one; if the objective function is multiple, the parameters corresponding to the obtained radiotherapy plan will also be multiple.

[0067] Furthermore, after obtaining the parameters corresponding to the radiotherapy plan, the computer equipment can optimize the parameters based on them, generate a radiotherapy plan and dose statistics results (including dose distribution, dose-volume histogram, ROI dose statistics, etc.), and save the generated radiotherapy plan. Users can search among the generated treatment plans using a scroll bar, display multiple generated radiotherapy plans, and determine the optimal radiotherapy plan from among the multiple radiotherapy plans.

[0068] In the aforementioned method for generating radiotherapy parameters, by inputting medical images into a preset neural network model, the model can obtain isodose line segmentation maps corresponding to the medical images. The pixel values ​​of each pixel in the isodose line segmentation map represent the probability values ​​of each pixel within the corresponding isodose line. Therefore, based on the isodose line segmentation map and a preset set of probability thresholds, an objective function can be constructed. Solving the constructed objective function according to preset constraints can yield a radiotherapy plan that better meets clinical requirements. In other words, solving the constructed objective function according to preset constraints improves the accuracy of the parameters corresponding to the obtained radiotherapy plan, thereby improving the accuracy of the generated treatment plan.

[0069] In the scenario described above, where an objective function is constructed based on isodose line segmentation maps and a preset set of probability thresholds, the computer device can construct the objective function based on isodose lines in the isodose line segmentation map and the preset set of probability thresholds, or it can obtain the organ volume percentage from the isodose line segmentation map and construct the objective function based on the organ volume percentage and the preset set of probability thresholds. In one embodiment, such as Figure 4 As shown, the above S202 includes:

[0070] S301, determine the target index based on the isodose line segmentation diagram.

[0071] Optionally, the target indicator in this embodiment can be a target isodose line or a volume percentage. Optionally, in this embodiment, the computer device can determine the target indicator from the isodose line segmentation map based on the type of region of interest (e.g., case type) in the medical image. For example, if the target indicator in this embodiment is a target isodose line, the computer device can determine the target isodose line from the isodose line segmentation map based on the type of region of interest (e.g., case type) in the medical image. For example, the isodose line of 5000 cGy is typically considered for the brainstem, and the isodose line of 4000 cGy is typically considered for the spine. Optionally, if the target indicator in this embodiment is a volume percentage, the computer device can determine the volume percentage from the isodose line segmentation map based on key clinical indicators of critical organs. For example, in lung cancer cases, the volume percentage exceeding 2500 cGy in the lungs is considered; in cardiac cases, the volume percentage exceeding 3000 cGy in the heart is considered. For example, as... Figure 5 As shown, the solid circle represents the isodose line of 2500 cGy, and the dashed circle represents the lung. The intersection of the solid and dashed circles represents the volume of the lung exceeding 2500 cGy, which can be represented by x. If the total volume of the lung is represented by V, then the proportion of the lung volume exceeding 2500 cGy in lung cancer cases can be represented as x / V.

[0072] S302, construct the objective function using the target index and the set of probability thresholds.

[0073] Optionally, in this embodiment, if the target indicator is a target isodose line, the computer device can select a probability threshold from the above probability threshold set and remove pixels in the target isodose line whose probability value is less than the selected probability threshold to obtain a processed isodose line. The processed isodose line is then used to construct the above objective function. Optionally, in this embodiment, if the target indicator is a volume percentage, the computer device can select a probability threshold from the above probability threshold set and obtain the volume percentage and dose value of the region of interest in the medical image whose pixel value is greater than the selected probability threshold. The obtained volume percentage and dose value are then used to construct the above objective function.

[0074] In this embodiment, the computer device can accurately determine the target index based on the isodose line segmentation map, thereby using the target index and a preset set of probability thresholds to accurately construct the target function, thus improving the accuracy of the target function constructed by the computer device.

[0075] In the scenario described above, where a target function is constructed using a target indicator and a preset set of probability thresholds, if the target indicator is a target isodose line, then in one embodiment, such as... Figure 6 As shown, the above S302 includes:

[0076] S401, Perform a traversal operation; the traversal operation includes: selecting a target probability threshold from the probability threshold set, removing pixels in the target isodose line whose pixel value is less than the target probability threshold, obtaining the processed target isodose line, and generating a penalty function corresponding to the target probability threshold based on the processed target isodose line.

[0077] Optionally, in this embodiment, the computer device can arbitrarily select a target probability threshold from the aforementioned probability threshold set, and remove pixels in the target isodose line whose pixel values ​​are less than the target probability threshold, thus obtaining the processed target isodose line. Further, as an optional implementation, the computer device can generate a penalty function corresponding to the target probability threshold based on the processed target dose line and a preset penalty function. Additionally, it should be noted that the aforementioned target probability threshold can be selected by the computer device from the probability threshold set, or it can be selected by the user from the probability threshold set.

[0078] S402, return to perform the traversal operation until all probability thresholds in the probability threshold set have been traversed, and obtain the penalty function corresponding to each probability threshold.

[0079] In this embodiment, the computer device then returns to perform the above traversal operation, that is, arbitrarily selects a target probability threshold from the above probability threshold set, generates the penalty function corresponding to the target probability threshold, until all probability thresholds in the probability threshold set have been traversed, and the penalty function corresponding to each probability threshold in the probability threshold set is obtained.

[0080] S403, construct the objective function based on the penalty function corresponding to each probability threshold.

[0081] Optionally, in this embodiment, the computer device can construct the above-mentioned objective function based on the product of the penalty function corresponding to each probability threshold and the optimization weight corresponding to each probability threshold. For example, as an optional implementation, the objective function constructed by the computer device can be as follows:

[0082]

[0083] In the formula, x represents the optimization variable of the machine parameters, such as flux size, multi-page grating blade position, dose rate, etc., d(x) represents the dose distribution under machine parameter x, and F k (d(x)) corresponds to the penalty function generated by the kth isodose line segmentation plot, w k These are the corresponding optimization weights, where κ represents the set of all dose line segmentation plots, |κ|=K, and K represents the number of isodose line segmentation plots; G(d(x))≤0 represents the clinically required constraint, x∈C represents the constraints on machine parameters, C represents the set of constraints on machine parameters, and F... kThe expression of d(x) is as follows: wherein f(·) is any penalty function, such as quadratic function, polynomial function, logarithmic function, Lp norm, etc., q ik represents the uncertainty of the i-th pixel on the k-th iso-dose contour segmentation map, {i:q ik <p} represents a pixel point set, wherein the uncertainty of each pixel point in the set is less than a given p, t ik is the optimization weight corresponding to the pixel, for example, t ik = 1 or t ik = q ik , etc.

[0084] In this embodiment, by selecting a target probability threshold from the probability threshold set and eliminating pixel points with pixel values less than the target probability threshold in the target iso-dose contour, the processed target iso-dose contour can be obtained, so that the penalty function corresponding to the target probability threshold can be generated according to the processed target iso-dose contour. This operation is repeatedly executed until all probability thresholds in the probability threshold set are traversed, and penalty functions corresponding to each probability threshold can be obtained, and then the objective function can be accurately constructed according to the penalty functions corresponding to each probability threshold, which improves the accuracy of the constructed objective function.

[0085] In the foregoing scenario of constructing an objective function by using a target indicator and a preset probability threshold set, if the foregoing target indicator is a volume ratio, in one embodiment, as shown in Figure 7 , the foregoing S302 includes:

[0086] S501, execute a traversal operation; the traversal operation includes: selecting a target probability threshold from the probability threshold set, acquiring the volume ratio and dose value of pixels with pixel values greater than the target probability threshold in a region of interest in a medical image, and generating a penalty function corresponding to the target probability threshold according to the volume ratio and the dose value.

[0087] Optionally, in this embodiment, a computer device may randomly select a target probability threshold from the foregoing probability threshold set, acquire the volume ratio and dose value of pixels with pixel values greater than the target probability threshold in a region of interest in a medical image, and generate a penalty function corresponding to the target probability threshold according to the volume ratio and dose value of pixels greater than the target probability threshold. Optionally, the computer device may generate the penalty function corresponding to the target probability threshold based on a preset penalty function according to the volume ratio and dose value of pixels greater than the target probability threshold.

[0088] S502, return to execute the traversal operation until all probability thresholds in the probability threshold set are traversed, and obtain penalty functions corresponding to each probability threshold.

[0089] In this embodiment, the computer device then returns to perform the above traversal operation, that is, arbitrarily selects a target probability threshold from the above probability threshold set, generates the penalty function corresponding to the target probability threshold, until all probability thresholds in the probability threshold set have been traversed, and the penalty function corresponding to each probability threshold in the probability threshold set is obtained.

[0090] S503, construct the objective function based on the penalty function corresponding to each probability threshold.

[0091] Optionally, in this embodiment, the computer device can construct the objective function based on the product of the penalty function corresponding to each probability threshold and the optimization weight corresponding to each probability threshold. For example, as an optional implementation, the objective function constructed by the computer device can be as follows: In this embodiment, the computer device can define a set of uncertainty thresholds for each critical organ in the medical image, with the threshold range in the range [0,1]. An arbitrary combination of uncertainty thresholds for a critical organ can be selected, for example, (P1,P2,…,P…). n ), where P n Let P be the uncertainty of the nth critical organ, based on the uncertainty threshold P of the nth critical organ. n Calculate the dose-volume histogram optimization objective for the critical organ. For the k-th isodose line segmentation, calculate the uncertainty exceeding P in the critical organ. n Volume percentage V k Determine the dose-volume histogram point (D). k V k ), through point (D) k V k Interpolation is performed on k = 1, 2, ..., K to obtain... Optionally, the interpolation method can be as follows: Figure 8 The examples shown are linear interpolation or cubic spline interpolation, etc. The resulting model is:

[0092]

[0093] In the formula, F n (d(x)) is defined as Among them, v n \ V represents the volume v in the nth critical organ. n Excluding volume The set of remaining pixels, where n∈N.

[0094] In this embodiment, by selecting a target probability threshold from the probability threshold set, the volume percentage and dose value of the pixel values ​​of the region of interest in the medical image that are greater than the target probability threshold are obtained. Based on the volume percentage and dose value that are greater than the target probability threshold, the penalty function corresponding to the target probability threshold can be accurately generated. This operation is repeated until all probability thresholds in the probability threshold set are traversed, and the penalty function corresponding to each probability threshold can be obtained. Then, based on the penalty function corresponding to each probability threshold, the target function can be accurately constructed, thereby improving the accuracy of the constructed target function.

[0095] Based on the above embodiments, users can also select the optimal radiotherapy plan from multiple generated radiotherapy plans. In one embodiment, such as... Figure 9 As shown, the above method also includes:

[0096] S601 generates multiple radiotherapy plans based on parameters.

[0097] Optionally, in this embodiment, if there is only one objective function, the computer device will obtain only one parameter corresponding to the radiotherapy plan; if there are multiple objective functions, the computer device will obtain multiple radiotherapy plan parameters. Optionally, the computer device can optimize the obtained parameters based on the obtained radiotherapy technical parameters to generate multiple optimized radiotherapy plans.

[0098] S602 determines the target radiotherapy plan from multiple optimized radiotherapy plans based on user-triggered selection instructions.

[0099] The target radiotherapy plan is the optimal radiotherapy plan among multiple optimized radiotherapy plans, and can be a plan that more closely matches actual needs. Optionally, in this embodiment, the computer device can receive a selection command triggered by the user, and determine the target radiotherapy plan from multiple optimized radiotherapy plans based on the selection command.

[0100] In this embodiment, the computer device can generate multiple optimized radiotherapy plans based on the acquired parameters, thereby determining the target radiotherapy plan from the multiple optimized radiotherapy plans based on the user-triggered selection command, ensuring that the determined target radiotherapy plan is the optimal radiotherapy plan among the multiple optimized radiotherapy plans.

[0101] It should be noted that the objective function constructed based on the target isodose line and the set of probability thresholds, and the objective function constructed based on the volume ratio and the set of probability thresholds, are two parallel schemes. Either scheme can be used in any situation, and there are no restrictions on the conditions for using either scheme.

[0102] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0103] Based on the same inventive concept, this application also provides a radiotherapy parameter generation apparatus for implementing the radiotherapy parameter generation method described above. The solution provided by this apparatus is similar to the implementation described in the above method; therefore, the specific limitations in one or more embodiments of the radiotherapy parameter generation apparatus provided below can be found in the limitations of the radiotherapy parameter generation method described above, and will not be repeated here.

[0104] In one embodiment, such as Figure 10 As shown, a radiotherapy parameter generation device is provided, comprising: a first acquisition module, a construction module, and a second acquisition module, wherein:

[0105] The first acquisition module is used to input medical images into a preset neural network model to obtain isodose line segmentation maps corresponding to the medical images;

[0106] The module is used to construct the objective function based on the isodose line segmentation map and a preset set of probability thresholds;

[0107] The second acquisition module is used to solve the objective function according to preset constraints and obtain the parameters corresponding to the radiotherapy plan.

[0108] The radiotherapy parameter generation device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, so it will not be described again here.

[0109] Based on the above embodiments, optionally, the above construction module includes: a determining unit and a construction unit, wherein:

[0110] The determination unit is used to determine the target index based on the isodose line segmentation map.

[0111] The building block is used to construct the objective function using the target metric and the set of probability thresholds.

[0112] The radiotherapy parameter generation device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, so it will not be described again here.

[0113] Based on the above embodiments, optionally, the determining unit is used to determine the target index from the isodose line segmentation map according to the type of region of interest in the medical image.

[0114] The radiotherapy parameter generation device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, so it will not be described again here.

[0115] Based on the above embodiments, optionally, the above-mentioned construction unit is used to perform a traversal operation; the traversal operation includes: selecting a target probability threshold from the probability threshold set, removing pixels in the target isodose line whose pixel value is less than the target probability threshold, obtaining a processed target isodose line, generating a penalty function corresponding to the target probability threshold based on the processed target isodose line; returning to perform the traversal operation until all probability thresholds in the probability threshold set have been traversed, obtaining the penalty function corresponding to each probability threshold; and constructing a target function based on the penalty function corresponding to each probability threshold.

[0116] The radiotherapy parameter generation device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, so it will not be described again here.

[0117] Based on the above embodiments, optionally, if the target indicator is volume percentage, the above-mentioned construction unit is used to perform a traversal operation; the traversal operation includes: selecting a target probability threshold from the probability threshold set, obtaining the volume percentage and dose value of the pixel value of the region of interest in the medical image that is greater than the target probability threshold, generating a penalty function corresponding to the target probability threshold based on the volume percentage and dose value; returning to perform the traversal operation until all probability thresholds in the probability threshold set have been traversed, obtaining the penalty function corresponding to each probability threshold; and constructing a target function based on the penalty function corresponding to each probability threshold.

[0118] The radiotherapy parameter generation device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, so it will not be described again here.

[0119] Based on the above embodiments, optionally, the above-mentioned construction unit is used to construct the objective function according to the product of the penalty function corresponding to each probability threshold and the optimization weight corresponding to each probability threshold.

[0120] The radiotherapy parameter generation device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, so it will not be described again here.

[0121] Based on the above embodiments, optionally, the above apparatus further includes: a generation module and a selection module, wherein:

[0122] The generation module is used to generate multiple optimized radiotherapy plans based on parameters.

[0123] The selection module is used to determine the target radiotherapy plan from multiple optimized radiotherapy plans based on user-triggered selection commands.

[0124] The radiotherapy parameter generation device provided in this embodiment can execute the above method embodiment, and its implementation principle and technical effect are similar, so it will not be described again here.

[0125] Each module in the aforementioned radiotherapy parameter generation device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0126] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0127] The medical images are input into a preset neural network model to obtain the isodose line segmentation map corresponding to the medical images;

[0128] Based on the isodose line segmentation map and the preset set of probability thresholds, an objective function is constructed;

[0129] The objective function is solved according to the preset constraints to obtain the parameters corresponding to the radiotherapy plan.

[0130] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0131] The medical images are input into a preset neural network model to obtain the isodose line segmentation map corresponding to the medical images;

[0132] Based on the isodose line segmentation map and the preset set of probability thresholds, an objective function is constructed;

[0133] The objective function is solved according to the preset constraints to obtain the parameters corresponding to the radiotherapy plan.

[0134] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0135] The medical images are input into a preset neural network model to obtain the isodose line segmentation map corresponding to the medical images;

[0136] Based on the isodose line segmentation map and the preset set of probability thresholds, an objective function is constructed;

[0137] The objective function is solved according to the preset constraints to obtain the parameters corresponding to the radiotherapy plan.

[0138] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0139] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0140] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0141] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for generating radiotherapy parameters, characterized in that, The method includes: The medical image is input into a preset neural network model to obtain an isodose line segmentation map corresponding to the medical image; the pixel value of each pixel in the isodose line segmentation map represents the probability value of each pixel within the corresponding isodose line; Based on the isodose line segmentation diagram and the preset set of probability thresholds, an objective function is constructed; The objective function is solved according to the preset constraints to obtain the parameters corresponding to the radiotherapy plan.

2. The method according to claim 1, characterized in that, The step of constructing an objective function based on the isodose line segmentation map and a preset set of probability thresholds includes: The target index is determined based on the isodose line segmentation diagram. The objective function is constructed using the target index and the set of probability thresholds.

3. The method according to claim 2, characterized in that, The step of determining the target index based on the isodose line segmentation map includes: The target index is determined from the isodose line segmentation map based on the type of region of interest in the medical image.

4. The method according to claim 3, characterized in that, The step of constructing the objective function using the target index and the set of probability thresholds includes: Perform a traversal operation; the traversal operation includes: selecting a target probability threshold from the probability threshold set, removing pixels in the target isodose line whose pixel value is less than the target probability threshold, obtaining the processed target isodose line, and generating a penalty function corresponding to the target probability threshold based on the processed target isodose line; Return to perform the traversal operation until all probability thresholds in the probability threshold set have been traversed, and obtain the penalty function corresponding to each probability threshold; The objective function is constructed based on the penalty function corresponding to each probability threshold.

5. The method according to claim 2, characterized in that, If the target indicator is volume percentage The step of constructing the objective function using the target index and the set of probability thresholds includes: Perform a traversal operation; the traversal operation includes: selecting a target probability threshold from the probability threshold set, obtaining the volume percentage and dose value of the pixel value of the region of interest in the medical image that is greater than the target probability threshold, and generating a penalty function corresponding to the target probability threshold based on the volume percentage and the dose value; Return to perform the traversal operation until all probability thresholds in the probability threshold set have been traversed, and obtain the penalty function corresponding to each probability threshold; The objective function is constructed based on the penalty function corresponding to each probability threshold.

6. The method according to claim 4 or 5, characterized in that, The step of constructing the objective function based on the penalty function corresponding to each probability threshold includes: The objective function is constructed by multiplying the penalty function corresponding to each probability threshold and the optimization weight corresponding to each probability threshold.

7. The method according to claim 1, characterized in that, The method further includes: Multiple optimized radiotherapy plans are generated based on the parameters; Based on the user-triggered selection command, the target radiotherapy plan is determined from the multiple optimized radiotherapy plans.

8. A radiotherapy parameter generation device, characterized in that, The device includes: The first acquisition module is used to input medical images into a preset neural network model to obtain isodose line segmentation maps corresponding to the medical images; the pixel value of each pixel in the isodose line segmentation map represents the probability value of each pixel within the corresponding isodose line. A construction module is used to construct an objective function based on the isodose line segmentation map and a preset set of probability thresholds; The second acquisition module is used to solve the objective function according to preset constraints to obtain the parameters corresponding to the radiotherapy plan.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images