Intelligent radiotherapy plan generation method and system

Abstract

The invention discloses an intelligent radiotherapy plan generation method and system, and the method comprises the steps: obtaining patient information, generating candidate angle information in # imgabs0 #, and obtaining an initial candidate angle group of a radiation treatment plan radiation field incident angle; performing a first round of screening on the initial candidate angle group, sorting the candidate angle groups according to depth and area information, and screening out a first candidate angle group; for the first candidate angle group, generating a radiotherapy plan by using a dose prediction model based on artificial intelligence, predicting dose distribution, scoring the predicted dose distribution, and screening out a second candidate angle group according to a score; for the second candidate angle group, setting the same constraint condition, performing optimization, screening out the candidate angle group corresponding to the radiotherapy plan with the optimal optimization result, and recording the candidate angle group as the optimal candidate angle group; and for the optimal candidate angle group, plan optimization is finely set, and a final radiotherapy plan is obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of radiotherapy planning, and in particular to an intelligent radiotherapy plan generating method and system. Background Art

[0002] Tumor radiotherapy is one of the three major methods of tumor treatment. Its basic goal is to maximize the therapeutic gain ratio, that is, to ensure that the target area receives a certain coverage rate while the prescribed dose is irradiated, while the surrounding normal tissues are protected from or less exposed to unnecessary radiation.

[0003] In the existing patent technology, in order to solve the beam angle optimization problem of radiotherapy, the Chinese invention patent with publication number CN115920257B belongs to the field of intensity modulated radiotherapy, and discloses an intensity modulated radiotherapy beam angle optimization system, which compresses the radiotherapy data and adopts sparse reading, and then calculates the dose deposition matrix; solves the intensity distribution matrix of the beam direction by the conjugate gradient algorithm; calculates the dose distribution matrix by the dose deposition matrix and the intensity distribution matrix; constrains the beam direction by the side inhibition function and the neighboring effect function; optimizes the dose distribution matrix by the voxel dose condition in the sub-field, and obtains the target dose of each beam direction; calculates the importance of the beam direction based on the target dose, inhibition value and neighboring effect value; eliminates the beam direction with the lowest importance until the number of remaining beam directions is equal to the number of preset angles, and then restores the beam direction to the beam angle, and outputs the beam angle and its corresponding target dose. The invention system improves the calculation speed, effectively avoids the excessive concentration and dispersion of the beam angle, and improves the radiotherapy effect.

[0004] The above-mentioned invention patent calculates the importance of beam direction through target dose, inhibition value and proximity effect value, thereby eliminating beam directions with low importance until a preset number of angles is reached. However, this method is essentially a greedy algorithm and does not take into account the effects of different beam combinations.

[0005] In view of this, the present invention patent is proposed. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides an intelligent radiotherapy plan generation method and system, which specifically adopts the following technical solutions: An intelligent radiotherapy plan generation method, comprising: Obtain patient information and generate The candidate angle information is used to obtain an initial candidate angle group of incident angles of the radiation therapy plan field; Perform a first round of screening on the initial candidate angle group, sort the candidate angle groups according to the depth and area information, and screen out the first candidate angle group; For the first candidate angle group, a dose prediction model based on artificial intelligence is used to generate a radiotherapy plan, predict the dose distribution and score the predicted dose distribution, and the second candidate angle group is screened out according to the scoring value; For the second candidate angle group, the same constraint conditions are set and optimized to screen out the candidate angle group corresponding to the radiotherapy plan with the best optimization result, which is recorded as the optimal candidate angle group; For the best candidate angle group, the plan optimization is refined to obtain the final radiotherapy plan.

[0007] As an optional embodiment of the present invention, in a method for generating an intelligent radiotherapy plan of the present invention, the patient information is obtained, and the generated The candidate angle information is obtained, and the initial candidate angle group of the incident angle of the radiotherapy plan includes: The number of fixed fields is preset as m; Obtain the patient's CT image, organ outline, anatomical structure and isocenter information, select angles at equal intervals around the isocenter, remove taboo angles to obtain n candidate angles, and thus obtain a total of candidate angle groups. Group.

[0008] As an optional embodiment of the present invention, in an intelligent radiotherapy plan generation method of the present invention, the first round of screening of the initial candidate angle group includes: For one angle direction in the initial candidate angle group, the isocenter is projected onto each CT slice, the beam is cut off at the projection point, and the distance from the projection point where the beam is cut off vertically to the intersection with the outline is defined as the depth; The depths of all CT slices are averaged to obtain the depth information of one angle direction in the initial candidate angle group; The depth information of each angle direction in a group of initial candidate angle groups is averaged to obtain the depth information of the group of initial candidate angle groups.

[0009] As an optional embodiment of the present invention, in an intelligent radiotherapy plan generation method of the present invention, the first round of screening is performed on the initial candidate angle group, and the calculation process according to the area information includes: For an angle direction in the initial candidate angle group, find the CT slice with the largest target area, calculate the percentage of the target area to the beam mask area, and obtain the area information of an angle direction in the initial candidate angle group; The area information of each angle direction in a group of initial candidate angle groups is averaged to obtain the area information of the group of initial candidate angle groups.

[0010] As an optional embodiment of the present invention, in an intelligent radiotherapy plan generation method of the present invention, the sorting of the candidate angle groups according to the depth and area information to screen out the first candidate angle group includes: Setting one parameter of the depth information and the area information as a first priority parameter and the other parameter as a second priority parameter; Sorting each initial candidate angle group according to the first priority parameter, and if the first priority parameters are the same, continuing to sort according to the second priority parameter; A screening threshold of N% is preset, and the first N% of the initial candidate angle groups in the sorting are selected to obtain the first candidate angle group.

[0011] As an optional embodiment of the present invention, in an intelligent radiotherapy plan generation method of the present invention, for the first candidate angle group, using an artificial intelligence-based dose prediction model to generate a radiotherapy plan, predicting the dose distribution and scoring the predicted dose distribution, and screening out the second candidate angle group according to the scoring score includes: Extract the key dose volume indicators of the target area and organs at risk, set a piecewise function around the clinical reference value of each indicator, with a score of 0 to 1, accumulate all indicator scores, and obtain the predicted dose distribution score of a candidate angle in the first candidate angle group; The predicted dose distribution scores of all first candidate angle groups are counted, and the first M first candidate angle groups with the highest scores are selected to obtain the second candidate angle group, where M is a preset positive integer.

[0012] As an optional embodiment of the present invention, in an intelligent radiotherapy plan generation method of the present invention, the same constraint conditions are set for the second candidate angle group, and optimization is performed to screen out the candidate angle group corresponding to the radiotherapy plan with the best optimization result, which is recorded as the optimal candidate angle group including: For each second candidate angle in the second candidate angle group, the same constraint condition is set and optimized to obtain an optimized dose distribution; The optimized dose distribution is scored, and the second candidate angle group with the highest score is selected and recorded as the optimal candidate angle group.

[0013] As an optional embodiment of the present invention, in an intelligent radiotherapy plan generation method of the present invention, the step of refining the plan optimization for the optimal candidate angle group to obtain the final radiotherapy plan includes: For the predicted dose distribution corresponding to each angle in the optimal candidate angle group, the final radiotherapy plan is obtained by building a ring outside the target area to control the dose drop, setting biological constraints to reduce the dose to organs at risk, and adjusting the constraint weights.

[0014] The present invention also provides an intelligent radiotherapy plan generation system for implementing the intelligent radiotherapy plan generation method, comprising: Patient information acquisition module, obtains patient information and generates The candidate angle information is used to obtain an initial candidate angle group of incident angles of the radiation therapy plan field; A first screening module performs a first round of screening on the initial candidate angle group, sorts the candidate angle groups according to the depth and area information, and screens out the first candidate angle group; The second screening module generates a radiotherapy plan for the first candidate angle group using an artificial intelligence-based dose prediction model, predicts dose distribution and scores the predicted dose distribution, and screens out a second candidate angle group according to the scoring value; The third screening module sets the same constraint conditions for the second candidate angle group, performs optimization, and screens out the candidate angle group corresponding to the radiotherapy plan with the best optimization result, which is recorded as the best candidate angle group; The plan tuning module fine-tunes the plan optimization for the optimal candidate angle group to obtain the final radiotherapy plan.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides an intelligent radiotherapy plan generation method and system, which generates The candidate angle information is obtained, and a large number of initial candidate angle groups of the incident angles of the radiation therapy plan field are obtained. By performing three-step screening on a large number of initial candidate angle groups, the optimal candidate angle group is obtained, which realizes accurate and efficient angle screening and greatly reduces the calculation cost. It provides a heuristic solution to the combinatorial problem of radiation therapy angle optimization and has important practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flow chart of a method for generating an intelligent radiotherapy plan according to an embodiment of the present invention; Figure 2 An example picture of calculating depth information and area information in an example of a method for generating an intelligent radiotherapy plan according to an embodiment of the present invention; Figure 3 An example of a method for generating an intelligent radiotherapy plan according to an embodiment of the present invention includes a function graph example diagram of constructing a piecewise function for each extracted index during the process of screening out the second candidate angle group; Figure 4 A modular block diagram of an intelligent radiotherapy plan generation system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0017] To make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be described clearly and completely in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them.

[0018] Therefore, the following detailed description of the embodiments of the present invention is not intended to limit the scope of the invention claimed for protection, but merely represents some embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0019] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features and technical solutions in the embodiments may be combined with each other.

[0020] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0021] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "upper", "lower", etc. is based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the invention product is usually placed when in use, or the orientation or positional relationship commonly understood by those skilled in the art. Such terms are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0022] See also Figure 1 As shown, a method for generating an intelligent radiotherapy plan in this embodiment is characterized by comprising: Obtain patient information and generate The candidate angle information is used to obtain an initial candidate angle group of incident angles of the radiation therapy plan field; Perform a first round of screening on the initial candidate angle group, sort the candidate angle groups according to the depth and area information, and screen out the first candidate angle group; For the first candidate angle group, a dose prediction model based on artificial intelligence is used to generate a radiotherapy plan, predict the dose distribution and score the predicted dose distribution, and the second candidate angle group is screened out according to the scoring value; For the second candidate angle group, the same constraint conditions are set and optimized to screen out the candidate angle group corresponding to the radiotherapy plan with the best optimization result, which is recorded as the optimal candidate angle group; For the best candidate angle group, the plan optimization is refined to obtain the final radiotherapy plan.

[0023] In this embodiment, an intelligent radiotherapy plan generation method is provided, by acquiring patient information, generating The candidate angle information is obtained, and a large number of initial candidate angle groups of the incident angles of the radiation therapy plan field are obtained. By performing three-step screening on a large number of initial candidate angle groups, the optimal candidate angle group is obtained, which realizes accurate and efficient angle screening and greatly reduces the calculation cost. It provides a heuristic solution to the combinatorial problem of radiation therapy angle optimization and has important practical value.

[0024] As an optional implementation of this embodiment, in a method for generating an intelligent radiotherapy plan in this embodiment, the patient information is obtained, and the generated The candidate angle information is obtained, and the initial candidate angle group of the incident angle of the radiotherapy plan includes: The number of fixed fields is preset as m; Obtain the patient's CT image, organ outline, anatomical structure and isocenter information, select angles at equal intervals around the isocenter, remove taboo angles to obtain n candidate angles, and thus obtain a total of candidate angle groups. Group.

[0025] The taboo angle in this embodiment refers to a beam angle that is prohibited from being used during radiotherapy due to limitations of the radiotherapy equipment itself or the target area of ​​the patient.

[0026] As an optional implementation of this embodiment, in a method for generating an intelligent radiotherapy plan in this embodiment, the first round of screening is performed on the initial candidate angle group, and the depth calculation process includes: For one angle direction in the initial candidate angle group, the isocenter is projected onto each CT slice, the beam is cut off at the projection point, and the distance from the projection point where the beam is cut off vertically to the intersection with the outline is defined as the depth; The depths of all CT slices are averaged to obtain the depth information of one angle direction in the initial candidate angle group; The depth information of each angle direction in a group of initial candidate angle groups is averaged to obtain the depth information of the group of initial candidate angle groups.

[0027] As an optional implementation of this embodiment, in a method for generating an intelligent radiotherapy plan in this embodiment, the first round of screening is performed on the initial candidate angle group, and the calculation process according to the area information includes: For an angle direction in the initial candidate angle group, find the CT slice with the largest target area, calculate the percentage of the target area to the beam mask area, and obtain the area information of an angle direction in the initial candidate angle group; The area information of each angle direction in a group of initial candidate angle groups is averaged to obtain the area information of the group of initial candidate angle groups.

[0028] Specifically, in a method for generating an intelligent radiotherapy plan in this embodiment, the sorting of the candidate angle groups according to the depth and area information to screen out the first candidate angle group includes: Setting one parameter of the depth information and the area information as a first priority parameter and the other parameter as a second priority parameter; Sorting each initial candidate angle group according to the first priority parameter, and if the first priority parameters are the same, continuing to sort according to the second priority parameter; A screening threshold of N% is preset, and the first N% of the initial candidate angle groups in the sorting are selected to obtain the first candidate angle group.

[0029] As an optional implementation of this embodiment, in an intelligent radiotherapy plan generation method of this embodiment, for the first candidate angle group, using an artificial intelligence-based dose prediction model to generate a radiotherapy plan, predicting the dose distribution and scoring the predicted dose distribution, and screening out the second candidate angle group according to the scoring score includes: Extract the key dose volume indicators of the target area and organs at risk, set a piecewise function around the clinical reference value of each indicator, with a score of 0 to 1, accumulate all indicator scores, and obtain the predicted dose distribution score of a candidate angle in the first candidate angle group; The predicted dose distribution scores of all first candidate angle groups are counted, and the first M first candidate angle groups with the highest scores are selected to obtain the second candidate angle group, where M is a preset positive integer.

[0030] Furthermore, in an intelligent radiotherapy plan generation method of this embodiment, the same constraint conditions are set for the second candidate angle group, and optimization is performed to screen out the candidate angle group corresponding to the radiotherapy plan with the best optimization result, which is recorded as the best candidate angle group including: For each second candidate angle in the second candidate angle group, the same constraint condition is set and optimized to obtain an optimized dose distribution; The optimized dose distribution is scored, and the second candidate angle group with the highest score is selected and recorded as the optimal candidate angle group.

[0031] In this embodiment, the optimized predicted dose distribution is scored, and the scoring and screening method for screening out the second candidate angle group with the highest score is the same as the scoring and screening method for the first candidate angle group.

[0032] In an intelligent radiotherapy plan generation method of this embodiment, the step of finely setting the plan optimization for the optimal candidate angle group to obtain the final radiotherapy plan includes: For the predicted dose distribution corresponding to each angle in the optimal candidate angle group, the final radiotherapy plan is obtained by building a ring outside the target area to control the dose drop, setting biological constraints to reduce the dose to organs at risk, and adjusting the constraint weights.

[0033] In an intelligent radiotherapy plan generation method of this embodiment, a radiotherapy plan is generated for the first candidate angle group using an artificial intelligence-based dose prediction model. The dose prediction model is a deep learning model that uses CT, key ROI delineation, other anatomical structures, and beam angle information as input to predict the dose distribution under the beam angle group.

[0034] See also Figure 4 As shown, this embodiment also provides an intelligent radiotherapy plan generation system for implementing the intelligent radiotherapy plan generation method, including: Patient information acquisition module, obtains patient information and generates The candidate angle information is used to obtain an initial candidate angle group of incident angles of the radiation therapy plan field; A first screening module performs a first round of screening on the initial candidate angle group, sorts the candidate angle groups according to the depth and area information, and screens out the first candidate angle group; The second screening module generates a radiotherapy plan for the first candidate angle group using an artificial intelligence-based dose prediction model, predicts dose distribution and scores the predicted dose distribution, and screens out a second candidate angle group according to the scoring value; The third screening module sets the same constraint conditions for the second candidate angle group, performs optimization, and screens out the candidate angle group corresponding to the radiotherapy plan with the best optimization result, which is recorded as the best candidate angle group; The plan tuning module fine-tunes the plan optimization for the optimal candidate angle group to obtain the final radiotherapy plan.

[0035] In this embodiment, a smart radiotherapy plan generation system is provided, wherein the patient information acquisition module acquires patient information and generates The candidate angle information is obtained to obtain a large number of initial candidate angle groups for the incident angles of the radiation therapy plan field. The first screening module, the second screening module, and the third screening module obtain the optimal candidate angle group through three-step screening of a large number of initial candidate angle groups, thereby achieving accurate and efficient angle screening, greatly reducing the calculation cost, and providing a heuristic solution to the combinatorial problem of radiation therapy angle optimization, which has important practical value.

[0036] This embodiment provides a specific example of optimizing beam angles during radiotherapy plan generation by an intelligent radiotherapy plan generation system, including the following implementation steps: 1) Obtain the user's CT, outline, other anatomical structures and isocenter information, and calculate the Take an angle, remove the taboo angles, obtain all candidate angles n, fix the number of radiation fields m according to the tumor type, and obtain the total number of candidate angles Group.

[0037] 2) For each candidate angle group, calculate the depth information and area information (such as Figure 2 As shown in the figure), first sort by depth information, and for angle groups with the same depth information, sort them again by area information, and set the threshold to 5%, that is, take the top 5% of angle groups to enter the second screening module.

[0038] 3) Use the multi-branch UNet model (MB-UNet) to predict the dose distribution under each angle group. The model has been trained on a large number of data sets containing angle information. For the predicted dose distribution, the key dose volume indicators of the target area and organs at risk are extracted, such as CI (conformity), HI (uniformity), , , organ-threatening , , construct a piecewise function for each indicator, with values ​​between 0 and 1, such as Figure 3 As shown, all scores are accumulated to obtain the score statistics of all angle groups, and the top 5 angle groups enter the third screening module.

[0039] 4) Only set the target area and Restraint and critical organs at risk Constraints are set, and the corresponding plans of the above five angle groups are optimized. The angle group corresponding to the optimal plan is obtained through the same scoring method.

[0040] 5) Refine the design of the optimal angle group plan, control the dose drop by building a ring outside the target area, set biological constraints to reduce the dose to organs at risk, adjust the constraint weights, and finally optimize the plan.

[0041] The above embodiments are only used to illustrate the present invention and are not intended to limit the technical solutions described in the present invention. Although the present invention has been described in detail with reference to the above embodiments, the present invention is not limited to the above specific implementation methods. Therefore, any modification or equivalent replacement of the present invention; and all technical solutions and improvements thereof that do not depart from the spirit and scope of the invention are included in the scope of the claims of the present invention.

Claims

1. A method for generating an intelligent radiotherapy plan, characterized in that: include: Get patient information and generate The candidate angle information is used to obtain an initial candidate angle group of incident angles of the radiation therapy plan field; Perform a first round of screening on the initial candidate angle group, sort the candidate angle groups according to the depth and area information, and screen out the first candidate angle group; For the first candidate angle group, a dose prediction model based on artificial intelligence is used to generate a radiotherapy plan, predict the dose distribution and score the predicted dose distribution, and the second candidate angle group is screened out according to the scoring value; For the second candidate angle group, the same constraint conditions are set and optimized to screen out the candidate angle group corresponding to the radiotherapy plan with the best optimization result, which is recorded as the optimal candidate angle group; For the best candidate angle group, the plan optimization is refined to obtain the final radiotherapy plan.

2. The method for generating an intelligent radiotherapy plan according to claim 1, characterized in that: The patient information is obtained and generated The candidate angle information is obtained, and the initial candidate angle group of the incident angle of the radiotherapy plan includes: The number of fixed fields is preset as m; Obtain the patient's CT image, organ outline, anatomical structure and isocenter information, select angles at equal intervals around the isocenter, remove taboo angles to obtain n candidate angles, and thus obtain a total of candidate angle groups. Group.

3. The intelligent radiotherapy plan generation method according to claim 1 or 2, characterized in that: The first round of screening of the initial candidate angle group includes the following steps according to the depth calculation process: For one angle direction in the initial candidate angle group, the isocenter is projected onto each CT slice, the beam is cut off at the projection point, and the distance from the projection point where the beam is cut off vertically to the intersection with the outline is defined as the depth; The depths of all CT slices are averaged to obtain the depth information of one angle direction in the initial candidate angle group; The depth information of each angle direction in a group of initial candidate angle groups is averaged to obtain the depth information of the group of initial candidate angle groups.

4. The method for generating an intelligent radiotherapy plan according to claim 3, characterized in that: The first round of screening of the initial candidate angle group and the calculation process based on the area information include: For an angle direction in the initial candidate angle group, find the CT slice with the largest target area, calculate the percentage of the target area to the beam mask area, and obtain the area information of an angle direction in the initial candidate angle group; The area information of each angle direction in a group of initial candidate angle groups is averaged to obtain the area information of the group of initial candidate angle groups.

5. The method for generating an intelligent radiotherapy plan according to claim 4, characterized in that: The step of sorting the candidate angle groups according to the depth and area information to select the first candidate angle group includes: Setting one parameter of the depth information and the area information as a first priority parameter and the other parameter as a second priority parameter; Sorting each initial candidate angle group according to the first priority parameter, and if the first priority parameters are the same, continuing to sort according to the second priority parameter; A screening threshold of N% is preset, and the first N% of the initial candidate angle groups in the sorting are selected to obtain the first candidate angle group.

6. The method for generating an intelligent radiotherapy plan according to claim 4, characterized in that: For the first candidate angle group, a dose prediction model based on artificial intelligence is used to generate a radiotherapy plan, predict the dose distribution and score the predicted dose distribution, and the second candidate angle group is screened out according to the scoring value, including: Extract the key dose volume indicators of the target area and organs at risk, set a piecewise function around the clinical reference value of each indicator, with a score of 0 to 1, accumulate all indicator scores, and obtain the predicted dose distribution score of a candidate angle in the first candidate angle group; The predicted dose distribution scores of all first candidate angle groups are counted, and the first M first candidate angle groups with the highest scores are selected to obtain the second candidate angle group, where M is a preset positive integer.

7. The method for generating an intelligent radiotherapy plan according to claim 6, characterized in that: The same constraint conditions are set for the second candidate angle group, and optimization is performed to screen out the candidate angle group corresponding to the radiotherapy plan with the best optimization result, which is recorded as the best candidate angle group including: For each second candidate angle in the second candidate angle group, the same constraint condition is set and optimized to obtain an optimized dose distribution; The optimized dose distribution is scored, and the second candidate angle group with the highest score is selected and recorded as the optimal candidate angle group.

8. The method for generating an intelligent radiotherapy plan according to claim 6, characterized in that: The method of finely setting the plan optimization for the optimal candidate angle group to obtain the final radiotherapy plan includes: For the predicted dose distribution corresponding to each angle in the optimal candidate angle group, the final radiotherapy plan is obtained by building a ring outside the target area to control the dose drop, setting biological constraints to reduce the dose to organs at risk, and adjusting the constraint weights.

9. An intelligent radiotherapy plan generation system implementing the intelligent radiotherapy plan generation method according to any one of claims 1 to 8, characterized in that: include: Patient information acquisition module, obtains patient information and generates The candidate angle information is used to obtain an initial candidate angle group of incident angles of the radiation therapy plan field; A first screening module performs a first round of screening on the initial candidate angle group, sorts the candidate angle groups according to the depth and area information, and screens out the first candidate angle group; The second screening module generates a radiotherapy plan for the first candidate angle group using an artificial intelligence-based dose prediction model, predicts dose distribution and scores the predicted dose distribution, and screens out a second candidate angle group according to the scoring value; The third screening module sets the same constraint conditions for the second candidate angle group, performs optimization, and screens out the candidate angle group corresponding to the radiotherapy plan with the best optimization result, which is recorded as the best candidate angle group; The plan tuning module fine-tunes the plan optimization for the optimal candidate angle group to obtain the final radiotherapy plan.

Images