Method and system for treatment planning

CN119007933APending Publication Date: 2024-11-22ELEKTA AB
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410570745.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2023-05-11
Filing Date
2024-05-09
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The creation process of existing radiation therapy plans is time-consuming and complex, especially when multiple anatomical structures exist, making it difficult to effectively optimize treatment plans to reduce planning time and computational burden.

Method used

By receiving the reference target, selecting the cost function associated with it and specifying the initial weight value, using the optimization process to minimize the cost function to determine the parameters of the radiation therapy plan, automatically adjusting the weights to handle the sensitivity of different cost functions, iteratively optimizing the radiation Treatment planning for clinically acceptable accuracy.

Benefits of technology

Improve the efficiency and accuracy of the creation of radiotherapy plans, reduce the burden on users in the optimization process, and improve the speed of optimization and the quality of treatment plans.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119007933A_ABST
    Figure CN119007933A_ABST
Patent Text Reader

Abstract

A computer-implemented method may be provided to assist in a radiotherapy plan, the method comprising: receiving a reference target, the reference target representing an objective to be achieved by a radiotherapy system; selecting a cost function associated with the reference target from a plurality of cost functions, the selected cost function associated with a specified initial weight value corresponding to a sensitivity of the example dose distribution relative to a change in the selected cost function; an optimization procedure is applied to a radiotherapy plan for radiotherapy of the patient according to the received reference target, where the optimization procedure uses the specified initial weight value of the selected cost function and seeks to minimize the selected cost function.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to methods and systems for treatment planning. More specifically, the present disclosure relates to a computer-implemented method for radiation therapy planning and a data processing device, a computer program and a non-transitory computer-readable storage medium configured to execute the method for radiation therapy planning. Background Art

[0002] Radiation therapy, or radiotherapy, can be described as the use of ionizing radiation to damage or destroy unhealthy cells in humans and animals. Ionizing radiation can be directed to tumors on the surface of the skin or deep inside the body. Common forms of ionizing radiation include X-rays and charged particles. An example of a radiotherapy technique is irradiating a patient's tumor with multiple, lower-intensity gamma rays focused at a higher intensity and with high precision on a targeted area, such as a tumor. (Gamma Another example of radiation therapy includes the use of a linear accelerator ("linac"), whereby a targeted area is irradiated with high energy particles (e.g., electrons, high energy photons, etc.). In another example, radiation therapy can be provided using a heavy charged particle accelerator (e.g., protons, carbon ions, etc.).

[0003] The placement and dosage of radiation beams can be precisely controlled to provide a prescribed dose of radiation to a targeted area (e.g., a tumor) and reduce damage to surrounding healthy tissue (referred to as organs at risk or OARs). One aspect of treatment planning involves determining appropriate characteristics of the radiation to be delivered to produce a safe and effective dose. For example, the characteristics of the radiation relate to the flux pattern or distribution. The flux pattern may depend on the beam arrangement, energy, and field size, which in turn are related to controllable parameters (which are optimizable). By determining appropriate values ​​for these parameters, an appropriate flux pattern may be obtained. The treatment plan may be determined by a treatment planning system.

[0004] An optimization procedure can be used to establish a radiation therapy treatment plan (treatment plan, or simply plan) to determine the optimal parameter values ​​or the set of optimal variable values ​​that are expected to deliver an appropriate dose. The optimization process can be based on clinical and dosimetric goals and constraints. Examples of clinical and dosimetric goals and constraints include maximum, minimum, and mean doses to the targeted area and surrounding areas (such as tumors and critical organs). Clinical and dosimetric goals and constraints can be referred to as treatment plan goals. Optimization is typically performed with respect to one or more treatment plan parameters to shorten beam on time, improve dose uniformity, etc.

[0005] The treatment planning process can include using images (two-dimensional or three-dimensional) of the patient to identify the target area and identify critical organs near the target area. Segmentation can be used to identify the target area (or area to be treated, for example, the Planned Target Volume (PTV) and the surrounding area (or Organ At Risk (OAR)). After segmentation, a dose plan can be created for the patient, which indicates the desired amount of radiation to be received by the target area and / or the surrounding area. The target area can have an irregular volume and can be significantly different in size, shape, and location.

[0006] In a practical example, there may be multiple anatomical structures (targeted areas and / or surrounding areas). For example, in head and neck treatment, there may be more than 20 anatomical structures. For each structure, it may be desirable to comply with various treatment plan goals. In order to arrive at a clinically acceptable plan, these structures and their goals may be assigned different priorities. Determining a treatment plan that meets the various goals is time-consuming and complex.

[0007] The creation of a radiation therapy plan is typically a time-consuming process in which the planner attempts to comply with various treatment goals or constraints, possibly taking into account their own importance, to produce a clinically acceptable radiation therapy plan. Common problems faced when creating radiation therapy plans involving optimization processes include lengthy optimization times and high computational burdens required to achieve safe and satisfactory results. These processes typically involve trial and error on the part of the user (e.g., treatment planner, dosimetrist, clinician, or healthcare worker), which can be time-consuming and further complicated as further goals and constraints are added.

[0008] Therefore, it is desirable to reduce planning time through improved radiotherapy treatment planning workflow. Summary of the invention

[0009] According to one aspect, a computer-implemented method for radiotherapy treatment planning is provided. The method includes receiving a reference target that represents a goal to be achieved by a radiotherapy system. In this case, the reference target can be an anatomically specific function that establishes a dose and / or biological response target. Such a reference target is not necessarily met in the optimization process - instead, the reference target is the goal that the optimization process attempts to achieve. For example, the reference target can be a dose-based reference target that establishes a target for a maximum absorbed dose (provided in SI units of gray (Gy)) for a targeted region of a patient. The reference target can be a biologically based reference target that takes into account the concept that a specific biological structure or tissue can tolerate a higher dose when the irradiated volume is reduced. The reference target can be a target for a targeted region (i.e., a target for radiotherapy, such as a tumor), or it can also be a target for a surrounding region (i.e., one or more regions surrounding a targeted region). The surrounding region (or adjacent region, or OAR) does not need to be in physical contact with the targeted region, but can include regions of the patient other than the targeted region where the user wishes to limit or otherwise control the absorbed dose.

[0010] The method also includes: selecting a cost function associated with a reference target from a plurality of cost functions. The selected cost function is associated with a specified initial weight value. That is, each of the plurality of cost functions can be mapped to a specified initial weight value. For example, a first cost function can be associated with a first specified initial weight value, a second cost function can be associated with a second specified initial weight value, and so on; of course, the first specified initial weight value and the second specified initial weight value can be the same value. The specified initial weight value corresponds to or maps to the sensitivity of the example or reference dose distribution to changes in the selected cost function (or its output). Equivalently, the specified initial weight value corresponds to or maps to the sensitivity of the selected cost function (or its output) to changes in the example dose distribution. That is, the specified initial weight value can be used to indicate the scale of change of the dose distribution (e.g., a dose map or dose distribution indicating the spatial distribution of radiation dose intensity) given a change in the output of the selected cost function.

[0011] The method further includes applying an optimization process to a radiation therapy plan for radiation therapy of the patient based on the received reference target. The optimization process uses specified initial weight values ​​for the selected cost function. For example, the optimization process scales the cost function values ​​(outputs of the cost function) by the specified initial weight values. The optimization process seeks to minimize the selected cost function (or its output). In this manner, the optimization process seeks to determine treatment parameters (or their values) that are expected to produce a treatment plan that is as close as practicable (i.e., in terms of processing and / or time constraints) to achieving the reference target (or the objective represented by the reference target).

[0012] As an example, a 1% change in the integrated intensity of the example dose map may be associated with a 5% change in the first cost function value (relative to the cost function value associated with the integrated intensity of the example dose map before any change). The same 1% change in the integrated intensity of the example dose map may be associated with a 50% change in the second cost function value. In this example, the first cost function may be considered non-sensitive and may be assigned or associated with a first weight that is higher than a second weight associated with the second cost function. Of course, other metrics (i.e., metrics other than integrated intensity) for quantifying dose distributions (and changes thereto) are also available.

[0013] Mathematically, the goal of the optimization process is to achieve a minimum cost function value. However, in radiotherapy, the inventors have recognized that the dose distribution, rather than the cost function value, is the key criterion when evaluating the quality of the treatment plan. Different cost functions have different mathematical properties with respect to the relationship between the cost function value and the dose distribution. For some cost functions, slightly different cost function values ​​may correspond to significantly different dose distributions, and vice versa. Therefore, the method for radiotherapy planning achieves a better balance (i.e., reliably handles cost functions of various mathematical properties). The automatic assignment of weights to the cost function makes it possible to indicate how sensitive the cost function value is expected to be when the dose distribution (or the treatment parameters that cause the dose distribution) is changed.

[0014] For example, higher weights may be assigned to cost functions for which small changes in cost function values ​​result in very different dose distributions. Such cost functions may be referred to as "insensitive cost functions" because changes in their cost function values ​​are not as significant as for other cost functions. The cost function values ​​may then be obtained by factoring in the weights. Similarly, lower weights may be assigned to cost functions for which larger changes in cost function values ​​result in only slightly different dose distributions. Without this case-by-case treatment of the cost function, the optimization process may not converge because the cost function values ​​are constrained and cannot be changed significantly without achieving the reference target.

[0015] Such a non-sensitive cost function can be determined (i.e., classified as non-sensitive) using any suitable technique. As an example, the resulting dose distribution (or various aspects thereof, such as the mean) can be plotted relative to the cost function value to determine how sensitive the cost function value is to changes in different dose distributions. Initial weights can then be assigned to the cost function based on various aspects of the relationship between the resulting dose distribution and the cost function value. For example, for a normal (sensitive) cost function, the initial weight can be set to 1; for a cost function that is considered to be moderately non-sensitive, the initial weight can be set to 100; for a cost function that is considered to be very non-sensitive, the initial weight can be set to 500.

[0016] The inventors have appreciated that automatic modification of the cost function based on its sensitivity routinely improves optimization speed and the accuracy of the generated treatment plan. In automating the process, the user need not have the burden of selecting or handling a cost function, the behavior of which may vary significantly between different cost functions.

[0017] Optionally, the optimization process includes determining a dose distribution based on a reference target, wherein the dose distribution indicates an expected dose in a targeted region of the patient. The expected dose is a dose of a radiotherapy system configured with the treatment plan parameters (or their values) determined by the optimization process (e.g., expected to be delivered to the targeted region). Examples of optimizable parameters include parameters related to characteristics of the radiation to be delivered by the radiotherapy system, such as radiation beam angles, radiation intensity levels at each angle, etc. Additionally or alternatively, the dose distribution indicates an expected dose in a surrounding region of the patient.

[0018] Optionally, the optimization process can be an iterative process that includes multiple optimization iterations or passes. For example, the iterative optimization process can use initial values ​​(e.g., initial values ​​of the radiation beam angle) to generate a sequence of improved approximate solutions for optimal parameter values ​​(e.g., improved values ​​of the radiation beam angle), where each subsequent approximation is derived from the previous approximation. In the first optimization iteration (of multiple optimization iterations), when seeking to minimize the selected cost function, the optimization process uses a specified initial weight value for the selected cost function. The selected cost function - and possibly other cost functions from a plurality of cost functions - can be associated or mapped to a specified weight processing parameter. During subsequent iterations of the optimization process, the specified initial weight value may be modified.

[0019] In the example where the specified initial weight value may be 0.75, the example specified weight processing parameters may reduce the specified weight value by 0.05 in each subsequent iteration of the optimization process. Of course, the specified weight processing parameters and associated functions may be more complex, such as including a threshold to stop modifying the specified weight value after a set number of iterations, or including some polynomial-based modification properties to increase the specified weight value by a polynomial factor (e.g., 0.05×n for n iterations). -1 ) to modify the specified weight value.

[0020] Specifying weight processing parameters may cause weights to increase during some or all subsequent iterations. After specifying initial weights for each cost function, the method may increase the weights when the achieved value is worse than the reference target. As an example, in the case of a sensitive cost function, if the achieved value is only a little worse than the reference target (e.g., worse by at most a predetermined value or percentage), the weight can be increased by a small amount (e.g., by 10%). If the achieved value is much worse than the reference target (e.g., worse by at least another predetermined value or percentage), the weight is increased by a large amount (e.g., by 10 times). In another example, in the case of a non-sensitive cost function, if the achieved value is only a little worse than the reference target, the weight can be increased by a small amount (e.g., by one or two times). If the achieved value is much worse than the reference target, the weight can be increased by a large amount (e.g., by 100 times).

[0021] Similarly, specifying weight processing parameters may cause weights to be reduced during some or all subsequent iterations. After specifying initial weights for each cost function, the method may reduce weights when the achieved value is better than the reference target. As an example, in the case of a sensitive cost function, if the achieved value is only a little better than the reference target, the weight can be reduced by a small amount (e.g., by 10%). If the achieved value is much better than the reference target, the weight can be reduced by a large amount (e.g., by 10 times). In another example, in the case of a non-sensitive cost function, if the achieved value is only a little better than the reference target, the weight can be reduced by a small amount (e.g., by one or half). If the achieved value is much better than the reference target, the weight can be reduced by a large amount (e.g., by 100 times).

[0022] For non-sensitive cost functions, some small changes in the cost function value may result in significant differences in the dose distribution, and the inventors have realized that the cost function value must be amplified by aggressively changing the weights to capture this behavior. For example, consider a scenario where the cost function value is 1 and the goal (reference target) is exactly achieved. For the first cost function, a cost function value of 1.1 indicates that the goal is almost achieved (thus, for the purposes of radiotherapy planning, a cost function value of 1.1 is acceptable). However, for the non-sensitive second cost function, a cost function value of 1.1 may indicate a significant violation of the goal. The inventors have realized that improvements can be seen if the optimization process is made "knowing" that a cost function value of 1.1 is acceptable for the first cost function, but not for the second cost function. By assigning a higher weight to the second cost function, the cost function value (i.e., the "actual" cost function value multiplied by the weight) becomes more sensitive to changes in the dose distribution.

[0023] For example, the optimization process seeks to reduce the cost function value. For the first cost function, 1.1 is not considered ideal (1 is considered ideal), but is considered acceptable. Instead, the weights may only need to be increased by a small factor (e.g., by 10%), thereby illustrating that the first cost function should not "give up" after reaching 1.1 because the goal is 1. Conversely, the optimization process can continue to work on the given first cost function by increasing the weights by a small amount. For the second cost function, alternatively, the weights can be increased more aggressively by a much larger factor (e.g., by 1000%), so that the optimization process will actually "focus" more on the second cost function. The mechanism of increasing or decreasing the weights is captured by the weight processing parameters.

[0024] Optionally, the optimization process can iteratively adjust the radiation therapy plan to minimize the selected cost function. That is, the optimization process can iteratively adapt the optimizable parameters of the radiation therapy plan to arrive at a radiation therapy plan that can treat the patient with clinically acceptable accuracy and expected efficacy.

[0025] Optimizable parameters may include a dose excess value, which is the amount of dose violation that is considered acceptable. Optimizable parameters may include beam weights; in a radiation beam, the beam may be divided into a plurality of beamlets, where the contribution under a (hypothetical) unit flux may be determined. The weight of the beamlet is then a scaling factor by which the unit flux of the beamlet may be scaled to reach another flux value. Optimizable parameters may include beam (or beamlet) angles, which may be the angle of the beam relative to a reference point (e.g., a radiation head of a radiotherapy system toward a targeted area). Optimizable parameters may include dose-volume histogram information, which provides information related to the cumulative dose per volume fraction of a target volume or OAR. "Fractions" may be derived using a "fractionalization" process, where a sequence of radiotherapy deliveries (e.g., 45 fractions) is provided over a predetermined period of time, where each therapy delivery includes a specified fraction of the total prescribed dose. Optimizable parameters may include the number of radiation beams. Optimizable parameters may include the dose of each radiation beam. Optimizable parameters may include segment or control point shapes. Optimizable parameters may include segment or control point weights. The optimization process may adapt any or all of these optimizable parameters.

[0026] Optionally, the optimization process may involve multiple objectives. That is, the multiple objectives may be the subject of multi-objective or Pareto optimization. The method may then involve receiving a second reference objective and selecting a cost function associated with the second reference objective from a plurality of cost functions. Of course, the second reference objective may be the same as the first reference objective. For example, the method may seek to optimize the doses involved in a treatment plan for two clinically similar target areas, in which case the same cost function may be appropriate. The selected second cost function may be associated with a second specified weight value, wherein the second specified weight value may correspond to the sensitivity of the example dose distribution to changes in the selected second cost function. The method may then apply the optimization process to a radiation therapy plan for radiation treatment of the patient based on the second reference objective (as well as the first reference objective).

[0027] Optionally, multiple cost functions may include a subset of cost functions, wherein the subset includes at least two cost functions that share a specified weight value. Additionally or alternatively, the at least two cost functions may share specified weight processing parameters. For example, each cost function in a subset may be marked with a flag or pointer indicating that each cost function will be processed in the same manner as each other (weight scaling or weight processing with respect to any cost function value). Of course, multiple subsets may be included, wherein a second subset, for example, may include cost functions that share a certain specified weight value and / or specified weight processing parameters. Example subset labels include "sensitive" and "non-sensitive".

[0028] Optionally, the reference target may include a reference dose value for a targeted area of ​​the patient. Additionally or alternatively, the reference target may include a reference dose value for a surrounding area.

[0029] Optionally, the reference target may include a reference volume value of a targeted area of ​​the patient. Additionally or alternatively, the reference target may include a reference volume value of a surrounding area.

[0030] Alternatively, the cost function may be an equivalent uniform dose cost function, a dose-volume histogram cost function or a partial volume cost function.

[0031] Optionally, the method for radiation therapy planning may include accepting user input to a graphical user interface, the user input including a reference target and / or a selected cost function.

[0032] Optionally, the method for radiation therapy treatment planning may include: outputting parameter values ​​corresponding to the treatment plan. The parameter values ​​may include any or all of the following: number of beams, angles of beams, dose per beam, beamlet weights, segment or control point shapes, segment or control point weights, dose-volume histogram information, dose excess values. In this way, the method may output parameter values ​​(and, of course, configuration settings) for a radiation therapy system, thereby enabling delivery of a radiation plan to a patient.

[0033] Embodiments of another aspect include a data processing apparatus including a memory storing computer-readable instructions and a processor. The processor (or controller circuit) is configured to execute the instructions to perform a computer-implemented method for radiation therapy planning.

[0034] Embodiments of another aspect include a computer program comprising instructions that, when executed by a computer, cause the computer to perform a computer-implemented method for radiation therapy planning.

[0035] Embodiments of another aspect include a non-transitory computer-readable storage medium including instructions that, when executed by a computer, cause the computer to perform a computer-implemented method for radiation therapy planning.

[0036] The disclosed techniques, systems, or operations may be implemented in digital electronic circuitry or in computer hardware, firmware, software, or a combination thereof. These techniques, systems, or operations may be implemented as a computer program or computer program product, i.e., a computer program tangibly embodied in a non-transitory information carrier (e.g., in a machine-readable storage device or in a propagated signal) for performing or controlling the operation of one or more hardware modules. The computer program may be in the form of a stand-alone program, a computer program portion, or a plurality of computer programs, and may be written in any form of programming language (including compiled or interpreted languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a data processing environment.

[0037] The present disclosure is described in specific embodiments or examples. Other embodiments are also within the scope of the appended claims. For example, the steps of the present disclosure can be performed in a different order and still achieve the desired results. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Reference is made, by way of example only, to the accompanying drawings in which:

[0039] Figure 1 is a flow chart of a method for radiation therapy treatment planning in general;

[0040] Figure 2Ais an example optimization process;

[0041] Figure 2B is an example optimization process for application in intensity modulated radiation therapy;

[0042] Figure 3A is a flow chart of a method for radiation therapy planning at a first optimization iteration according to a working example;

[0043] Figure 3B is a flow chart of a method for radiation therapy planning at a subsequent optimization iteration according to a working example; and

[0044] Figure 4A is the expected dose distribution associated with a radiotherapy treatment plan determined according to conventional procedures;

[0045] Figure 4B is an expected dose distribution associated with a radiation therapy plan determined according to an embodiment;

[0046] Figure 5A is another expected dose distribution for spinal cancer treatment associated with a radiation therapy plan determined according to conventional procedures;

[0047] Figure 5B is another expected dose distribution for spinal cancer treatment associated with the radiation therapy plan determined according to the embodiment;

[0048] Figure 6 is a dose-volume histogram comparing the expected doses for spinal cancer treatment planning;

[0049] Figure 7 is a radiotherapy system adapted to implement a radiotherapy plan determined according to an embodiment; and

[0050] Figure 8 It is a radiotherapy device or apparatus suitable for implementing the radiotherapy plan determined according to the embodiment. DETAILED DESCRIPTION

[0051] Treatment planning typically involves performing an optimization process to optimize several parameters with the goal of providing a sufficiently high dose to the targeted region (or Planned Target Volume (PTV)) while reducing the dose to surrounding healthy tissue. A clinically appropriate plan can be obtained by balancing different treatment planning goals, for example by prioritizing treatment planning goals related to dose coverage in the targeted region and then addressing treatment plans related to the remaining surrounding areas. However, with traditional treatment planning methods, users are often required to have advanced knowledge of how the specific cost function used in the optimization process works in terms of the resulting dose distribution. This places an unnecessary burden on the user and can ultimately significantly slow down the treatment planning process.

[0052] Figure 1 A general method for radiation therapy treatment planning is shown that addresses the burden borne by users with cost function expertise. In step 102, a computer receives a reference target that represents some objective to be achieved by a radiation therapy system (and associated radiation therapy process). In step 104, the computer selects a cost function associated with the reference target from a plurality of cost functions. The selected cost function is associated with a specified initial weight value, wherein the specified initial weight value corresponds to the sensitivity of an example dose distribution to changes in the selected cost function. In step 106, the computer applies an optimization process to a radiation therapy treatment plan for a patient's radiation treatment based on the received reference target. The optimization process uses the specified initial weight values ​​for the selected cost function and seeks to minimize the selected cost function.

[0053] Any optimization procedure that accepts a reference target that represents some goal to be achieved by the radiotherapy system and seeks to minimize a cost function (or equivalently, maximize a reward function) is suitable for inclusion in the method according to various embodiments. For example, the method may implement the optimization procedure described in European patent application publication EP 3681600A1.

[0054] The compliance with the objectives and constraints involved in the optimization process can be quantified using a cost function, where the cost function output is a measure of the difference between the dose achieved by the treatment plan (e.g., the dose in a given region) and a reference dose (e.g., the maximum acceptable dose in that region).

[0055] FIG2 shows a block diagram of an example optimization process. The optimization process is applicable to the optimization of a set of parameters (i.e., one or more parameters) for a radiotherapy system. The one or more parameters may be referred to as optimizable parameters, or may be referred to as decision variables. Examples of optimizable parameters include parameters related to the characteristics of the radiation to be delivered by the radiotherapy system, such as the angle of the radiation beam, the level of radiation intensity at each angle, etc. Optimization is typically performed with respect to one or more parameters to shorten the beam on time, improve dose uniformity, etc. An example of a radiotherapy system will be described below.

[0056] The optimization module 200 may include an optimizer 212. Given the anatomical model 204 and one or more parameters 206, the purpose of the optimizer 212 is to provide a set of optimized parameters 216 that minimize the cost function 204. The optimizer includes at least one optimization algorithm. Examples of optimization algorithms include a simplex algorithm, a gradient-based algorithm, or an interior point algorithm, etc. or a combination thereof. Other optimization algorithms are also possible. The optimization example is shown in FIG. Training Guide, Document ID: LTGMON0530 (Elekta AB).

[0057] The cost function value 214 is the value of the cost function 204 when evaluated using the set of optimization parameters 216. The cost function value 214 is calculated by the optimizer 212 during optimization. The cost function value 214 may be referred to as a penalty. Optionally, the cost function value 214 is output by the optimizer 212.

[0058] The model 202 is a representation of the physical problem. The model 202 may correspond to an anatomical structure or a region thereof. The model 202 may include a targeted region (or PTV) and surrounding regions of healthy tissue or other critical structures (referred to as organs at risk or OARs). These regions may be determined and / or defined during a process called segmentation performed on a two-dimensional or three-dimensional image or model of the anatomical structure. These regions may be in the form of shell structures, which are structures generated to tune the dose delivered to tissue surrounding the target region, and they may be used to control dose distribution conformality. The dose distribution may be determined by a pencil beam algorithm, a convolution-based algorithm, and / or a Monte Carlo (MC)-based algorithm. More details on how the dose may be determined are provided in Training Guide, Document ID: LTGMON0530 (Elekta AB).

[0059] The one or more parameters 206 can be optimizable parameters (also called decision variables) for which the optimizer 212 attempts to find an optimal value. The one or more parameters 206 can be initialized to predetermined values. In one example, for Intensity-Modulated Radiation Therapy (IMRT), the parameters 206 are weights of beamlets, and the parameters 206 can be initialized to predetermined weights. The weights of beamlets can be mapped to actual beamlet intensities during radiation therapy, where the intensity can be changed, for example, by changing the aperture size or the pulse time. Alternatively, the parameters 206 can be initialized to values ​​determined in a previous step. For example, when performing further optimization procedures (or passes), the parameters can be initialized to values ​​determined in previous optimization procedures (or passes).

[0060] The cost function 204 is a mathematical expression relating the parameters 206 and the model 202. The cost function 204 may be referred to as an objective function. The cost function 204 relates the dose distribution in the model 202 to a single value, the cost function value 214. For example, the cost function value 214 defines a penalty for violating an objective and / or constraint. The penalty may be evaluated by the optimizer 212 during optimization. Optionally, in constrained optimization, only the objective contributes to the cost function (i.e., the constraints remain "fixed" and the optimization process does not allow for constraint violations, since violating the constraint 208 would simply result in the treatment plan being rejected).

[0061] Constraints 208 include one or more conditions that must be satisfied by the optimization parameters 216. Constraints may include hard constraints (conditions that the parameters 206 need to satisfy). Constraints limit the set of available solutions. Constraints are used to define what is physically or clinically acceptable, rather than mathematically possible. For example, for IMRT, a hard constraint may be that the weights of beamlets must be non-negative (since negative beamlet weights are impossible). Note that constraints 208 are an optional feature.

[0062] Reference target 210 includes what optimizer 212 is trying to achieve. For example, in IMRT, reference targets can relate to dose-based targets and / or volume-based targets. Dose-based targets can include dose values ​​(in Gy), such as the maximum, minimum, or average dose to a given area. For example, a dose-based target can be set in units of 10 mm. 3 The dose absorbed by the volume should not exceed 25 Gy. Volume-based targets can include relative volumes (fractions or percentages of volume) or absolute volumes (or in the case of two dimensions, absolute areas). For example, a volume-based target might state that the dose absorbed by 50% of the OAR volume should not exceed 5 Gy. Volumes (or areas) can be defined in terms of physical dimensions (e.g., in mm 2 、cm2 、mm 3 、cm 3 In units). Additionally, the reference target 210 may include an indication of the cost function 204 to be used. Further, the reference target 210 may include an indication of the anatomical structure applicable to the objective to be achieved. The optimizer 212 will then determine the optimization parameters 216 that will result in an achieved objective value 218 close to the reference target 210. The reference target may be referred to as a treatment plan objective. Note that the reference target 210 is an optional feature.

[0063] Thus, reference target 210 may be anatomy-specific functions that establish dose and / or bioresponse targets. Constraints 208 are anatomy-specific functions that must be satisfied in order to achieve optimized fusion. When constraints are used with targets, the constraints are always satisfied, while the targets may not necessarily be satisfied (rather, they are goals that optimizer 212 attempts to achieve). In the case of multiple targets, priorities or preferences may be assigned to the targets, for example, a primary target is considered in a first optimization run, a secondary target is considered in a second optimization run following the first optimization run, a tertiary target is considered in a third optimization run following the second optimization run, and so on.

[0064] The cost function 204 together with the model 202, parameters 206, constraints 208 (optional), and reference target 210 may be considered to represent an optimization problem to be solved. The cost function 204, constraints 208, reference target 210, alone or in combination, may be referred to as a treatment planning goal.

[0065] The optimizer 212 aims to find a set of optimization parameters 206 that minimizes the cost function 204. The optimization parameters 216 are the output of the optimizer 212.

[0066] Optionally, optimizer 212 outputs cost function value 214. Cost function value 214 is the value of cost function 204 when evaluated using optimization parameters 216.

[0067] Optionally, the optimizer 212 outputs a target value 218. The target value 218 is comparable to a reference target. The target value 218 represents a value having the same unit as a given reference target. The target value 218 may be different from the cost function value 214. Unlike the cost function value 214, which may be a numerical value of an evaluation of a cost function, the target value 218 may have a physical meaning. In one example, when the reference target includes a target for dose measurement, the target value is also related to the dose (e.g., it has the unit Gy and / or the same physical meaning as the reference target). In one example, when the reference target includes a reference volume and a dose value, the target value is also related to the reference volume and the dose value (i.e., it has the same physical meaning as the reference target). The cost function value 214 may be a numerical value corresponding to a cost function evaluated using a set of optimization parameters. The target value 218 will be a percentage of the volume that receives a predetermined dose. In other words, the target value 218 is comparable to the reference target.

[0068] Figure 2B An example of a two-stage optimization process 250 for IMRT is shown. Although this example is applied to IMRT, it should be noted that the optimization process can be applied to other radiation therapy modes. For example, the optimization process can also be applied to Volumetric Modulated Arc Therapy (VMAT).

[0069] In IMRT, one or more radiation beams are directed to the tumor (targeted area). The intensity of each beam profile is not uniform. The goal of the optimization process 250 is to modify the intensity distribution so that a sufficiently high dose is delivered to the tumor while reducing the dose delivered to healthy organs (surrounding areas).

[0070] The optimization method of IMRT 250 includes two stages. The first stage 252 is Fluence Map Optimization (FMO), and the second stage 254 is to determine the configuration of the radiotherapy system. In FMO (stage 252), the optimal flux map is determined. Then, in the second stage, the optimal flux map is used to determine the configuration of the radiotherapy system.

[0071] For FMO, each beam is divided into a number of beamlets. Then, the contribution of each beamlet to the voxel per unit flux is calculated. During optimization, the weight of each beamlet is adjusted to minimize the cost function. The total dose distribution can be obtained by multiplying the weight with the contribution of each beamlet per unit flux. The total dose distribution can be compared with any constraints and / or reference targets to determine whether the optimization solution is appropriate.

[0072] Optimization consists in minimizing the cost function f(x) by determining suitable values ​​of the parameters x based on certain constraints g. For FMO, the parameters x correspond to the weights of the beamlets (the parameters x may also be referred to as decision variables). The output of FMO (stage 252) includes the parameters x.

[0073] The constraints g include limits. Examples of limits are the minimum dose at a voxel corresponding to a target, or the maximum dose at a voxel corresponding to an OAR.

[0074] In one example, the cost function f(x) is:

[0075] f(x)=T1+T2+…+T3, (Equation 1)

[0076] Among them, for example, T1 = ∑x n ·d n , where T1 = ∑x n ·d n Corresponding to the weight of the sub-bundle, d n It represents the dose delivered to a voxel per beam at unit intensity. n is a non-optimizable parameter (e.g., it may depend on machine configuration and / or organizational properties). n Corresponding to the above model 202. In one example, the dose d n It can be determined by a pencil beam algorithm, a convolution-based algorithm, and / or a Monte Carlo (MC)-based algorithm. Optionally, the dose calculation uses a pencil beam algorithm or a convolution-based algorithm (which is faster but has reduced accuracy). n More details are in Training Guide, Document ID: LTGMON0530 (Elekta AB) is provided. The output of Phase 1 may include dose distribution.

[0077] In this example, the cost function f(x) is the sum of the total doses (T1, T2, ..., T3), where each T1, T2, ..., T3 represents the dose on a different structure. Note that alternative formulations of the problem to be solved can be used (e.g., by defining different cost functions or constraints).

[0078] In one example, the requirement that the dose at the target be above a certain amount can be stated as a constraint. Alternatively, such a requirement can be stated as a reference target.

[0079] Step 252 provides a combination of optimized parameters (e.g., beamlet weights x) that will provide an optimal intensity map (or flux map). In order to be ready for delivery by a radiotherapy system, a further stage (stage 153) is required to convert each optimized beamlet weight into a machine configuration that will deliver the optimal flux.

[0080] In stage 254, the output from stage 252 is converted into a configuration of a radiotherapy system. The configuration can be used by a radiotherapy device (examples of which are described herein) to deliver radiotherapy. For example, the configuration of the radiotherapy system includes a set (i.e., one or more) of through-hole configurations. The shape of the through-hole configuration can be selected to meet the same goals as in the first stage. The through-hole configuration can be achieved by a beam shaping device.

[0081] The through-hole configuration may be referred to as a control point or segment. The control point and / or segment includes radiation information (eg, energy, dose) and geometric information (segment or control point shape), such as gantry angle and blade position.

[0082] The shape and weights of the vias (ie, segment or control point weights) may be determined by applying algebraic and trigonometric considerations to the placement of the vias to implement the optimized beamlet weights of stage 252 .

[0083] Alternatively, stage 254 includes performing a second stage optimization flow to determine an optimized via configuration that will implement the optimized flux pattern determined in stage 252. The second stage optimization flow may be referred to as via optimization or via refinement.

[0084] Via optimization can include the following:

[0085] - A set of receive beamlet weights (e.g., from stage 252)

[0086] - Perform optimization to determine the optimal via shape. For example, optimization can be performed using Figure 2B The optimization process shown in is performed.

[0087] Other optimization processes are possible.

[0088] When the beam shaping device includes a multi-leaf collimator (MLC), through-aperture optimization may include:

[0089] - Receive a set of beamlet weights from the FMO and / or flux profile (eg, from stage 252).

[0090] - Convert the received profile into beamlet widths (openings between blade pairs). This results in segmentation.

[0091] - Optimize the weights of the resulting segments.

[0092] - Optionally, optimize the resulting segmented shapes (using a process called segmented shape optimization).

[0093] Note that the optimization step in stage 254 may include calculation of a dose distribution. The dose distribution may be calculated using a pencil beam algorithm, a convolution-based algorithm, and / or a Monte Carlo (MC)-based algorithm. Optionally, the dose calculation in stage 254 uses an MC-based algorithm (which is more accurate but computationally expensive). Further details of through-hole optimization are provided in Training Guide, Document ID: LTGMON0530 (Elekta AB).

[0094] Notice, Figure 2B The process may involve a specific angle (radiation beam angle). In one approach, a previous stage may be to determine the angle at which the radiation is to be delivered. The Beam Angle Optimization (BAO) process may be performed in Figure 2B , then Figure 2B The process can be performed for each determined angle.

[0095] Although Figure 2B The example involves a two-stage optimization process for IMRT (FMO at stage 252 and determination of configuration at stage 254), but it should be noted that alternative optimization methods such as Direct Machine Parameter Optimization (DMPO) can be used. In DMPO, the decision variables x correspond to parameters of the machine delivering radiation (e.g., MLC blade positions).

[0096] Returning to the cost function 202 in the optimization process, examples of cost functions include: target EUD, target penalty, secondary overdose, secondary underdose, serial, parallel, maximum dose, overdose DVH, underdose DVH. The cost function is briefly defined as follows. Each cost function provides a different calculation method. When the cost function includes any one of target penalty, parallel, overdose DVH and underdose DVH, the reference target is a dose value and a relative volume component (e.g., a percentage) or an absolute volume (e.g., in cc).

[0097] When the cost function includes any of target EUD, secondary overdose, seriality, maximum dose, or conformality, the reference target is the dose value.

[0098] Typically, the optimization process seeks to minimize a composite cost function consisting of the sum of cost functions that may be objectives or constraints. The cost function relates the non-uniform dose distribution to a single value. This value can be used to define a penalty for violating the objective or constraint. The system can evaluate this penalty value during the optimization process.

[0099] Cost functions such as serial, parallel, quadratic overdose, overdose DVH, or maximum dose are intended to limit the dose at the anatomical structure (e.g., if the dose exceeds a threshold, a penalty value is applied). Cost functions such as quadratic underdose or underdose DVH are intended to increase the dose at the anatomical structure. Cost functions such as target EUD and target penalty are intended to increase the dose at the structure. Some of the above cost functions also take a unitless number as input. The unitless number (k) can be, for example, a power law exponent.

[0100] The Target EUD cost function defines the structure as a target volume and represents the probability that the targeted cells will survive a given dose. The cost function requires a specified dose as input (i.e., the reference target of the cost function is the dose value). The specified dose (in Gy) is the Equivalent Uniform Dose (EUD). EUD is a uniform dose that, if delivered at the anatomical structure, has the same clinical effect as provided by a non-uniform dose distribution.

[0101] The target penalty cost function takes a prescribed dose and a minimum volume as input (i.e., the reference targets of the cost function are dose and relative volume). The target penalty is a quadratic penalty that starts with a threshold dose. It produces a steeper dose gradient after the target threshold is met. The target penalty is used to define a requirement that at least some fraction of the anatomical volume should receive at least the target dose.

[0102] The Quadratic Overdose (QO) cost function is a cost function used to limit the dose in the structure to which it is applied. QO can be applied to a target or an OAR. The QO cost function can be used to limit hot spots in a target. The QO cost function takes the maximum dose and the root-mean square (RMS) dose exceedance as input (i.e., the reference target includes two dose values). The maximum dose defines the dose that will trigger a penalty if exceeded. The RMS dose exceedance defines the acceptable amount of violation.

[0103] The quadratic underdose cost function is a cost function applied to the target volume. The quadratic underdose function implements a quadratic penalty. The cost function takes as input the minimum dose in Gy and the underdose in Gy (i.e., the reference target includes two dose values). The minimum dose is the minimum dose allowed in the target volume, and it represents the dose below which a penalty is triggered. The underdose is similar to the RMS dose overrun in that it defines the amount of acceptable violation in the prescription.

[0104] The serial cost function (or equivalent uniform dose cost function) is often used with serial OARs (e.g., an example of an insensitive cost function, especially when a lower p-value is used). Serial anatomical structures are those that will be harmed by high doses even if they are confined to a small volume. Examples are the spinal cord and the intestine. This cost function imposes large penalties on hot spots, even if they are small in volume. The cost function takes as input the EUD in Gy and a power law exponent p (i.e., the reference target includes both the dose value and a unitless number). An example mathematical expression for the serial EUD cost function is as follows:

[0105]

[0106] where the cost function represents the sum of the i voxels in the patient's volume V, and where V i is the volume of the ith voxel, D i is the dose delivered to the i-th voxel, Δ is a reference dose value, which is the maximum radiation dose in the target area of ​​the patient and / or the surrounding area of ​​the patient, and ε is a constant. The inventors have realized that the sensitivity of the cost function (relative to the resulting dose map) can be controlled in part using a power law exponent. For example, p<3 typically results in an insensitive EUD cost function (i.e., the EUD cost function value does not change significantly with a significant change in the resulting dose map, and vice versa).

[0107] Parallel cost functions (or partial volume cost functions; an example of an insensitive cost function) are often used for parallel OARs. Parallel structures are those for which very high doses to a small volume can be tolerated if the rest of the organ is preserved. Examples are lungs, parotid glands, kidneys, liver. The cost function takes as input a reference dose in Gy, the mean organ damage (i.e., a fraction of the volume of the structure that may be sacrificed), and a power law exponent p (i.e., the reference target includes dose values, relative volumes, and unitless numbers). The cost function further takes a mean organ damage parameter ζ in percentage form PV, which is biologically equivalent to the fraction of the volume of the structure that can be sacrificed. For example, the function of any organ can be divided into a large number of subvolumes (functional subunits, tissue repair units) that work in parallel. The effect of radiation is considered to knock down the subvolumes, thereby reducing the function of the organ. As a result, the function is degraded on a continuous scale. Since each subvolume has a continuous dose response (that is, each subvolume can lose function continuously), 50% of the organ function can be lost in two extreme ways: either 50% of the organ volume is irradiated to a dose that causes it to lose function, or the entire organ is irradiated to a dose that causes 50% of the function of each subvolume to lose function. Therefore, the total damage to the organ is the average of the functional loss of all subvolumes. An example mathematical expression of a partial volume (PV) parallel cost function is as follows:

[0108]

[0109] where the cost function represents the sum of the i voxels in the patient's volume V and where V i is the volume of the ith voxel, D i is the dose delivered to the i-th voxel, and Δ is a reference dose value, which is the maximum radiation dose in the target area of ​​the patient and / or the surrounding area of ​​the patient.

[0110] The Maximum Dose (MXD) cost function is effectively a hard barrier that can be applied to the targeted region (structure) or the surrounding region (OAR). The Maximum Dose cost function has a penalty that takes effect whenever a voxel exceeds the maximum dose threshold. The cost function takes as input the maximum dose in Gy (i.e., the reference target includes the dose value). An example mathematical expression for the MXD cost function is as follows:

[0111]

[0112] where the cost function represents the sum of the i voxels in the patient's volume V, and where V i is the volume of the ith voxel, D i is the dose delivered to the i-th voxel, Δ is a reference dose value which is the maximum radiation dose in the target region of the patient and / or the surrounding region of the patient, and α is a maximum dose threshold.

[0113] The overdose DVH cost function (an example of a non-sensitive cost function) converts the target dose in Gy and the maximum volume ζ ODVHAs input (i.e., a reference target including dose values ​​and relative volumes). This cost function is applied to the OAR. The goal is to keep the volume that receives more than the target dose below the relative volume. For example, if it is desired that no more than 50% of the structure receive a dose in excess of 60 Gy, this cost function can be used. An example mathematical expression for Overdose DVH (ODVH) is as follows:

[0114]

[0115] Among them, b i =(D i -Δ+d o ) / d o

[0116] where the cost function represents the sum of the i voxels in the patient's volume V, and where V i is the volume of the ith voxel, D i is the dose delivered to the i-th voxel, Δ is a reference dose value, which is the maximum radiation dose in the target area of ​​the patient and / or the surrounding area of ​​the patient, and d o is a constant or a fraction of Δ. As an example, a hard-coded d0 value of 0.04×Δ has been shown to be effective.

[0117] The underdose DVH cost function (an example of a non-sensitive cost function) converts the target dose in Gy and the minimum volume ζ UDVH As input (i.e., a reference target including dose values ​​and relative volumes). This cost function is applied to the target. The goal is to keep the targeted volume that receives less than the target dose greater than a minimum volume. For example, if it is desired that at least 80% of the structure receive a dose of at least 60 Gy, this cost function can be used. An example mathematical expression for Underdose DVH (UDVH) is as follows:

[0118]

[0119] Among them, b i =(-D i +Δ+d o ) / d o

[0120] where the cost function represents the sum of the i voxels in the patient's volume V, and where V i is the volume of the ith voxel, D i is the dose delivered to the i-th voxel, Δ is a reference dose value, which is the maximum radiation dose in the target area of ​​the patient and / or the surrounding area of ​​the patient, and d ois a constant or a fraction of Δ. As an example, a hard-coded d0 value of 0.04×Δ has been shown to be effective.

[0121] Further details of the cost function are given in Training Guide, Document ID: LTGMON0530 (Elekta AB).

[0122] The inventors have realized that the EUD cost function, the PV cost function, the ODVH cost function and the UDVH cost function are good examples of non-sensitive cost functions. Of course, other types of cost functions can also be good candidates for non-sensitive cost functions.

[0123] Figure 3A A method for radiotherapy planning is shown, which involves cost function processing at the beginning of the optimization process. Consider the case of prostate cancer radiotherapy planning. The user may want to have a radiotherapy plan that can deliver at least 60Gy to at least 80% of the tumor in the prostate region of the patient. After entering this reference target, in step 310, the user can then select a suitable cost function (e.g., an overdose DVH cost function as described above).

[0124] At step 320, the computer implementing the method then determines whether the selected cost function is classified as non-sensitive (i.e., the cost function value is not expected to change significantly with a significant change in the resulting dose map, or vice versa). This classification can be performed during the course of the method, or it can be predetermined - for example, the determination at step 320 can involve using a lookup table to identify known cost functions that have been classified as non-sensitive.

[0125] In the event that the cost function is classified as sensitive (ie, step 320: No), the computer may then instruct to perform the optimization process using default initial weights at step 330. Here, the default initial weights may simply be 1, in which case the cost function value is not scaled.

[0126] In the case where the cost function is classified as non-sensitive (i.e., step 320: yes), the computer may then instruct to perform the optimization process using the increased initial weights at step 340. For example, the optimization process may be performed while taking into account that the cost function value is scaled by a factor of 2. In this way, unnecessary extension of the optimization process is avoided, and the optimization process may reach convergence faster.

[0127] In step 350, the computer initializes the optimization process using initial weights according to the selected cost function.

[0128] Figure 3BA method for radiation therapy treatment planning is shown that involves processing of a cost function during an iterative optimization process. As described above, the user can then select an appropriate cost function at step 310. In this example, assume that an optimization process is ongoing (e.g., according to Figure 3A During subsequent optimization iterations (ie, not the initial iteration), the computer implementing the method then determines, at step 360, whether the selected cost function is classified as non-sensitive.

[0129] In the event that the cost function is classified as sensitive (i.e., step 360: No), the computer may then instruct the optimization process to modify the weight values ​​from the previous optimization iteration using the default weight processing parameters, at step 370. Here, for example, the default weight processing parameters may simply be "decreased by 0.01", in which case - and given that the weight value in the previous optimization iteration was 1 - the weight value is reduced to 0.99, and the cost function value is scaled by 0.99.

[0130] In the event that the cost function is classified as non-sensitive (i.e., step 360: yes), in step 380, the computer can then instruct to perform the optimization process using increased weight processing parameters. For example, the weight processing parameters associated with the non-sensitive cost function (and, for example, all non-sensitive cost functions) can be "reduced by half." Consider again that the weight value in the previous optimization iteration was 1, and then the weight value is reduced to 0.5, and the cost function value is scaled by 0.5. By applying a more aggressive weight processing, the method compensates for the cost function being non-sensitive. In this way, during optimization, the optimization process automatically reverses direction by increasing or decreasing the weight of individual objectives depending on the sensitivity level of the associated cost function value.

[0131] In step 350 , the computer continues the optimization process using the adjusted weights adjusted according to the weight processing parameters of the selected cost function.

[0132] Figure 4A It is a dose map of a patient, which is overlaid on a medical image of the patient acquired using a computed tomography, CT scan. The diagram shows the dose distribution in the form of a color painting, where different colors represent different doses. Generally, darker colors indicate higher doses and vice versa. If two locations have exactly the same color, the dose in these two locations is the same. The diagram includes isodose lines, where each of these connecting curves shows a certain dose level. Any position at the periphery of a particular curve has the same dose. For example, one isodose curve may represent the dose at 95% of the defined dose value, while another isodose curve may represent the dose at 80% of the defined dose value.

[0133] The dose map indicates the expected radiation dose according to the radiotherapy plan (determined using the optimization process) for treating tumors in the oropharynx while seeking to minimize the dose delivered to the posterior neck. The radiotherapy plan is determined using a traditional process. As shown in the figure, the target area (which surrounds the gross and microscopic tumors and potentially affected lymph nodes) is the area delimited by the line marked with "A". The surrounding area of ​​concern is the area within the area delimited by the dotted line marked with "B", which includes the posterior neck. A peak dose value of 49.22Gy is delivered to the target area. In the surrounding area of ​​concern, the expected undesired dose in the central part is about 35Gy.

[0134] Figure 4B 1 is a dose map for the same patient, where the associated radiotherapy plan is determined using the method according to the embodiment instead. As shown, the expected dose in the target area (again demarcated by the line marked "A") is essentially unchanged; the average dose to the parotid gland is only 4Gy different compared to the radiotherapy plan determined using the conventional process. However, the expected dose delivered to the surrounding area (area "B") is significantly reduced.

[0135] Figure 5A is another example dose map for a patient, which is overlaid on a medical image of the patient acquired using a computed tomography scan, CT scan. In this example, the patient is undergoing treatment for spinal cancer, and therefore the targeted area is in the spine (the area towards the center in the lower part of the dose map). The targeted area is bounded by curve 10. In this scan, the patient's healthy lungs form two peripheral regions (OARs), roughly outlined by crescent-shaped gray lines; the peripheral region of the right lung is bounded by curve 12, and the peripheral region of the left lung is bounded by curve 14 (note that the image is flipped left to right). An ideal treatment plan will limit the dose in the peripheral region (lungs) while delivering the expected dose to the targeted area (spine).

[0136] In this example, both the left and right peripulmonary regions are subjected to the optimization process of the serial cost function with p=1, and in the case of the parallel cost function with p=2, considering the maximum prescribed dose of 5 Gy to be provided. Both cost functions are examples of non-sensitive cost functions. The dose map of this example is determined using a conventional method that does not consider the cost function weight values ​​and / or weight value processing parameters.

[0137] As shown, the optimization process seeks to minimize the two selected cost functions and provides optimizable parameter values ​​suitable for delivering the displayed dose map. Iso-dose lines (and colored painted lines, although this aspect is lost in the grayscale rendering) radiating outward from the targeted region represent the expected dose levels at 90%, 80%, 70%, etc. of the reference target (prescribed dose). As shown, in this example, the dose map has good conformality to the targeted region, but there is still a large spread of expected dose in the left and right lung peripheries.

[0138] Figure 5B is a dose map for the same patient, where an associated radiation therapy plan is determined using a method according to an embodiment instead. That is, the dose map is determined using a method that takes into account cost function weight values ​​and / or weight value processing parameters. The isodose lines are equivalent to the isodose lines shown in the dose map determined using conventional methods in both absolute and relative terms (e.g., Figure 5A The 90% isodose line in Figure 5B ). As shown in the figure, under the weight values ​​and processing parameters selected according to the embodiment, the radiotherapy plan significantly reduces the expected dose in the left lung periphery and the right lung periphery, while minimizing the change in the dose to the target area, relative to the treatment plan determined using the conventional process.

[0139] Figure 6 Draw the Figure 5A and Figure 5B Dose-volume (DVH) and spinal cancer radiotherapy plans for the same patient shown in . The solid lines (curves 20, 30, 40) correspond to the treatment plans determined without processing the weights of the non-sensitive cost function (conventional process). The dotted lines (curves 22, 32, 42) correspond to the treatment plans determined with processing the weights of the non-sensitive cost function. Curves 20 and 22 are DV curves for the targeted area (spinal tumor). Curves 30 and 32 are DV curves for the left lung periphery. Curves 40 and 42 are DV curves for the right lung periphery.

[0140] As shown, the target dose volumes for the two plans are very similar, with each plan delivering approximately 30 Gy to the targeted region. The expected doses to the left and right lungs associated with the conventional method without weight processing are significantly higher than the expected doses associated with the treatment plan determined using the method according to the embodiment. For example, for the right lung peripheral region (curves 40, 42), without weight processing, approximately 25% of the right lung receives a dose of approximately 10 Gy. With weight processing, only approximately 10% of the right lung receives a dose of approximately 10 Gy. Accordingly, automatic processing of the cost function that takes into account the sensitivity of the cost function is shown to improve the accuracy of the generated treatment plan.

[0141] Figure 7 is a block diagram of an embodiment of a radiation therapy system 500 suitable for performing a method for radiation therapy treatment planning according to an embodiment. The example radiation therapy system 500 includes a computing system 510 within which a set of instructions for causing the computing system 510 to perform any one or more of the methods (or steps thereof) discussed herein can be executed. The computing system 510 can implement a treatment planning system. The computing system 510 can also be referred to as a computer. In particular, the methods described herein can be implemented by a processor or controller circuit 511 of the treatment planning system 510.

[0142] Computing system 510 should be considered as including any number or combination of machines, such as (one or more) computing devices, which independently or jointly execute a group (or multiple groups) of instructions to perform any one or more methods discussed herein. That is, in a computing system, hardware and / or software can be provided in a single computing device, or distributed in multiple computing devices. In some embodiments, one or more elements of a computing system can be connected (e.g., networked) to other machines such as a local area network (LAN), an intranet, an extranet, or the Internet. One or more elements of a computing system can be operated with the ability of a server or client machine in a client-server network environment, or operated as a peer-to-peer (peer-to-peer) (or distributed) network environment. One or more elements of the computing system may be a personal computer (PC), a tablet computer, a set-top box (STB), a personal digital assistant (PDA), a cellular phone, a web appliance, a server, a network router, a switch or a bridge, or any machine capable of executing a set of instructions (sequential or otherwise) that specify actions to be taken by the machine.

[0143] The computing system 510 includes a controller circuit 511 and a memory 513 (e.g., a read-only memory (ROM), a flash memory, a dynamic random access memory (DRAM), such as a synchronous DRAM (SDRAM) or a Rambus DRAM (RDRAM), etc.). The memory 513 may include a static memory (e.g., a flash memory, a static random access memory (SRAM), etc.) and / or an auxiliary memory (e.g., a data storage device) that communicate with each other via a bus (not shown).

[0144] The controller circuit 511 represents one or more general-purpose processors, such as a microprocessor, a central processing unit, or an accelerated processing unit. More specifically, the controller circuit 511 may include a complex instruction set computing (CISC) microprocessor, a reduced instruction set computing (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, a processor implementing other instruction sets, or a processor implementing a combination of instruction sets. The controller circuit 511 may also include one or more special processing devices, such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a digital signal processor (DSP), or a network processor. One or more processors of the controller circuit may have a multi-core design. The controller circuit 511 is configured to execute processing logic for executing the operations and steps discussed herein.

[0145] The computing system 510 may also include a network interface circuit 515. The computing system 510 may be communicatively coupled to an input device 520 and / or an output device 530 via an input / output circuit 516. In some embodiments, the input device 520 and / or the output device 530 may be an element of the computing system 510. The input device 520 may include an alphanumeric input device (e.g., a keyboard or a touch screen), a cursor control device (e.g., a mouse or a touch screen), an audio device (e.g., a microphone), and / or a tactile input device. The output device 530 may include an audio device (e.g., a speaker), a video display unit (e.g., a liquid crystal display (LCD) or a cathode ray tube (CRT)), and / or a tactile output device. In some embodiments, the input device 520 and the output device 530 may be provided as a single device, or as separate individual devices.

[0146] In some embodiments, the computing system 510 may include an image processing circuit 514. The image processing circuit 514 may be configured to process image data 580 (e.g., image or imaging data), such as medical images acquired from one or more imaging data sources, the treatment device 550, and / or the image acquisition device 540. The image processing circuit 514 may be configured to process or pre-process the image data 580. For example, the image processing circuit 514 may convert the received image data to a specific format, size, or resolution, etc. In some embodiments, the image processing circuit 514 may be combined with the controller circuit 511.

[0147] In some embodiments, the radiotherapy system 500 may also include an image acquisition device 540 and / or a treatment device 550. The image acquisition device 540 and the treatment device 550 are provided as a single device. In some embodiments, the treatment device 550 is configured to perform imaging, for example, in addition to providing treatment and / or during treatment. The treatment device 550 comprises the main radiation delivery component of the radiotherapy system.

[0148] The image acquisition device 540 can be configured to perform positron emission tomography (PET), computed tomography (CT), magnetic resonance imaging (MRI), single positron emission computed tomography (SPECT), X-ray, etc.

[0149] Image acquisition device 540 may be configured to output image data 580, which may be accessed by computing system 510. Treatment device 550 may be configured to output treatment data 560, which may be accessed by computing system 510.

[0150] The computing system 510 may be configured to access or acquire treatment data 560, planning data 570, and / or image data 580. The treatment data 560 may be acquired from an internal data source (e.g., from the memory 513) or an external data source (e.g., the treatment device 550 or an external database). The planning data 570 may be acquired from the memory 513 and / or an external source (e.g., a planning database). The planning data 570 may include information acquired from one or more of the image acquisition device 540 and the treatment device 550.

[0151] The above methods can be implemented by a computer program. The computer program may include computer code (e.g., instructions) arranged to instruct a computer to perform the functions of one or more of the above methods. Figure 1 , FIG. 2 and / or FIG. 3, the steps of each method described can be performed by computer code. The steps of the above method can be performed in any suitable order. The computer program and / or code for performing these methods can be provided to a device (e.g., a computer) on one or more computer-readable media (or more generally, a computer program product). The computer-readable medium can be transient (transitory) or non-transitory (non-transitory). One or more computer-readable media can be, for example, electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, or propagation media for data transmission (e.g., for downloading code via the Internet). Optionally, one or more computer-readable media can take the form of one or more physical computer-readable media, such as semiconductor or solid-state memory, magnetic tape, removable computer disk, random access memory (Random Access Memory, RAM), read-only memory (Read-Only Memory, ROM), rigid disk and optical disk, such as CD-ROM, CD-R / W or DVD. The instructions may also reside, in whole or at least in part, within the memory 513 and / or within the controller circuit 511 during execution thereof by the computing system 510 , the memory 513 and the controller circuit 511 also constituting computer-readable storage media.

[0152] In one embodiment, the modules, components, and other features described herein may be implemented as discrete components or integrated into the functionality of hardware components such as ASICS, FPGA, DSP, or similar devices.

[0153] A "hardware component" is a tangible (e.g., non-transitory) physical component (e.g., a set of one or more processors) that is capable of performing certain operations and may be configured or arranged in some physical manner. A hardware component may include dedicated circuitry or logic that is permanently configured to perform certain operations. A hardware component may include a dedicated processor, such as an FPGA or ASIC. A hardware component may also include programmable logic or circuitry that is temporarily configured by software to perform certain operations.

[0154] In addition, modules and components can be implemented as firmware or functional circuits in hardware devices. In addition, modules and components can be implemented with any combination of hardware devices and software components, or only with software (e.g., code stored or otherwise embodied in a machine-readable medium or a transmission medium).

[0155] Figure 8A radiotherapy device suitable for implementing a radiation treatment plan determined according to an embodiment is depicted. A cross section through the radiotherapy device 600 includes a radiation head 610 and a beam receiving device 602, both of which are attached to a gantry 604. The radiation head 610 includes a radiation source 612 that emits a radiation beam 606. The radiation head 610 also includes a beam shaping device 618 that controls the size and shape of the radiation field associated with the beam.

[0156] The beam receiving device 602 is configured to receive radiation emitted from the radiation head 610 for the purpose of absorbing and / or measuring the radiation beam. In the illustrated view, the radiation head 610 and the beam receiving device 602 are positioned diametrically opposite each other.

[0157] The gantry 604 is rotatable and supports the radiation head 610 and the beam receiving device 602 so that they can rotate about a rotation axis 608, which can coincide with the patient longitudinal axis. The gantry provides rotation of the radiation head 610 and the beam receiving device 602 in a plane perpendicular to the patient longitudinal axis (e.g., a sagittal plane). Three gantry directions X can be defined. G , Y G , Z G , where Y G Direction perpendicular to the frame rotation axis. G The direction extends from the point on the gantry corresponding to the radiation head toward the gantry rotation axis. Therefore, with respect to the patient reference frame, Z G The direction changes as the rack rotates.

[0158] The radiation therapy device 600 also includes a support surface 620 on which a subject (or patient) is supported during radiation therapy. The radiation head 610 is configured to rotate about the rotation axis 608 so that the radiation head 610 directs radiation toward the subject from various angles around the subject to establish a prescribed dose of radiation at the targeted area while spreading the radiation dose received by healthy tissue to a larger area of ​​healthy tissue.

[0159] The radiotherapy apparatus 600 is configured to deliver a radiation beam toward a radiation isocenter that is substantially located on the rotation axis 608 at the center of the gantry 604 regardless of the angle at which the radiation head 610 is positioned.

[0160] The rotatable gantry 604 and radiation head 610 are sized to allow for the presence of a central aperture 622. The central aperture 622 provides an opening sufficient to allow a subject to be positioned therethrough without the possibility of the subject being accidentally contacted by the radiation head 610 or other mechanical components as the gantry rotates the radiation head 610 about the subject.

[0161] The radiation head 610 emits a radiation beam 606 along a beam axis 624 (or radiation axis or beam path), wherein the beam axis 624 is used to define the direction in which the radiation head emits radiation. The radiation beam 606 is incident on a beam receiving device 602, which may include at least one of a beam blocker and a radiation detector. The beam receiving device 602 is attached to the gantry 604 on radially opposite sides of the radiation head 610 to attenuate and / or detect the radiation beam after the beam passes through an object.

[0162] Radiation beam axis 624 may be defined, for example, as the center or point of maximum intensity of radiation beam 606 .

[0163] The beam shaper 618 defines the spread of the radiation beam 606. The beam shaper 618 is configured to adjust the shape and / or size of the radiation field generated by the radiation source. The beam shaper 618 does this by defining a variable shaped through hole (also referred to as a window or opening) to collimate the radiation beam 606 to a selected cross-sectional shape. In this example, the beam shaper 618 may be provided by a combination of an aperture and an MLC. The beam shaper 618 may also be referred to as a beam modifier.

[0164] The radiation therapy device 600 can be configured to deliver radiation therapy treatments in a coplanar mode and a non-coplanar mode (also called an oblique mode). In coplanar treatment, radiation is emitted in a plane that is perpendicular to the axis of rotation of the radiation head 610. In non-coplanar treatment, radiation is emitted at an angle that is not perpendicular to the axis of rotation. To deliver coplanar treatment and non-coplanar treatment, the radiation head 610 can be moved between at least two positions, one position in which radiation is emitted in a plane perpendicular to the axis of rotation (coplanar configuration) and one position in which radiation is emitted in a plane that is not perpendicular to the axis of rotation (non-coplanar configuration).

[0165] In a coplanar configuration, the radiation head is positioned to rotate about an axis of rotation and in a first plane. In a non-coplanar configuration, the radiation head is tilted relative to the first plane so that the radiation field generated by the radiation head is directed at an oblique angle relative to the first plane and the axis of rotation. In a non-coplanar configuration, the radiation head is positioned to rotate in a respective second plane that is parallel to the first plane and offset from the first plane. Since the radiation beam is emitted at an oblique angle relative to the second plane, the beam sweeps out a cone as the radiation head rotates.

[0166] When the radiotherapy device is in both the coplanar mode and the non-coplanar mode, the beam receiving device 602 remains in the same position relative to the rotatable gantry. Therefore, the beam receiving device 602 is configured to rotate about the rotation axis in the same plane in both the coplanar mode and the non-coplanar mode. This can be the same plane as the plane in which the radiation head rotates.

[0167] The beam shaping device 610 is configured to reduce the spread of the radiation field in a non-coplanar configuration compared to a coplanar configuration.

[0168] Radiation therapy device 600 includes controller 630 programmed to control radiation source 612, beam receiving device 806, and gantry 802. Controller 840 may perform functions or operations such as treatment planning, treatment delivery, image acquisition, image processing, motion tracking, motion management, and / or other tasks related to the radiation therapy process.

[0169] The controller 630 is programmed to control features of the device 600 according to a radiotherapy treatment plan for irradiating a targeted area (also referred to as a targeted tissue) of a patient. The treatment plan includes information about a specific dose to be applied to the targeted tissue, as well as other parameters such as beam angles, dose-volume histogram information, the number of radiation beams to be used during therapy, the dose of each beam, etc. The controller 630 is programmed to control various components of the device 600, such as the gantry 604, the radiation head 610, the beam receiving device 602, and the support surface 620, according to the treatment plan. Using a method according to an embodiment, a treatment plan can be determined.

[0170] The hardware components of the controller 630 may include one or more computers (e.g., general-purpose computers, workstations, servers, terminals, portable / mobile devices, etc.); processors (e.g., central processing units (CPUs), graphics processing units (GPUs), microprocessors, digital signal processors (DSPs), field programmable gate arrays (FPGAs), dedicated or specially designed processors, etc.); memory / storage devices, such as memory (e.g., read-only memory (ROM), random access memory (RAM), flash memory, hard disk drives, optical disks, solid-state drives (SSDs), etc.); input devices (e.g., keyboards, mice, touch screens, models, buttons, knobs, trackballs, levers, handles, joysticks, etc.); output devices (e.g., displays, printers, speakers, vibration devices, etc.); circuits; printed circuit boards (PCBs); or other suitable hardware. The software components of the controller 630 may include operating device software, application software, etc.

[0171] The radiation head 610 can be connected to a head actuator 614, which is configured to actuate the radiation head 610, for example, between a coplanar configuration and one or more non-coplanar configurations. This can involve translation and rotation of the radiation head 610 relative to the frame. In some embodiments, the head actuator can include a curved guide rail along which the radiation head 610 can be moved to adjust the position and angle of the radiation head 610. The controller 630 can control the configuration of the radiation head 630 via the head actuator 614.

[0172] The beam shaping device 618 includes a shaping actuator 616. The shaping actuator is configured to control the position of one or more elements in the beam shaping device 618 to shape the radiation beam 606. In some embodiments, the beam shaping device 618 includes an MLC, and the shaping actuator 616 includes a device for actuating the blades of the MLC. The beam shaping device 618 may also include an aperture, and the shaping actuator 616 may include a device for actuating a stop of the aperture. The controller 630 may control the beam shaping device 618 via the shaping actuator 616.

[0173] The treatment plan may include positioning information of the beam shaping device 618. The positioning information of the beam shaping device 618 may include information indicating the configuration of one or more elements of the beam shaping device 618, such as the configuration of blades of the MLC of the beam shaping device 618, the configuration of an aperture of the beam shaping device 618, and / or the configuration of an opening (e.g., a window or a through hole) of the MLC, etc.

[0174] As described herein, the objective of an optimizer (or an optimization process) is to find a set of optimization parameters that minimizes a cost function. However, it should be understood that, alternatively and equivalently, the objective of an optimizer (or an optimization process) can be considered to be to find a set of optimization parameters that maximizes a reward function.

[0175] Unless otherwise specifically stated, as will be apparent from the following discussion, it should be understood that throughout the specification, discussions utilizing terms such as "receiving," "determining," "comparing," "enabling," "maintaining," "identifying," "obtaining," "accessing," etc., refer to the actions and processes of a computer system or similar electronic computing device that manipulates and transforms data represented as physical (electronic) quantities within the computer system's registers and memories into other data similarly represented as physical quantities within the computer system's memories or registers or other such information storage, transmission, or display devices.

[0176] Although certain embodiments have been described, these embodiments are presented only by way of example and are not intended to limit the scope of the present disclosure. In fact, the novel methods and devices described herein can be embodied in various other forms; in addition, various omissions, substitutions, and changes can be made in the form of the methods and devices described herein.

[0177] The following statements may be helpful in understanding the present disclosure:

[0178] Statement 1: A computer-implemented method for radiation therapy treatment planning, the method comprising: receiving a reference target, the reference target representing a goal to be achieved by a radiation therapy system; selecting a cost function associated with the reference target from a plurality of cost functions, the selected cost function being associated with a specified initial weight value, the specified initial weight value corresponding to a sensitivity of an example dose distribution to changes in the selected cost function; applying an optimization process to a radiation therapy treatment plan for radiation treatment of a patient based on the received reference target, wherein the optimization process uses the specified initial weight value for the selected cost function and seeks to minimize the selected cost function.

[0179] Statement 2: The method according to statement 1, wherein the optimization process includes: determining a dose distribution based on the reference target, the dose distribution indicating the expected dose in the target area of ​​the patient and / or the surrounding area of ​​the patient.

[0180] Statement 3: A method according to Statement 1 or Statement 2, wherein the selected cost function is associated with specified weight processing parameters, and the optimization process includes multiple optimization iterations, wherein: the first optimization iteration uses the specified initial weight value of the selected cost function and seeks to minimize the selected cost function; and subsequent optimization iterations use the specified weight processing parameters of the selected cost function to modify the specified initial weight value.

[0181] Statement 4: A method as recited in any preceding statement, wherein said optimization process iteratively adjusts said radiation therapy plan to minimize a selected cost function.

[0182] Statement 5: A method as recited in any preceding statement, wherein the optimization process seeks to minimize a selected cost function with respect to one or more optimizable parameters.

[0183] Statement 6: The method of Statement 5, wherein the one or more optimizable parameters include one or more of: dose excess, beamlet weights, beam angles, dose-volume histogram information, number of radiation beams, and dose per beam.

[0184] Statement 7: The method according to any of the preceding statements further includes: receiving a second reference target; selecting a cost function associated with the second reference target from the multiple cost functions, the selected second cost function is associated with a second specified weight value, the second specified weight value corresponds to the sensitivity of the example dose distribution to changes relative to the selected second cost function; and applying an optimization process to the radiation therapy plan for radiation treatment of the patient based on the second reference target.

[0185] Statement 8: A method as recited in any preceding statement, wherein said plurality of cost functions comprises a subset of cost functions, said subset comprising at least two cost functions that share a specified weight value and / or a specified weight processing parameter.

[0186] Statement 9: A method as described in any preceding statement, wherein the reference target comprises a reference dose value for a targeted area and / or a surrounding area of ​​the patient.

[0187] Statement 10: A method as recited in any preceding statement, wherein said reference target comprises reference volume values ​​of a targeted region and / or a surrounding region of said patient.

[0188] Statement 11: A method as in any preceding statement, wherein said plurality of cost functions comprises any or all of the following:

[0189] Equivalent uniform dose cost function, which is proportional to:∑ i∈V |V i |(D i / Δ) p ;

[0190] Partial volume cost function, which is proportional to: as well as

[0191] Dose volume histogram cost function, which is proportional to:

[0192] in and wherein the cost function represents the sum of i voxels in the patient's volume V, and wherein V i is the volume of the ith voxel, D i is the dose delivered to the i-th voxel, Δ is a reference dose value, which is the maximum radiation dose in the target area of ​​the patient and / or the surrounding area of ​​the patient, p is a power law exponent, and d0 is a constant or a fraction of Δ.

[0193] Statement 12: A method as recited in any preceding statement, further comprising: accepting user input to a graphical user interface, said user input comprising said reference target and / or said selected cost function.

[0194] Statement 13: The method according to any of the preceding statements further includes: determining parameter values ​​corresponding to the radiotherapy plan, wherein the parameter values ​​include one or more of the following: dose excess value, sub-beam weight, beam angle, dose-volume histogram information, number of radiation beams and dose per beam; and outputting the parameter values.

[0195] Statement 14: A data processing apparatus comprising: a memory storing computer executable instructions; and a processor configured to execute the instructions to perform the method according to any one of Statements 1 to 13.

[0196] Statement 15: A computer program comprising instructions which, when executed by a computer, cause the computer to carry out a method according to any one of Statements 1 to 13.

[0197] Statement 16: A non-transitory computer-readable storage medium comprising instructions which, when executed by a computer, cause the computer to perform the method according to any one of Statements 1 to 13.

Claims

1. A computer-implemented method for radiation therapy planning, the method comprising: receiving a reference target, the reference target representing a goal to be achieved by the radiotherapy system; selecting a cost function associated with the reference target from a plurality of cost functions, the selected cost function being associated with a specified initial weight value corresponding to a sensitivity of the example dose distribution relative to changes in the selected cost function; Based on the received reference target, an optimization procedure is applied to a radiation therapy plan for radiation therapy of a patient, wherein the optimization procedure uses the specified initial weight values ​​of the selected cost function and seeks to minimize the selected cost function.

2. The method according to claim 1, wherein: The optimization process includes determining a dose distribution based on the reference target, the dose distribution indicating an expected dose in a targeted region of the patient and / or a surrounding region of the patient.

3. The method according to claim 1, wherein: The selected cost function is associated with specified weight processing parameters, and the optimization process includes a plurality of optimization iterations, wherein: A first optimization iteration uses the specified initial weight values ​​for the selected cost function and seeks to minimize the selected cost function; and Subsequent optimization iterations use the specified weight processing parameters of the selected cost function to modify the specified initial weight values.

4. The method according to claim 1, wherein: The optimization process iteratively adjusts the radiation therapy plan to minimize a selected cost function.

5. The method according to claim 1, wherein: The optimization process seeks to minimize a selected cost function with respect to one or more optimizable parameters.

6. The method according to claim 5, wherein: The one or more optimizable parameters include one or more of the following: number of beams, beam angle, dose per beam, beamlet weights, segment or control point shape, segment or control point weights, dose-volume histogram information, dose excess value.

7. The method according to claim 1, further comprising: receiving a second reference target; selecting a cost function associated with the second reference target from the plurality of cost functions, the selected second cost function being associated with a second specified weight value corresponding to a sensitivity of the example dose distribution relative to changes in the selected second cost function; as well as Based on the second reference target, an optimization procedure is applied to the radiation therapy plan for radiation therapy of the patient.

8. The method according to claim 1, wherein: The plurality of cost functions includes a subset of cost functions, the subset including at least two cost functions that share a specified weight value and / or a specified weight processing parameter.

9. The method according to claim 1, wherein: The reference target includes reference dose values ​​of a targeted area and / or surrounding area of ​​the patient.

10. The method according to claim 1, wherein: The reference target includes reference volume values ​​of a target region and / or a surrounding region of the patient.

11. The method according to claim 1, wherein: The plurality of cost functions may include any or all of the following: Equivalent uniform dose cost function, which is proportional to: ∑ i∈V |V i |(D i / Δ) p ; Partial volume cost function, which is proportional to: as well as Dose volume histogram cost function, which is proportional to: in and wherein the cost function represents the sum of i voxels in the patient's volume V, and wherein V i is the volume of the ith voxel, D i is the dose delivered to the i-th voxel, Δ is a reference dose value, and wherein the reference dose value corresponds to or includes the maximum radiation dose in the target area of ​​the patient and / or the surrounding area of ​​the patient, p is a power law exponent, and d0 is a constant or a fraction of Δ.

12. The method according to claim 1, further comprising: User input is accepted to a graphical user interface, the user input including the reference target and / or the selected cost function.

13. The method according to claim 1, further comprising: determining parameter values ​​corresponding to the radiation therapy plan, wherein the parameter values ​​include one or more of the following: number of beams, beam angles, dose per beam, beamlet weights, segment or control point shapes, segment or control point weights, dose-volume histogram information, dose excess values; and Output the parameter value.

14. A data processing device, comprising: a memory storing computer executable instructions; as well as A processor configured to execute the instructions to: receiving a reference target, the reference target representing a goal to be achieved by the radiotherapy system; selecting a cost function associated with the reference target from a plurality of cost functions, the selected cost function being associated with a specified initial weight value corresponding to a sensitivity of the example dose distribution to changes in the selected cost function value; as well as Based on the received reference target, an optimization procedure is applied to a radiation therapy plan for radiation therapy of a patient, wherein the optimization procedure uses the specified initial weight values ​​of the selected cost function and seeks to minimize the selected cost function.

15. A computer program comprising instructions which, when executed by a computer, cause the computer to: receiving a reference target, the reference target representing a goal to be achieved by the radiotherapy system; selecting a cost function associated with the reference target from a plurality of cost functions, the selected cost function being associated with a specified initial weight value corresponding to a sensitivity of the example dose distribution to changes in the selected cost function value; Based on the received reference target, an optimization procedure is applied to a radiation therapy plan for radiation therapy of the patient, wherein: The optimization process uses the specified initial weight values ​​for a selected cost function and seeks to minimize the selected cost function.

16. A non-transitory computer-readable storage medium comprising instructions which, when executed by a computer, cause the computer to: receiving a reference target, the reference target representing a goal to be achieved by the radiotherapy system; selecting a cost function associated with the reference target from a plurality of cost functions, the selected cost function being associated with a specified initial weight value corresponding to a sensitivity of the example dose distribution to changes in the selected cost function value; Based on the received reference target, an optimization procedure is applied to a radiation therapy plan for radiation therapy of the patient, wherein: The optimization process uses the specified initial weight values ​​for a selected cost function and seeks to minimize the selected cost function.