Methods and systems for robust radiotherapy treatment planning to address dose mapping uncertainty

By generating a robust radiotherapy treatment plan, and utilizing multi-image registration and error estimation, the dose mapping uncertainty caused by image registration in radiotherapy is solved, thereby improving the accuracy and robustness of dose delivery.

CN116785601BActive Publication Date: 2026-11-10RAYSEARCH LAB
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202310286794.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2022-03-22
Filing Date
2023-03-21
Publication Date
2026-11-10
Estimated Expiration
2043-03-21

AI Technical Summary

Technical Problem

Existing technologies have failed to effectively address dose mapping uncertainties caused by image registration in radiotherapy, resulting in inaccurate dose delivery accuracy and distribution. Furthermore, existing correction methods are time-consuming and dependent on user expertise.

Method used

By generating a robust radiotherapy treatment plan method, multiple image registration algorithms and error estimations are used to construct a distribution of mapped doses, and an optimization function is used to account for uncertainties, thus generating a robust dose distribution.

Benefits of technology

This technology ensures adequate total dose delivery in regions with steep dose gradients and image registration uncertainty, improving the reliability of dose mapping and the robustness of treatment planning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116785601B_ABST
    Figure CN116785601B_ABST
Patent Text Reader

Abstract

This invention relates to a method and system for robust radiotherapy treatment planning for dose mapping uncertainty. A method for generating a robust radiotherapy treatment plan for a treatment volume of a subject, the treatment volume being defined using a plurality of voxels, the method comprising the steps of: receiving a first image of the treatment volume; receiving at least one second image of the treatment volume; generating a distribution of mapped doses in the first image by mapping doses defined in the at least one second image to the first image using image registration; defining an optimization problem using at least one optimization function for the total dose associated with the radiotherapy treatment; calculating at least one optimization function value based on the doses of at least two mapped doses in the distribution of mapped doses in the first image; and generating a radiotherapy treatment plan by optimizing the at least one optimization function value evaluated by taking into account the doses of at least two mapped doses in the distribution of mapped doses.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This disclosure generally relates to the field of radiotherapy, and in particular to the generation, optimization, and evaluation of radiotherapy treatment plans. Background Technology

[0002] Medical imaging is widely used in radiotherapy treatment planning to identify and delineate targets (tumors) and organs at risk (OARs) within the patient's body. Along with other radiotherapy planning parameters, such as the minimum or target dose delivered to the tumor and the maximum dose for the OAR, the delineated targets and OARs from acquired images are used as input to optimize the dose distribution of the plan, concentrating radiation on the target while limiting exposure to the OAR.

[0003] In many treatment scenarios, multiple images of the patient are acquired at different times and / or spatial points. Examples include, but are not limited to, adaptive planning, where the original plan is adapted during radiotherapy, retreatment, and 4D robust optimization, where the patient undergoes additional radiotherapy during retreatment, and where multiple images are acquired, for example, within a single respiratory cycle during 4D robust optimization.

[0004] To compare medical images of the same patient acquired at different times and / or from different angles and to obtain information about the delivered dose, the images undergo image registration (IR) to transform the different datasets into a single coordinate system, such as a first image. However, during radiotherapy, changes can occur that affect the accuracy of dose delivery and the spatial distribution of the effective dose. Such changes can include tumor displacement, shrinkage, or expansion, or changes in the shape of the tumor and surrounding anatomy within the patient, for example, due to weight loss / gain.

[0005] To mitigate the effects of anatomical variations, deformable image registration (DIR) is used, where individual volumetric elements are mapped from one image to another using non-rigid translation and / or rotation, resulting in a deformable map or deformable vector field (DVF). Nevertheless, current methods of DIR are ambiguous because there is no perfectly good deformable model that deforms all anatomical regions, and therefore any dose mapping based on DIR is inherently uncertain.

[0006] Existing techniques do not accurately account for this uncertainty. Instead, the choice is between trusting the deformation vector field (DVF) in a region or rejecting it entirely, thus applying potentially overly conservative estimates as a dose of the mapping.

[0007] WO 2012 / 069965A1 discloses a system and method for manually or automatically correcting deformable mappings resulting from the registration of two planned image sets. However, the correction process is time-consuming, requires user expertise, and cannot adequately overcome uncertainties in dose mapping. Summary of the Invention

[0008] The objective of this disclosure is to provide an improved solution, wherein a robust optimization method is provided to address dose mapping uncertainties in image registration. This objective is achieved in a first aspect of this disclosure, wherein a method is provided for generating a robust radiotherapy treatment plan for a treatment volume of an object, the treatment volume being defined using a plurality of voxels, the method comprising the steps of:

[0009] - Receive the first image of the treatment volume;

[0010] - Receive at least one second image of the treatment volume;

[0011] - By using image registration, the dose defined in at least one second image is mapped to the first image to generate the distribution of the mapped dose in the first image;

[0012] - Define the optimization problem using at least one optimization function for the total dose associated with radiotherapy treatment, where the total dose is a function of the dose defined in the first image and the dose defined in at least one second image;

[0013] - Calculate at least one optimization function value based on the doses of at least two mappings in the dose distribution of the mappings in the first image; and

[0014] - A radiotherapy treatment plan is generated by optimizing at least one optimization function value that is evaluated by taking into account the doses of at least two mapped doses in the distribution of mapped doses.

[0015] By considering the dose-varying frequency (DVF) as an uncertainty within reasonable constraints, or by considering multiple credible DVFs, this disclosure enables the acquisition of a robust total dose to the modeled dose-mapping uncertainty. A key advantage is that by considering the distribution of the mapped dose rather than just the nominal mapped dose, the optimizer can be guided to a solution that ensures a sufficient total dose, even in regions with steep dose gradients and IR uncertainties.

[0016] This disclosure provides robustness regarding uncertainties in doses already delivered, whereas conventional methods only consider doses to be delivered in the future. Furthermore, the generation of the distribution of mapped doses in the first image can be accomplished in several ways and is not limited to any particular dependency structure between voxels in the treatment volume. In this context, the distribution can be defined as a set of mapped doses, and optionally as the corresponding probability of each member in the set. Alternatively, the distribution can be defined as nominal values ​​and error estimates. For example, these values ​​could be the mean and standard deviation of each voxel, or alternatively, the maximum deviation of each voxel.

[0017] In one embodiment, deformable image registration (DIR) is used to perform dose mapping from at least one second image to a first image. DIR can interpret non-rigid translations, rotations, and / or other variations in anatomical structures.

[0018] In one embodiment, at least two distinct image registrations are used to perform the mapping of doses in at least one second image to a first image, wherein the distribution of the mapped doses is a set containing doses from at least two mappings. Increasing the number of image registrations, for example, by using several different algorithms or by selecting algorithm parameters, provides a more reliable estimate of dose mapping uncertainty.

[0019] In one embodiment, mapping a dose from at least one second image to a first image is performed by performing error estimation for each vector in the deformed vector field generated by image registration, wherein the distribution of the mapped dose is based on at least two mapped doses generated from deformed vector fields perturbed according to the error estimation using at least two different perturbations. One of the perturbations can be zero perturbation.

[0020] In one embodiment, the dose in at least one second image is a dose from a previous treatment, treatment fraction, or partial treatment fraction, and wherein the step of generating a radiotherapy treatment plan includes generating a radiotherapy retreatment plan or adapting a pre-existing radiotherapy treatment plan.

[0021] In one embodiment, the dose in at least one second image is a portion of the beam dose in a 4D radiotherapy treatment plan.

[0022] In one embodiment, the optimization problem includes constraints defining parameters that are maintained during optimization. Constraints can be, for example, in the form of a predetermined dose within a defined subvolume (e.g., the target), which is not changed during optimization. In this way, the target dose is maintained to ensure a certain dose distribution, but the remainder of the radiotherapy treatment plan is robustly optimized.

[0023] In one embodiment, the optimization problem includes biological or physical objectives. Preferably, physical objectives include dose constraints on the target and organ at risk (OAR) within the treatment volume, dose-volume histogram (DVH) constraints, LET constraints, particle stop location and / or homogeneity and conformity index. In this way, biological uncertainties can also be combined with other (physical) objectives. The planning optimization and evaluation should be able to combine different biological models as well as physical optimization functions and objectives. The optimization problem can be a combination of physical objectives, such as minimum and maximum doses to the target and organ at risk respectively, and DVH constraints, as well as biological objectives such as BED, EQD2, EUD, TCP, and NTCP.

[0024] In one embodiment, physical objectives include dose limitation on targets and organs at risk (OARs) within the treatment volume, dose-volume histogram (DVH) limitation, linear energy transfer (LET) limitation, particle stopping location and / or uniformity and conformity index.

[0025] In one embodiment, the optimization includes: a stochastic programming method, wherein the expected value of the optimization function is minimized at at least two mapped doses in the distribution of mapped doses; a minimax method, wherein the maximum value of the optimization function is minimized at at least two mapped doses in the distribution of mapped doses; or any combination of the two, commonly referred to as minimax stochastic programming; or a voxel-by-voxel worst-case method, in which the worst-case dose is optimized for each voxel individually considered.

[0026] In one embodiment, at least two mapped doses in the distribution of mapped doses are further combined with an additional set of error scenarios, wherein the additional set of error scenarios represents a specific realization of uncertainty in one or more parameters related to the treatment plan, including particle range, spatial location of treatment volume, radiotherapy device setup, density of irradiated tissue, interaction effects, organ movement, and / or biological model parameter values.

[0027] According to a second aspect, a computer program product is provided, the computer program product including a computer-readable code means that, when run on a computer, causes the computer to perform the method according to the first aspect.

[0028] According to a third aspect, a computer system is provided, the computer system including a processor coupled to a memory, the memory storing computer-readable instructions that, when executed by the processor, cause the processor to perform the method according to the first aspect.

[0029] According to the fourth aspect, a treatment planning system is provided, which includes the aforementioned computer system. Attached Figure Description

[0030] These and other features, aspects and advantages of this disclosure will be further explained in the following description with reference to the accompanying drawings, in which:

[0031] Figure 1 Three mappings from the delivered dose on the old image to the new image are shown;

[0032] Figure 2 A flowchart illustrating the steps of a computer-based method for generating a robust radiotherapy treatment plan according to an embodiment of the present disclosure is shown;

[0033] Figure 3aand Figure 3b The illustration shows the generation of a dose distribution of a mapping according to an embodiment of the present disclosure;

[0034] Figure 4a and Figure 4b The illustration shows the generation of a dose distribution of a mapping according to another embodiment of the present disclosure;

[0035] Figure 5 A computer-based system for evaluating, visualizing, generating, and improving radiotherapy treatment plans is illustrated schematically according to one embodiment of the present disclosure.

[0036] In this document, the same reference numerals are used to denote the same elements wherever possible, which are common to multiple figures. Furthermore, the images in the figures are simplified for illustrative purposes and are not necessarily depicted to scale. Detailed Implementation

[0037] Now for reference Figure 1 The diagram illustrates a fundamental problem related to dose mapping uncertainty. Figure 1 An image of the patient is shown on the left side, indicating the delivered dose. Different shades of shading indicate the amount of dose delivered to a specific area of ​​the treatment volume. Figure 1 The right-hand side shows three different images of the same treatment volume at different time points. To compare or integrate data from these different images, for example, to map the delivered dose from the old image to the new image, the old image is registered with the new image using image registration. Each of the three images uses a different algorithm to represent a different image registration, resulting in three different mapped doses on the new image. Therefore, there is an inherent uncertainty in dose mapping that affects radiotherapy treatment planning.

[0038] Figure 2 This is a flowchart illustrating an embodiment of the method according to the present disclosure, which can be used with the generation of radiotherapy treatment plans. In one embodiment, the starting point is an initial treatment plan and a number of scenarios to be considered, and the method aims to obtain an improved treatment plan based on the initial treatment plan to modify the initial plan with some constraints, or to obtain a deliverable treatment plan where the initial treatment plan does not meet all machine limitations. Depending on the type of data included in the plan, additional input data, such as patient-related data, may be required for dose calculation. The initial treatment plan can be obtained by any means known in the art, including scenario-based and non-scenario-based methods.

[0039] The purpose of generating a treatment plan is to provide radiotherapy for a treatment volume on a subject (patient), which may be an organ and may include a target that is a tumor or a cluster of tumor cells. The treatment volume is defined using multiple voxels, as known in the art.

[0040] In step S100, a first image of the treatment volume is received. The first image may be a reference image representing the current state of the patient to be used in the radiotherapy treatment plan. In step S102, at least one second image of the treatment volume is received. The second image may be a target image representing the patient's state at a previous time point (e.g., during a previous treatment or fractionated treatment). Alternatively, the second image may be a stage image representing the patient's state at different stages (e.g., at different stages of the patient's respiratory cycle).

[0041] Both the first and second images can be obtained using any suitable medical imaging technique, including but not limited to X-ray computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET), ultrasound, and single-photon emission computed tomography (SPECT).

[0042] In step S104, image registration is used to map the dose defined in at least one second image to the first image, thus generating a distribution of the mapped dose in the first image. In one embodiment, deformable image registration is used to account for non-rigid translations, rotations, and / or other changes in anatomical structures. Image registration produces a deformable vector field, where vectors representing the deformation between the first and second images are assigned to each voxel in the treatment volume.

[0043] In step S106, an optimization problem is defined using at least one optimization function for the total dose associated with radiotherapy treatment. The total dose is a function of the dose defined in the first image and the dose defined in at least one second image.

[0044] In step S108, an optimization function value is calculated based on the doses of at least two mapped doses in the distribution of mapped doses in the first image. The optimization function value calculation can be performed using the mapped doses and the doses defined in the first image as input variables, thereby producing an optimization function value. Alternatively, each of the mapped doses can be used as an input variable, where the first function value is based on the dose of the first mapped dose and the second function value is based on the dose of the second mapped dose.

[0045] In step S110, the optimization function value, which is evaluated by considering the doses of at least two mapped doses in the distribution of mapped doses, is used to generate a radiotherapy treatment plan.

[0046] The distribution of the mapped dose can be generated in several ways and is not limited to any specific dependency structure between voxels. Now refer to Figure 3a and Figure 3bThe illustration shows an embodiment in which at least two different image registrations are used to map a dose in at least one second image to a first image. Figure 3a A first image is shown, which includes point A, and indicates that... Figure 3b The four points A', A”, A”', and A”” in the second image shown are mapped to four arrows representing point A in the first image. Each of the four arrows is then a vector in the corresponding deformation vector field generated by the registration of the four different images.

[0047] The set of mapped doses can be considered as the distribution of mapped doses. By increasing the number of image registrations, i.e., using several different image registration algorithms, or using the same algorithm with different settings or added perturbations, a more reliable estimate of dose mapping uncertainty is achieved.

[0048] Now for reference Figure 4a and Figure 4b Another embodiment of how to generate the distribution of the mapped dose is shown, wherein each vector in the deformable vector field is considered to be erroneous. Figure 4a A first image is shown, comprising two points A and B, and a vector assigned to each point, representing the... Figure 4b Points A' and B' in the second image shown are mapped to points A and B in the first image. Figure 4a The estimation of the error of the two vectors is given here by having radii ε respectively. A and ε B The circles surrounding points A' and B' are then used to represent the vectors. Multiple reliable implementations are then generated for each vector using the error estimate to produce a distribution of the mapped dose. More specifically, the distribution of the mapped dose can be based on at least two mapped doses generated from a deformed vector field perturbed according to the error estimate using at least two different perturbations. One of the perturbations can be a zero perturbation, i.e., where the deformed vector field obtained from image registration is considered without perturbing the vector.

[0049] The method according to this disclosure can be used for various planning scenarios. In one embodiment, the method is used to generate a radiotherapy retreatment plan or to adapt a pre-existing radiotherapy treatment plan. In such a case, the dose defined in at least one second image is a dose from previous treatment, fractionated treatment, or partially fractionated treatment, also referred to herein as the background dose. The total dose is then the sum of the background dose and the new dose to be delivered in the retreatment plan or the adapted plan.

[0050] In one embodiment, the method is used to generate a 4D radiotherapy treatment plan, wherein the treatment volume undergoes periodic spatial deformation during different phases (e.g., the patient's respiratory cycle). In such a case, the dose defined in at least one second image is a portion of the beam dose delivered during a specific phase. The total dose is then the sum of the portion of the beam dose to be delivered in the 4D radiotherapy treatment plan.

[0051] The optimization function can be included in the optimization as a constraint. Alternatively, the optimization function can be included in the optimization as an objective function. Typically, objectives are set for the treatment, and these objectives are used to define the objective function, constraints, or a combination thereof. The objective function is the desired objective that the optimization should strive to achieve or that the optimization should satisfy as much as possible, while constraints are strict objectives or conditions that must be precisely satisfied, such as the minimum dose to the tumor or the maximum dose to the OAR, or the boundaries of variables controlling the objective function.

[0052] Various types of optimization methods for achieving robustness can be combined with the methods according to this disclosure. For example, minimization-maximization (or "composite worst-case") optimization can be used, where the worst-case scenario on the composite objective function is optimized. The optimization problem is then formulated as follows:

[0053]

[0054] Where X is the set of feasible optimization variables (e.g., the set of allowed point weights, MLC leaf positions, etc.), and S is the set of scenarios enumerating doses for different mappings.

[0055] f(x;s)

[0056] It is a composite objective that is a function of the optimization variable x under scenario s. For example, f(x; s) can be expressed as g(d(x) + d... b (s) is given, where d(x) is the new dose generated by the optimization variable x, d b (s) is the background dose (i.e., the mapped dose) in scene s, and g is the dose with respect to the total dose d(x) + d b (s) related functions.

[0057] Another type of optimization method that achieves robustness is expectation optimization, where the expected value with respect to uncertainty is optimized. The optimization problem is formulated as follows:

[0058]

[0059] Here, E is the expectation operator, and Y is a random variable that represents the value from the scene set S.

[0060] The third alternative is a voxel-by-voxel worst-case optimization method. In this method, the two artificial worst-case dose distributions d are calculated based on the mapped dose. high and d low Here, d high The dose is calculated as the highest mapping of the scene to each individually considered voxel, and d low The dose calculated as the lowest mapping for each individually considered voxel in the scene pair, i.e.

[0061]

[0062]

[0063] Where, d b,i (s) represents the dose mapped to voxel i under scenario s, and N is the number of voxels.

[0064] Then, the optimization problem is formulated as follows:

[0065]

[0066] Among them, f high It is a composite objective function, the composition of which is used to avoid overdose (e.g., the target of OAR in organs at risk), and f low It is a composite objective function whose composition is used to avoid underdosing (e.g., the minimum dose requirement of the target).

[0067] Another alternative is to minimize the objective function h(x) which is not necessarily (but may be) related to the entire set S of scenarios, and to the constraints of the function f(x; s) over all s in S, i.e.

[0068]

[0069] It is subject to the condition f(x; s)≤0, s∈S.

[0070] The objective function h(x) can be formulated using any of the methods described above, but it can also be formulated to consider only the nominal scenario corresponding to no error.

[0071] Other methods, such as stochastic minimax methods, which are a combination of worst-case optimization and expectation optimization, can also be used and are known in the art.

[0072] Now for reference Figure 5 , Figure 5A simplified schematic representation of a computer-based system 100 for generating a radiotherapy treatment plan 114 according to the present disclosure is shown. The computer-based system 100 includes a memory or database 110 having the radiotherapy treatment plan 114 stored thereon, and a computer program 116 for generating an improved radiotherapy treatment plan 118. The memory 110 can be any volatile or non-volatile memory device, such as a flash drive, hard disk drive, optical disk drive, dynamic random access memory (DRAM), static random access memory (SRAM), and any other suitable device for storing information and for subsequent information retrieval and data processing. Furthermore, the system 100 includes one or more hardware processors 120 for performing data processing, which are capable of accessing the memory 110. The hardware processors 120 can be made of one or more of a central processing unit (CPU), digital signal processor (DSP), reduced instruction set computer (RISC), application-specific integrated circuit (ASIC), complex programmable logic device (CPLD), field-programmable gate array (FPGA), parallel processor system, or combinations of these different hardware processor types.

[0073] Computer program 116 is composed of computer-readable instructions that can be passed to and executed by hardware processor 120. When executed on hardware processor 120, the computer-readable instructions perform methods for generating an improved radiotherapy treatment plan 118. The results of the processing performed by hardware processor 120 while executing computer program 116 (e.g., the improved radiotherapy treatment plan 118, and associated data) can be stored in memory 110. Hardware processor 120 can also access memory 110 via direct memory access (DMA) and can also use cache memory to store temporary processing results. Computer program 116 can also be stored on non-transitory computer-readable medium 130, such as a Universal Serial Bus (USB) flash drive, optical data carriers such as CD-ROM, DVD-ROM and Blu-ray discs, floppy disks, exchangeable hardware drives, USB external hard disk drives (HDDs), or any other portable information storage device, so that computer program 116 can be transferred to different computing systems and can also be loaded into memory 110 of system 100. This can be accomplished by connecting the computer-readable medium 130 to the system 100 via a data reader / writer 140 (e.g., optical drive, USB interface, etc.).

[0074] In addition, system 100 includes a display unit 150 having a display driver that allows visualization of the results of data processing, such as a three-dimensional (3D) representation of a target volume of a patient containing, for example, tumors or cancer cells, as well as healthy, at-risk organs to which dose delivery must be prevented, 3D contour data, or two-dimensional (2D) slice representations of various cross directions and LET distributions for both the target volume and the at-risk organs, biological effects (e.g., the possibility of damage / cell death / side effects), etc. For example, it can display a 3D computer-generated reconstruction of a CT scan. Furthermore, display unit 150 can display a dose-volume histogram (DVH) summarizing the 3D dose distribution using a graphical 2D format. For example, display unit 150 is configured to display comparative DVH plots for the patient volume showing the dose contribution of radiotherapy treatment plan 114 and for the same volume for an optimized or improved radiotherapy treatment plan 118, so that LET distributions can also be visually compared.

[0075] Display unit 150 is used to display 3D scans of the patient taken before, during, or after treatment. For example, it can display a 3D computer-generated reconstruction of a CT scan. Furthermore, display unit 150 can display LET, dose, and / or DVH summarizing the 3D dose distribution using a graphical 2D format or a numerical format. For example, display unit 150 is configured to display a comparative LET plot for a patient volume, showing cancer cell destruction or dose contribution of the radiotherapy treatment plan 114. This shows and compares optimized or improved radiotherapy treatment plans for the same volume, enabling visual comparison of improvements. Additionally, display unit 150 may be equipped with a touchscreen and capable of displaying a graphical user interface for operating system 100.

[0076] Additionally, the computer system 100 has a system bus 160 that connects the hardware processor 120, memory 110, data reader 140, touchscreen, and various other data input / output interfaces and peripheral devices (not shown). For example, the computer system 100 can be connected to a keyboard 170 for user input and can be connected to an external radiotherapy planning device 180, such as a powerful dedicated computer, which has already created radiotherapy treatment plans. Furthermore, the system 100 can be connected to a CT scanner (not shown). For example, the external device 180 that creates the radiotherapy treatment plan 114 can be capable of developing a dose and LET distribution calculation algorithm encoded in software, and can access radiation data regarding the prescribed dose distribution, machine calibration data, and patient-specific information regarding the patient's target volume and organs at risk. The external device 180 can then deliver the radiotherapy treatment plan 114 to the computer system 100 for evaluation, visualization, creation of new plans, and improvement of existing plans taking into account the LET distribution. However, it is also possible that the computer program 116 runs on the external device itself, thereby generating not only the radiotherapy treatment plan 114 but also an improved radiotherapy treatment plan 118.

[0077] Furthermore, a computer program product for performing parameter optimization is introduced. The computer program product 130 includes a computer-readable code means that, when run in a computer, executes the method described above.

[0078] Example 1: Retreatment plan with robustness to background dose uncertainty

[0079] Some patients who have already undergone radiotherapy may experience tumor recurrence near the volume previously treated. In these cases, a second course of radiotherapy, known as retreatment or re-radiation, can be considered. First images in the form of planning images are then acquired regarding the planned retreatment. Second images in the form of old planning images of the previous treatment are also considered. To avoid exceeding a certain total fractionated corrected dose for healthy tissue, the retreatment plan must take into account the already delivered dose present in the second image. The delivered dose can be mapped from the second image to the first image after image registration to serve as the background dose. However, image registration and the background dose can also be indeterminate or ambiguous. Different image registrations may result in different background doses.

[0080] Question: Which mapped dose can be trusted as the background dose for a total dose-based retreatment plan?

[0081] Recommended solution: Generate at least two background dose distributions and robustly optimize the total dose to the OAR to account for dose mapping uncertainty.

[0082] - Define a new treatment plan based on the first image received.

[0083] - Define at least one objective for the total dose on at least one OAR, which will be robustly optimized for uncertainties in the dose mapping.

[0084] Examples of these objectives include:

[0085] maximum total physical dose from o to OAR

[0086] o to the maximum total biological equivalent dose (BED) of OAR, such as EQD2

[0087] - Define at least one target for the retreatment dose to the target.

[0088] - Optionally, define additional error scenarios, which represent a specific implementation of one or more additional relevant parameters used in the radiotherapy treatment plan, such as setting errors or range errors.

[0089] - Receive a second image from the original treatment plan and perform image registration between the first and second images to map the delivery dose in the second image to the first image.

[0090] - Generate the dose distribution of the mapping and define the scene based on the dose distribution of the mapping, for example, in one of the following two ways:

[0091] o Perform at least one additional image registration using an alternative algorithm or with parameters of the alternative algorithm. Define the scene for the dose of each resulting mapping.

[0092] o calculates an error estimate for image registration and uses the error estimate to characterize the distribution of the mapped dose. The scenario is defined based on an artificial worst-case dose distribution, for example, the stochastic minimax method described above.

[0093] - Optimize the retreatment plan using a preferred robust optimization framework.

[0094] Example 2: Adaptive treatment planning with robustness to background dose uncertainty

[0095] During radiotherapy, it may be desirable to adjust the original treatment plan to account for changes in the patient's anatomy or previous errors in treatment delivery. In this case, a first image in the form of a replanning image is acquired. At least one second image in the form of a fractionated image is also considered for delivery. The delivered dose can be mapped from at least one second image to the first image after image registration to be used as a background dose. However, image registration and background dose may also be indeterminate or ambiguous. Different image registration techniques may result in different background doses.

[0096] questionWhich mapped dose can be trusted to be used as the background dose for adaptive treatment planning based on total dose?

[0097] Suggested solution Generate at least two background dose distributions and robustly optimize the total dose to the OAR to account for dose mapping uncertainties.

[0098] - Define a new, adjusted treatment plan based on the first image received.

[0099] - Define at least one objective for the total dose, which will be robustly optimized for uncertainties in the dose mapping.

[0100] Examples of these objectives could be:

[0101] maximum total physical dose from o to OAR

[0102] o to OAR's maximum total BED

[0103] Minimum physical dose to target volume

[0104] - Optionally, define additional error scenarios, which represent a specific implementation of one or more additional relevant parameters used in the radiotherapy treatment plan, such as setting errors or range errors.

[0105] - Receive at least one second image from a previously delivered sequence, and perform image registration between the first image and at least one second image to map the delivery dose in the at least one second image to the first image.

[0106] - For each second image, generate the distribution of the dose mapped in the first image, for example, in one of the following two ways:

[0107] o Perform at least one additional image registration using an alternative algorithm or by using parameters of an alternative algorithm.

[0108] o is used to calculate the error estimate for image registration, and this error estimate is used to characterize the distribution of the mapped dose.

[0109] - Define the scene based on the uncertainty in the distribution of doses from each mapping of at least one second image.

[0110] - Optimize the retreatment plan using a preferred robust optimization framework.

[0111] Example 3: 4D treatment planning with robustness to dose mapping uncertainty

[0112] When treating tumors in parts of the body that undergo periodic motion, robustness to that phase of the motion cycle (dose delivery during the motion cycle) can be expected. For this purpose, multiple images from various phases of the motion cycle are sometimes acquired. The dose to be optimized is then a function of the dose in each phase image, mapped by DIR to a common reference image.

[0113] question Which dose mapping can be trusted for use as a phase dose in a 4D-optimized treatment plan?

[0114] Suggested solution Generate doses for at least two mappings from at least one stage image to a reference image and robustly optimize the dose mappings for dose mapping uncertainties.

[0115] - Define 4D treatment plans within the patient volume undergoing exercise.

[0116] - Define at least one objective for the dose in the reference phase image, which will be robustly optimized for uncertainties in the dose mapping.

[0117] Examples of these objectives are:

[0118] maximum total physical dose from o to OAR

[0119] Minimum total physical dose to the target volume

[0120] - Optionally, define additional error scenarios, which represent a specific implementation of one or more additional relevant parameters used in the radiotherapy treatment plan, such as setting errors or range errors.

[0121] - Receive at least one second image as a phase image from different phases in the motion cycle, and perform image registration between the first image and the at least one second image to be used to map the dose in the at least one second image to the first image.

[0122] - For each second image, generate the dose distribution mapped in the first image.

[0123] In this case, the mapped dose will be a function of the optimization variables and will change in each iteration of the optimization. The distribution of the mapped dose can be generated, for example, in one of the following ways:

[0124] o Perform at least one additional image registration using an alternative algorithm or by using parameters of an alternative algorithm.

[0125] o is used to calculate the error estimate for image registration, and this error estimate is used to characterize the distribution of the mapped dose.

[0126] - Define the scene based on the uncertainty in the distribution of doses from each mapping of at least one second image.

[0127] - Optimize the retreatment plan using a preferred robust optimization framework.

[0128] Preferred embodiments of the methods and systems for generating radiotherapy treatment plans have been disclosed above. However, those skilled in the art will recognize that variations are possible within the scope of the appended claims without departing from the inventive concept.

[0129] Provided that the combinations are not contradictory, all alternative embodiments or multiple parts of the embodiments described above can be freely combined or used separately without departing from the inventive concept.

[0130] The following abbreviations were used:

[0131] BED Bioequivalent Dose

[0132] CT computed tomography scan

[0133] CTV Clinical Tumor / Target Volume

[0134] DICOM Medical Digital Imaging and Communications

[0135] DVH dose-volume histogram

[0136] EHR Electronic Health Record System

[0137] EQD equivalent standard fractionated dose

[0138] EQD2 2Gy fractionated equivalent dose

[0139] EUD equivalent uniform distribution

[0140] eMIX Electronic Medical Information Exchange System

[0141] GUI (Graphical User Interface)

[0142] GTV (Total Tumor Volume / Target Volume)

[0143] HIS (Hospital Information System)

[0144] HIM Health Information Management System

[0145] IGRT image-guided radiotherapy

[0146] Intensity-modulated radiotherapy (IMRT)

[0147] LET (Linear Energy Transfer)

[0148] MLC Multileaf Collimator

[0149] MRI magnetic resonance imaging system

[0150] MU monitoring unit

[0151] NTCP normal tissue complication probability

[0152] OAR (Organ in Danger)

[0153] PBS pencil beam scanning

[0154] PET Positron Emission Tomography

[0155] PTV (Planned Tumor / Target Volume)

[0156] QA Quality Assurance

[0157] QC Quality Control

[0158] US ultrasound examination

[0159] RBE relative bioavailability

[0160] ROI (Region of Interest)

[0161] RVS Recording and Verification System

[0162] SPECT (Single Photon Positron Emission Tomography)

[0163] TCP tumor control probability

Claims

1. A method for generating a robust radiotherapy treatment plan for a treatment volume of a subject, the treatment volume being defined using a plurality of voxels, the method comprising the steps of: - Receive (S100) a first image of the treatment volume; - Receive (S102) at least one second image of the treatment volume; - By using image registration to map the dose defined in the at least one second image to the first image, a distribution of the mapped dose in the first image is generated (S104); - Define (S106) an optimization problem using at least one optimization function for the total dose associated with the radiotherapy treatment, wherein the total dose is a function of the dose defined in the first image and the dose defined in the at least one second image; - Calculate (S108) at least one optimization function value based on the doses of at least two mappings in the distribution of the doses mapped in the first image; and - A radiotherapy treatment plan is generated by optimizing the value of at least one optimization function, which is evaluated by taking into account the doses of at least two maps in the distribution of the mapped doses (S110).

2. The method according to claim 1, wherein, Deformable image registration is used to perform the mapping of the dose in the at least one second image to the first image.

3. The method according to claim 1 or 2, wherein, The at least two different image registrations are used to perform mapping of the dose in the at least one second image to the first image, wherein the distribution of the mapped dose is a variation between the at least two mapped doses.

4. The method according to claim 1 or 2, wherein, Mapping the dose in the at least one second image to the first image is performed by performing error estimation for each vector in the deformed vector field generated by the image registration, wherein the distribution of the mapped dose is based on at least two mapped doses generated from deformed vector fields perturbed with at least two different perturbations according to the error estimation.

5. The method according to claim 1 or 2, wherein, The dose in the at least one second image is a dose from a previous treatment, treatment fraction, or partial treatment fraction, and the step of generating a radiotherapy treatment plan includes generating a radiotherapy retreatment plan or adapting a pre-existing radiotherapy treatment plan.

6. The method according to claim 1 or 2, wherein, The dose in at least one second image is a portion of the beam dose in a 4D radiotherapy treatment plan.

7. The method according to claim 1 or 2, wherein, The optimization problem includes constraints on parameters that are defined and maintained during the optimization process.

8. The method according to claim 1 or 2, wherein, The optimization problem includes biological or physical objectives.

9. The method according to claim 8, wherein, The physical objectives include dose limitation on targets and organs at risk (OARs) within the treatment volume, dose-volume histogram (DVH) limitation, linear energy transfer (LET) limitation, particle stopping location and / or uniformity and conformity index.

10. The method according to claim 1 or 2, wherein, The optimization includes: a random programming method, wherein the expected value of the optimization function is minimized at at least two mapped doses in the distribution of mapped doses; a minimax method, wherein the maximum value of the optimization function is minimized at at least two mapped doses in the distribution of mapped doses; or any combination of the two, commonly referred to as minimax random programming; or a voxel-by-voxel worst-case method, in which the worst-case dose is optimized for each voxel individually considered.

11. The method according to claim 1 or 2, wherein, The doses of at least two of the mapped dose distributions are further combined with an additional error scenario, wherein the additional error scenario represents the realization of uncertainty in one or more parameters related to the treatment plan, including particle range, spatial location of the treatment volume, radiotherapy device setup, density of irradiated tissue, interaction effects, organ movement, and / or biological model parameter values.

12. A computer program product (130) comprising computer-readable instructions that, when executed on a computer, cause the computer to perform the method according to any one of claims 1 to 11.

13. A computer system (100) comprising a processor (120) coupled to a memory (110) having computer-readable instructions stored thereon, the computer-readable instructions causing the processor to perform the method according to any one of claims 1 to 11 when executed by the processor.

14. A radiotherapy treatment planning system, the radiotherapy treatment planning system comprising the computer system (100) according to claim 13.

Citation Information

Patent Citations

  • Interactive deformation map corrections

    WO2012069965A1

  • Adaptive replanning based on multimodality imaging

    CN113842566A

  • Method and system for uncertainty based radiotherapy treatment planning

    US20170209712A1