robust treatment planning or plan evaluation
By identifying the structures of uncertainty in radiotherapy treatment planning, defining multiple scenarios and performing robust evaluation and optimization, treatment plans adapted to different material properties and locations are generated. This solves the problem of treatment inaccuracy caused by uncertainty in traditional radiotherapy and achieves more reliable treatment results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- RAYSEARCH LAB
- Filing Date
- 2021-06-14
- Publication Date
- 2026-06-12
Smart Images
Figure CN115702021B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a computer-based method for calculating or evaluating radiotherapy treatment plans, particularly for robust treatment planning. Background Technology
[0002] Traditional radiotherapy treatment planning involves providing one or more medical images of the patient, determining the desired dose distribution, and optimizing the plan to be as close as possible to the desired dose distribution.
[0003] This plan considers various factors, such as the patient's geometry, the material properties (e.g., density) of different areas of the patient, and other factors (e.g., patient movement). It is recognized that the precise location and internal geometry of the patient on the treatment bed are often impossible to ascertain accurately. Methods for robust treatment planning have been developed to account for uncertainties such as the patient's position on the treatment bed, the precise location of tumor cells, and potential anatomical changes that may occur during or between treatment segments. Robust treatment plans should be insensitive to any errors arising from such uncertainties.
[0004] Robust programming methods are known that directly consider uncertainties in the patient's position and / or geometry during optimization. Possible realizations of the uncertainty are typically discretized into multiple scenarios, each corresponding to a specific realization of the uncertainty. As a simple example, different scenarios can be defined corresponding to different possible set errors based on different rigid translations of the patient.
[0005] For example, the same applicant disclosed in WO2016 / 070938 a method for generating robust radiotherapy treatment plans in the face of uncertainties related to treatment planning (including extent uncertainty, patient setup uncertainty, and deformations in organ motion and patient geometry). For example, the plan may include different possible locations of the target volume, and voxels may be weighted according to how many of the possible locations of the target volume are included in the voxels.
[0006] In some cases, the material properties of the patient area are also uncertain. Material properties, such as density, will affect the propagation of radiation through the patient. For example, in photon radiotherapy, a structure in the beam path with a higher actual density than the planned assumed structure will reduce the actual dose delivered to the target. In proton therapy, a higher or lower actual density will shift the position of the Bragg peak, causing the dose to be delivered at different locations. Both scenarios are undesirable.
[0007] The uncertainties that may arise in the properties of materials include:
[0008] • Implants or prosthetic devices made of unknown materials. Such devices are typically made from one of a limited number of suitable materials, such as titanium alloys or various biocompatible alloys. Their shapes are usually well-defined.
[0009] Other additive structures, such as dental fillings, may have undefined shapes and a wide range of materials.
[0010] • Nasal fillers, air, or mucus. Again, the shape of the nasal cavity is relatively constant, but they can become more or less blocked.
[0011] • Colonic bubble formation
[0012] Breast swelling
[0013] In some cases, including the latter two, the shape of the unknown material structure is also uncertain. Attempts have been made to address uncertainties in patient geometry, such as bubble formation, by considering segmented images to obtain up-to-date information on this aspect, but this does not always provide reliable material property values. Furthermore, even using the most recent images, bubbles can be able to move around after the images are taken. Additionally, in some cases, capturing new images of the patient on the day of treatment segmentation is undesirable or impractical. For such situations, performing robust planning in advance may be advantageous. For implants or prosthetic devices, standard practice is to select beam angles that avoid these areas. If the beam is planned to pass through or near these areas, a suitable set of material properties is typically selected and used during planning, such as the most probable material or the average of different probable materials. The dose produced by the plan can be evaluated, and if necessary, a new plan can be developed using a different set of material values.
[0014] Van de Water et al., “Anatomical robust optimization to deal with variation in nasal cavity filling during IMPT,” *Radiotherapy and Oncology*, Vol. 123, Supplement 1, pp. 437–438, discuss different approaches to processing intersegmental variations in nasal cavity filling based on a large number of artificial CT images. The proposed methods vary the planned target volume margin, employ anatomically robust optimization, or online planning adaptation, where a new treatment plan is generated for each artificial image. Summary of the Invention
[0015] The purpose of this invention is to provide more reliable treatment plans when the material properties of one or more areas within or around a patient are unknown. The areas around the patient may include one or more of a support, fixation device, injection device, and / or a bed.
[0016] This invention relates to a computer-based method for generating radiotherapy treatment plans for patients, the method comprising the following steps:
[0017] a. Obtain images of the patient and the desired dose for at least one part of the patient.
[0018] b. Identify at least one structure in the image where at least one parameter has uncertainty.
[0019] c. For the at least one parameter, define two or more different scenarios for the structure, each scenario including a set of material coverage values in the structure, the values corresponding to different possible values of the at least one parameter.
[0020] d. Perform calculations based on the at least two scenarios to provide robust assessment data for each of the at least two scenarios and / or a robustly optimized treatment plan with respect to all values in the set of material coverage values.
[0021] Therefore, this invention utilizes the functionality of a treatment planning system that allows setting material property values in an image to define different scenarios for one or more portions of the image where uncertainty exists, using different material property values. Uncertainty may exist in the materials or material composition used in the region, but this uncertainty may also involve uncertainty in the position or shape of the portion.
[0022] In some embodiments, the at least one parameter relates to the material properties of the structure, and a set of material coverage values relates to at least one material property of the structure. This is relevant, for example, in the case of implants where the material of the implant is unknown, but may be one of a variety of known materials. At least one material property of the structure may include one or more material properties, as will be discussed below.
[0023] The structure can be an implant or prosthetic device already inserted into the patient's body. It can also be a natural part of the patient's body, whose shape and / or contents may vary, such as the intestines, bladder, or nasal cavity. Alternatively, the structure can be external to the patient, such as a bed, chair, fixation device, or injection device, which can also affect dose delivery.
[0024] In some embodiments, at least one structure is added to an image by changing the material settings in a portion of the image, and the parameters are related to the location of that portion of the image, such that the material property values are changed for different portions of the image in different scenes, and the method further includes the step of obtaining a definition of the structure.
[0025] As is common in robust planning, different weights can be assigned to different possible values based on the probability of different possible values during the calculation process.
[0026] If the calculations are performed to provide robust assessment data, the method may further include a step of evaluating at least two scenarios using the robust assessment data. This will provide an assessment of the program's effectiveness in each different scenario, for example, to indicate whether the program is clinically acceptable for all scenarios. If it is found that the program is not clinically acceptable for all scenarios, the program can be abandoned.
[0027] In some embodiments, the invention is based on providing robust data for optimizing or evaluating treatment plans, while taking into account uncertainties in material properties, shape, and / or location due to possible material settings in the data. For optimization, this means that possible material property values have been considered during the planning process to develop a plan that will be adequately applicable to all different scenarios of the unknown structure, including possible variations in material properties and / or structural location and / or shape. This is achieved by providing the planning device with a list of different material properties and / or shapes and taking all of these material properties into account during optimization. In this context, it is achieved as a material overlay. Other systems may have different names for the same function, such as density overlay or CT overlay.
[0028] The method according to the invention can also be used to address uncertainties in patient placement on treatment supports (such as couches or chairs) and / or other external devices (such as fixation devices or injection devices), which may have different materials and / or different thicknesses in different areas. The support is typically added to the images during planning because the planning images are taken with the patient on another type of support. This can also be achieved through material overlay or corresponding functionality, depending on the planning system used. In some cases, synthetic CT data will be interpolated into the CT images. Movement of the support and / or movement of the patient on the couch will result in different material properties, which will affect any beams delivered through the support.
[0029] In the simplest embodiment, the material property can be either electron density or mass density, depending on the type of radiation, where mass density is used for photons and electron density for electrons. Other types of individual or combined material property information may also be available, such as atomic number Z or mass number A, ionization energy, and mass density. For composite materials, this would include the atomic composition of the material and the relative fraction of the atomic number Z or mass number A, and / or the average ionization energy and mass density of the material. Robust optimization of material properties can be combined with conventional robust optimization of patient geometry. The at least one parameter can also be related to the shape or location of the structure, where the material property function is used to set different scenarios for the material properties in the structural region based on different shapes or locations.
[0030] The present invention also relates to a computer program product comprising computer-readable code means that, when executed in a computer, cause the computer to perform methods according to any embodiment disclosed herein. The computer program product may be stored on a non-transitory memory device. The present invention also relates to a computer system comprising a processor and a program memory, wherein the program memory includes such a computer program product to be executed in the processor.
[0031] In a specific embodiment, the present invention relates to a computer-based method for generating radiotherapy treatment plans for patients, the method comprising the following steps:
[0032] • Obtain images of the patient and the expected dose for at least one part of the patient.
[0033] • Identify at least one structure in or near the patient that has unknown material properties.
[0034] • Obtain a set of different possible values for at least one material property.
[0035] • Perform robust optimizations for different possible values.
[0036] In various embodiments, the present invention relates to a computer-based method for generating radiotherapy treatment plans for patients, the method comprising the following steps:
[0037] • Obtain images of the patient and the expected dose for at least one part of the patient.
[0038] • Identify at least one structure to be added to the image by changing the material settings and having an unknown precise location relative to the patient.
[0039] • Obtain a set of different possible values for the position.
[0040] • Perform robust optimizations for the different possible values.
[0041] In both cases, robust optimization ensures that the final plan has sufficiently high quality for all different possible material property values or location values. Attached Figure Description
[0042] The invention will now be described in more detail with reference to examples and the accompanying drawings.
[0043] Figure 1 and Figure 2 The different cases where the material properties of the patient area are unknown are shown.
[0044] Figure 3 This illustrates a situation where the exact location of a structure to be added to a medical image is unknown.
[0045] Figure 4 This is a flowchart of a method according to an embodiment of the present invention.
[0046] Figure 5 This is a schematic diagram of a computer system in which the method of the present invention can be implemented. Detailed Implementation
[0047] Figure 1 A cross-section of a medical image passing through patient 11 is disclosed, including some internal organs 13, hip joint 15, and left hip joint prosthesis 17. As can be seen, any beam entering patient 11 from the left side will be affected by prosthesis 17, and the material properties of the prosthesis will affect the dose and / or the positioning of dose delivery within the patient. Different materials are used in implants and prosthetic devices, including many biocompatible metals and alloys. Different materials have different material properties, including density, which will have different effects on radiation and should therefore be considered in treatment planning. If the material and its properties are unknown, according to embodiments of the invention, a list of possible materials commonly used in such implants and the material property values of these different materials are obtained, and different scenarios are defined using different material property values for use in optimization processes that are arranged to perform robust optimization with respect to the different material property values.
[0048] Foreign body materials can exist as implants in various parts of the body or as dental fillings. All of these foreign body materials will affect any radiation passing through them, and the manner of effect will depend on the shape of the implant and the properties of the material. Therefore, the shape of the implant and the properties of the material need to be considered for different parts of the body in radiation therapy planning. The list of possible materials and their properties may differ for different types of prostheses or implants. For example, for hip replacement, metals such as stainless steel or titanium and their alloys are commonly used. In other applications, ceramics such as zirconium oxide or calcium compositions, or polymers such as silicone or collagen are used. Dental fillings can include a variety of different materials, including plastics, ceramics, dental amalgam, and gold.
[0049] The properties of the material may vary in certain areas of the body. For example, Figure 2 A schematic diagram of a human head 21 is shown as an example. The head has multiple nasal cavities 23, which may be filled with different proportions of air and mucus, which will affect radiation passing through the nasal cavities in different ways. In order to obtain a treatment plan that is effective when the nasal cavities are mostly filled with air and more or less filled with mucus, different material property values of air, mucus, and combinations of both can be used in robust planning.
[0050] Other body regions where material properties may change include: the chest, which may contain more or less fluid; the breasts, which may be more or less swollen; and the intestines, which may contain air pockets. For each of these regions, a set of possible material property values covering a range of different possibilities can be obtained and fed into the optimization process.
[0051] Figure 3 Patient 31 is shown placed on a couch 33 of a radiotherapy delivery system. As mentioned above, the patient's CT images do not include the couch, and the couch used for imaging is typically different from the couch used for treatment delivery. Therefore, depending on the type of planning system used, the couch is added during treatment planning as a material covering or a corresponding function. The shape and material of the couch itself may be known, but the position of the couch relative to the patient is often uncertain. Figure 3 As can be seen, different parts of the bed have different shapes. For example, there is a support structure 35 and a mechanism 37 for moving the bed beneath the bed 33. Different parts of the bed may also include different types of materials. Therefore, different positions of the patient relative to the bed will affect any beams passing through the bed differently. This can be addressed by obtaining a definition of the bed itself to be added as material coverage to the patient's medical image and setting multiple different possible positions for the bed in the material coverage information.
[0052] Other types of structures can be added to medical imaging, either as a substitute for or supplement to a bed. For example, the patient can be placed in a chair or other type of support, and / or external devices (such as injection devices) can be placed in the radiation path. Fixation devices can also be applied to ensure the patient is in the correct shape and positioned in a non-moving manner. For each of these fixation devices, uncertainties in the relative position between the patient and the added structure can be addressed by setting different positions for the added structure in the material cover or corresponding function.
[0053] In some cases, the structures involved may have unknown shapes, or their shapes may change, which can affect treatment in ways that cannot be precisely known.
[0054] Figure 4 This is a flowchart of the method according to the present invention. In the first step S41, input data is provided for the process. This input data includes a medical image of the patient and a desired dose for at least a portion of the image (e.g., a tumor).
[0055] In step S42, a structure is identified for which one or more of the following factors—material properties, location, and shape—are uncertain. For the same structure, two or three of these factors may be uncertain, and the situation could be: uncertainty in the location of one structure, uncertainty in the material properties of another structure, and uncertainty in the shape and material properties of a third structure. Of course, these are just examples; any combination of uncertainties may exist in one or more structures.
[0056] In step S43, a set of possible values for the structure is defined. Given the location and shape of the structure, such as... Figure 1 and Figure 2 In the cases shown, such possible values will be related to material properties, as discussed below. In other words, for one or more material properties discussed, there will be at least a first set of possible values and a second set of possible values. Given the shape and material of the structure, such as... Figure 3 As shown, the possible values will be related to the location of the structure. In other words, there will be at least a first and a second possible location in the image where the structure should be added, each corresponding to the possible relative position of the bed and the patient.
[0057] In step S44, optimization is performed using an optimization problem that takes into account a set of possible values for the desired dose and structure, such that the result of the optimization will be a plan that is good enough for all possible values.
[0058] The choice of material properties to consider can be tailored to specific circumstances, such as the type of radiation used. For example, proton or ion therapy may require more specific material property information than photon therapy. Material property values may be related to one or more of the following:
[0059] Density / mass density
[0060] • Atomic composition along with the relative weights of different atoms
[0061] • Atomic number Z of an atom
[0062] • Atom mass number A
[0063] Average ionization energy
[0064] Such as combination Figure 3 The placement of structures, particularly those to be added to an image, can also present uncertainties. For example, uncertainties may exist regarding the shape of the structure, such as in the case of dental fillings or air bubbles in the intestines. One way to address this issue in an image is by applying material overlay to the areas where the structure should be placed. According to embodiments of the invention, multiple different positions and / or shapes can be set to a set of values for material overlay, and robust optimization can be performed with respect to these values.
[0065] As is common in robust planning, different possible values within a set of values may be assigned different weights depending on their likelihood of occurrence. For example, the value indicating air overload in the sinuses mixed with a certain amount of mucus might be considered the most likely and therefore given the highest weight. Similarly, the relative position of placing the patient close to the center of the bed could be assigned the highest weight.
[0066] Figure 5 This is a schematic diagram of a computer system capable of executing the treatment planning method of the present invention. Computer 51 includes a processor 53, a data memory 54, and a program memory 56. Preferably, one or more user input devices 58, 59 are also present, in the form of a keyboard, mouse, joystick, voice recognition device, or any other available user input device. The user input devices may also be arranged to receive data from external memory units.
[0067] Data memory 54 includes the necessary data for performing the method, such as the desired dose distribution and segmented patient images. Program memory 56 stores the computer program configured to cause the computer to perform actions such as... Figure 2 The method steps outlined according to some embodiments of the present invention are described below.
[0068] As will be understood, data memory 54 and program memory 56 are schematically shown and discussed. There may be several data memory units, each storing one or more different types of data, or a single data memory storing all data in a suitably structured manner, and the same applies to program memory. Both programs and data can reside in one or more memories within the computer system or in another unit accessible from the computer system.
[0069] In the simplest embodiment, the material property can be either electron density or mass density, depending on the type of radiation, where mass density is used for photons and electron density for electrons. Other types of individual or combined material property information may also exist, such as atomic number Z or mass number A, ionization energy, and mass density. For composite materials, this would include the atomic composition of the material along with the relative fraction of the material's atomic number Z or mass number A, as well as the material's average ionization energy and mass density. Robust optimization of material properties can be combined with conventional robust optimization of patient geometry. This at least one parameter can also be related to the shape or location of the structure, where the material property function is used to set different scenarios for the material properties in the structural region based on different shapes or locations.
Claims
1. A computer program product comprising computer-readable code means, which, when executed in a computer, causes the computer to perform a computer-based method for generating a radiotherapy treatment plan for a patient, the computer-based method comprising the following steps: a. Obtain images of the patient and the desired dose for at least one part of the patient. b. Identify at least one structure in the image where at least one parameter exhibits uncertainty, wherein the structure includes at least one of the following: an implant or prosthetic device, a nasal cavity, a bubble, a breast swelling, or a structure external to the patient. c. For the at least one parameter, two or more different scenarios are defined for the structure, each scenario including a set of material coverage values for the structure, the values corresponding to different possible values of the at least one parameter, and d. Perform calculations based on the at least two scenarios to provide robust assessment data for each of the at least two scenarios and / or a robustly optimized treatment plan with respect to all values in the set of material coverage values.
2. The computer program product according to claim 1, wherein, The at least one parameter is related to the material properties of the structure, and the set of material coverage values is related to at least one material property of the structure.
3. The computer program product according to claim 2, wherein, The at least one material property of the structure includes at least one of the following: density, atomic composition, one or more atomic numbers A, one or more mass numbers Z, and average ionization energy.
4. The computer program product according to any one of claims 1 to 3, wherein, The at least one structure is added to the image by changing the material settings in a portion of the image, and the parameters are related to the position of the portion of the image. The computer-based method further includes the step of obtaining a definition of the structure.
5. The computer program product according to claim 1, wherein, The structure is a couch, chair, fixing device, or injection device.
6. The computer program product according to any one of claims 1 to 3, wherein, The at least one parameter is related to the shape of the structure.
7. The computer program product according to any one of claims 1 to 3, wherein if the calculation is performed to provide robust evaluation data, the method includes the step of using the robust evaluation data to evaluate the at least two scenarios.
8. The computer program product according to any one of claims 1 to 3, comprising the step of assigning different weights to the different possible values during the calculation process.
9. A computer system comprising a processor and a program memory, wherein, The program memory includes a computer program product according to any one of claims 1 to 8.