Method for treatment plan used in radiation therapy implemented in computer structure
The convex optimization method in radiotherapy planning addresses the challenge of balancing target dose delivery with healthy tissue protection by penalizing low-dose volumes and adjusting beam settings, enhancing treatment efficacy and safety.
Patent Information
- Application Number
- JP2024156983
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2018-02-21
- Filing Date
- 2024-09-10
- Publication Date
- 2024-11-15
AI Technical Summary
Current radiotherapy planning methods struggle with optimizing treatment plans that balance delivering high doses to targets while minimizing exposure to surrounding healthy tissues, particularly in cases with multiple targets close to each other, leading to non-convex optimization problems that are difficult to solve efficiently.
A convex optimization approach that manipulates delivered radiation by penalizing voxels receiving doses above a threshold, using a frame description to suppress low-dose volumes and adjust beam shape settings, collimator configurations, and isocenter positions to create an optimal treatment plan.
This method effectively reduces the dose to surrounding healthy tissues and organs at risk, improving treatment efficiency and reducing adverse effects, while ensuring adequate target coverage.
Smart Images

Figure 2024161301000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to the field of radiation therapy, in particular to a method and system for planning and optimizing a patient treatment session in a radiation therapy system. [Background technology]
[0002] The development of surgical techniques has made great strides over the years. For example, patients requiring brain surgery can now instead undergo non-invasive procedures that dramatically reduce trauma to the patient.
[0003] Systems for non-invasive surgery include the Leksell GammaKnife® Icon™ and the Leksell GammaKnife® Perfexion, which enable such surgery by gamma irradiation. Radiation is emitted from multiple fixed radioactive sources and is focused toward a defined target or treatment volume by collimators, i.e., passages or channels to obtain a beam of limited cross-section. Each radiation source delivers a dose of gamma radiation that is not sufficient by itself to damage the intervening tissue. However, tissue destruction occurs when radiation beams from multiple radiation sources intersect or converge, causing the radiation to reach tissue-destroying levels. The point of convergence is hereafter referred to as the "focal point."
[0004] Treatment plan optimization in radiation therapy, including, for example, Gamma Knife radiosurgery, aims to deliver a sufficiently high dose to a target volume within a patient (e.g., in the treatment of a tumor) while simultaneously minimizing the dose delivered to adjacent normal tissue. In treatment plan optimization, at least three competing factors must be considered: delivering a sufficiently high dose to the target volume, sparing surrounding normal or healthy tissue, and keeping treatment time as short as possible.
[0005] Treatment plan optimization is a process that involves optimizing the relative isocenter position or beam direction, beam shape settings (e.g., collimator configuration), and fluence. For example, in the Leksell GammaKnife® Icon™ and Leksell GammaKnife® Perfexion, treatment plan optimization may include optimizing the number of shots used, the collimator configuration for each shot, the shot time, and the location of the shot. Irregularities and size of the target volume greatly affect the relative isocenter position or beam direction, beam shape settings (e.g., collimator configuration), and fluence used to optimize treatment.
[0006] In treatment planning, inverse treatment planning has gained increasing interest. Inverse planning generally refers to a stage in treatment planning where an achievable treatment plan is sought such that several criteria are met. Inverse planning can be contrasted with forward planning, where the practitioner manually places, adjusts and shapes the shots. Inverse planning promises shorter planning times and higher quality plans. Inverse planning is sometimes tightly integrated with forward planning, for example in LeksellGammaPlan®, the software that accompanies the Leksell Gamma Knife. It is based on relative isodose, using a metric that is well known in radiosurgery. This makes the transition from forward to inverse planning easier and is thought to be one of the reasons why inverse planning has been widely adopted. The drawback of the complexity of relative isodose-based inverse planning means and objectives is that the resulting optimization problem is inherently difficult to solve. In realistic cases, a compromise between computation time and the risk of poor local optima is necessary. This makes it difficult to explore what trade-offs are achievable, especially in complex cases with multiple conflicting objectives, such as multi-metastatic cases where at least one metastasis is adjacent to an organ at risk. Incidentally, in such cases it may be desirable to specify some criteria that must be met. Current backward planning tools do not allow for setting organ at risk (OAR) criteria.
[0007] Historically, in inverse treatment planning for Gamma Knife radiosurgery, relative isodose is the fundamental objective of interest. It is a heuristic problem solution motivated by the fact that the dose falloff is steepest at a certain isodose level that should coincide with the target boundary. Incidentally, this is true for single shots, but not necessarily when the dose distribution is the sum of contributions from multiple shots. Note that utilizing steepness assumes high positional accuracy. For the isocenter, the optimization variables are the position, the collimator configuration, and the beam-on time. The isocenter position is moved during the optimization, and the collimator configuration is treated as an individual element in the set of all possible collimator configurations. Organs at risk are not explicitly handled in the objective function, which can be a severe limitation. Obviously, the tolerated dose for the organs at risk is given in absolute doses, but in the current planning mode, the absolute dose is only assigned after completing the plan. This results in an optimization problem that is very difficult in the sense that any solution requires extensive computations or runs the risk of returning an unsatisfactory solution.
[0008] In an improved inverse treatment planning method provided by the applicant, multiple objectives reflecting clinical criteria of the region of interest, including targets to be treated during treatment of a patient, organs at risk and / or healthy tissues are set, and radiation dose profiles to be delivered to the targets are generated. A convex optimization problem is provided to manipulate the radiation to be delivered according to the objectives, and dose profiles for specific treatment configurations including beam shape settings for the radiation dose profiles are calculated using the convex optimization problem. A treatment plan is then created that includes determining the radiation dose profiles to be delivered during treatment based on the treatment configurations, each radiation dose profile is modeled by a spatial dose volume distribution of radiation, the shape of the spatial distribution depends on the beam shape settings, and an optimal treatment plan that meets the clinical criteria is selected.
[0009] However, there is a wealth of clinical data showing that adverse cognitive effects can occur when large amounts of normal tissue are also irradiated at relatively low doses. Therefore, a key objective of planning methods is to achieve low dose delivery to normal tissue surrounding one or more target volumes, which is particularly challenging when multiple targets are close to each other. The main problem in optimization is that the set of voxels within the irradiated volume depends on the dose distribution and changes during optimization, which results in a non-convex optimization problem that is generally difficult to solve.
[0010] Thus, there remains a need for more efficient methods for planning and optimizing treatment. Summary of the Invention [Problem to be solved by the invention]
[0011] It is an object of the present invention to provide an improved method and system for planning and optimizing treatment sessions of a patient treated with radiation therapy. [Means for solving the problem]
[0012] This and other objects are achieved by the invention as defined by the independent claims. Preferred embodiments are defined by the dependent claims.
[0013] The term "target" or "target volume" refers to a representation of a target in a patient being treated during radiation therapy. The target may be a tumor being treated by radiation therapy. Typically, the representation of the target is obtained by non-invasive image capture, for example, using X-rays or nuclear magnetic resonance.
[0014] The term "shot" refers to the delivery of radiation to a predetermined location within a target volume with a predetermined level and spatial distribution of radiation. A shot is delivered for a predetermined period of time ("beam-on" time) through at least one sector of a collimator of a treatment system using one of the sector's states. A "composite shot" refers to the delivery of radiation to a focal spot using different collimator sizes for different sectors.
[0015] The term "beam on time" refers to the predetermined period of time during which a shot is delivered to a target volume.
[0016] The term "constraint" refers directly to a constraint on an optimization variable, e.g., enforcing a non-negative beam on-time, or indirectly to enforcing delivery of a minimum dose to a particular volume. Constraints can also refer to constraints that must not be violated (hard constraints) and / or constraints that are allowed to be violated but with a penalty in the objective function (soft constraints).
[0017] The term "voxel" is used in the context of this application to refer to a volume element on a lattice that may be anisotropic in three-dimensional space.
[0018] The term "frame description" includes at least an objective function and / or constraints for a set of voxels.
[0019] The present invention can be used, for example, in radiation therapy. Radiation therapy is used to treat cancer and other diseases in mammalian (e.g., human and animal) tissues. One such radiation therapy device is the Gamma Knife, which delivers a large amount of low-intensity gamma radiation to the patient that is focused with high intensity and precision at the target (e.g., tumor). Another radiation therapy device uses a linear accelerator that delivers high-energy particles (e.g., photons, electrons, etc.) to the tumor. Yet another radiation therapy device, the cyclotron, uses protons and / or ions. Another form of radiation therapy is brachytherapy, in which a radiation source is placed inside or next to the area needing treatment. The direction and shape of the radiation beam must be precisely controlled to ensure the tumor receives the prescribed radiation dose, and radiation from the beam must minimize damage to surrounding healthy tissue, often referred to as the organ at risk (OAK)(s). The radiation beam parameters can be controlled using a treatment plan, and the radiation therapy device performs the treatment by delivering a spatially varying dose distribution to the patient.
[0020] The invention is for example used in connection with treatment planning of a treatment provided by a radiotherapy system having a collimator body with several groups or sets of collimator passages, each set designed to provide a radiation beam of a respective specific cross section directed to a fixed focal spot. Suitably, the inlets of each set of collimator passages have a pattern that essentially corresponds to the pattern of the source on the source carrier configuration. These sets of collimator passage inlets can be configured such that they can be changed from set to set, thereby changing the resulting beam cross section and the spatial dose distribution surrounding the focal spot. The number of sets of collimator passages with different diameters can be more than two, for example three or four, or even more. A typical embodiment of the collimator includes eight sectors, each with four different states (beam off, 4 mm, 8 mm, and 16 mm). To change the spatial distribution of radiation around the focal spot, the sectors can be adjusted individually, i.e. different states can be selected for each sector.
[0021] The present invention may be used in treatment planning for intensity modulated radiation therapy (IMRT) treatments.
[0022] The present invention can further be used in brachytherapy. Brachytherapy is a form of radiation therapy in which a radiation source is placed inside or next to the area in need of treatment. Brachytherapy involves precise placement of a short-range radiation source (a radioisotope, e.g., iodine-125) directly at the site of the cancerous tumor. Brachytherapy treatment planning often includes optimization methods to calculate the dwell time and dwell position of the radioactive source along a specified applicator path. Inverse planning methods for brachytherapy aim to obtain adequate target coverage and maximum preservation of vital structures. In geometric optimization, the relative dwell time is determined by the implant geometry by assigning individual weighting factors for the dwell time at each dwell position that are inversely proportional to the dose contribution from adjacent source positions.
[0023] Thus, the optimization problem, which in a preferred embodiment is a convex optimization problem, manipulates the delivered radiation according to the objective, and a dose profile for a particular treatment configuration, including source strength and / or dwell time, is calculated using the optimization problem. A treatment plan can then be created that includes determining the radiation dose profile to be delivered during treatment based on the treatment configuration, where each radiation dose profile is modeled by a spatial dose volume distribution of radiation, where the shape of the spatial distribution depends on the setting, and an optimal treatment plan that meets the criteria can be selected.
[0024] According to an aspect of the present invention, a method is provided for dose or treatment planning of a radiation therapy system including a radiation therapy unit, where the delivered spatial dose can be changed by adjusting beam shape settings, and an optimization problem is used to determine the delivered radiation. The optimization problem manipulates the delivered radiation according to an objective reflecting a criterion for a region of interest including at least one of at least one target, organ at risk, and / or healthy tissue to be treated during the treatment of a patient. The method includes estimating a voxel set that receives a dose higher than a predetermined threshold dose level, the voxel set including voxels from at least one target volume. Further, a low-dose voxel set is determined, and a frame description is provided for voxels in the low-dose voxel set, and voxels that receive a dose above a predetermined threshold are penalized according to the frame description such that the dose delivered to the low-dose voxel set is suppressed. The frame description is then used in the optimization problem to manipulate the delivered radiation.
[0025] The optimization problem, which in a preferred embodiment is a convex optimization problem, manipulates the delivered radiation according to an objective, and a dose profile for a particular treatment configuration, including a beam shape setting of the radiation dose profile, is calculated using the optimization problem. A treatment plan can then be created that includes determining a radiation dose profile to be delivered during treatment based on the treatment configuration, where each radiation dose profile is modeled by a spatial dose volume distribution of radiation, the shape of the spatial distribution depending on the beam shape setting, and an optimal treatment plan that meets the criteria can be selected. According to another aspect of the present invention, a method is provided for treatment planning of a radiation therapy system comprising a radiation therapy unit with a fixed radiation focal spot, where the spatial dose distribution surrounding the focal spot can be changed by adjusting the beam shape setting, including the collimator setting. The collimator is configured into a plurality of sectors and has a plurality of collimator passage entrances that direct radiation emitted from a radioactive source of a source carrier configuration of the therapy system toward the focal spot. The method includes estimating a set of voxels that will receive a dose higher than a predetermined threshold dose level, where the set of voxels includes at least one target volume. A low-dose voxel set is then determined, and a frame description is determined that reflects criteria for voxels in the low-dose voxel set such that the dose delivered to the low-dose voxel set is constrained. The frame description is then used in an optimization problem that steers the delivered radiation.
[0026] The optimization problem, which in a preferred embodiment is a convex optimization problem, manipulates the delivered radiation according to the objective, and the dose rate for a particular treatment configuration, including sector and collimator settings and exposure time at the isocenter, is calculated using the optimization problem. A treatment plan can then be created that includes determining the shots to be delivered during treatment based on the treatment configuration, with each shot being modeled by a spatial dose volume distribution of radiation represented by a three-dimensional voxel representation. The shape of the spatial distribution depends on the particular sector and collimator settings and exposure time. Finally, an optimal treatment plan that meets the criteria can be selected.
[0027] According to an embodiment of the present invention, a dose distribution is estimated. A first estimate of a low-dose voxel set is determined. At least one optimization with either a weight setting to be used in a final optimization or a standardized weight setting is performed, and a low-dose voxel set defined as a volume between two dose levels is updated based on the optimized dose distribution in the previous optimization step. The frame description is modified according to the updated low-dose voxel set. A final optimization is performed using the updated low-dose voxel set and the modified frame description from the previous optimization.
[0028] In an embodiment of the invention, the dose distribution is estimated by estimating the dose distribution outside each target independently or by estimating the combined dose distribution outside the targets.
[0029] In particular, low-dose rings or volumes are generated using an iterative approach. First, for each of the N targets, the dose distribution d i (r) is estimated, i.e., for i=1,...,N. The estimate may depend on geometric factors, such as the target area A, the target volume V, and the vector r from a point on the target to the location of interest (or derived properties such as distance and / or direction from the target). Furthermore, the dose distribution estimate can be expressed as the prescribed dose D T In the case of multiple targets, all dose distributions outside the targets are combined by a possible weighted sum using weights w. In one embodiment, the combined dose distribution is given by:
number
[0030] Based on the combined dose distribution, a first estimate of the low dose volume is determined. In an example, two dose levels, e.g. the volume between these levels, can be selected as 40%-60% of the minimum prescribed dose. Then, for example, a threshold dose D is added to the objective function. lrThe frame description is modified by adding a penalty term that penalizes voxels with doses above . For example, the penalty term might be:
number
[0031] According to an embodiment of the present invention, the method further comprises determining a low-dose voxel set by applying a predetermined dose distribution.
[0032] In an embodiment of the present invention, the step of determining the low-dose voxel set by applying a predetermined dose distribution includes the steps of calculating a first shell surface outside the voxel set based on dose levels higher than a threshold dose level, calculating a second shell surface outside the voxel set based on dose levels lower than the threshold dose level, and defining the low-dose voxel set as the voxel set between the first shell surface and the second shell surface.
[0033] According to an embodiment of the present invention, the step of determining the low-dose voxel set includes the steps of calculating a first shell surface outside the voxel set based on a predetermined geometric shape, calculating a second shell surface outside the voxel set based on the predetermined geometric shape, and defining the low-dose voxel set as the voxel set between the first shell surface and the second shell surface.
[0034] In an embodiment of the invention, calculating an outer first shell surface of the set of voxels comprises uniformly shrinking an outer surface of the set of voxels that receive a dose higher than a predefined threshold dose level.
[0035] In an embodiment of the invention, calculating the outer second shell surface of the set of voxels comprises uniformly expanding the outer surface of the set of voxels that receive a dose above a predefined threshold dose level.
[0036] In an embodiment of the invention, the method further comprises the steps of finding an isocenter position, determining a shot collimator configuration for every isocenter position, setting a dose to be delivered to at least one target, including determining exposure times for every shot to be delivered, and calculating a dose distribution based on the determined shot configurations and the dose to be delivered.
[0037] In an embodiment of the present invention, the low-dose voxel set is a spherical shell with uniform voxel thickness.
[0038] According to an embodiment of the present invention, the calculating step includes applying a distance model according to:
number
[0039] In an embodiment of the present invention, the calculating step comprises applying a distance model according to:
number
[0040] According to an embodiment of the invention, the method further includes using the optimization problem to calculate a dose profile for a particular treatment configuration including a beam shape setting of the radiation dose profile, creating a treatment plan including determining a radiation dose profile to be delivered during treatment based on the treatment configuration, where each radiation dose profile is modeled by a spatial dose volume distribution of radiation, where the shape of the spatial distribution depends on the beam shape setting, and selecting an optimal treatment plan that meets a criteria.
[0041] In an embodiment of the invention, the method includes using an optimization problem to calculate dose rates for a particular treatment configuration including sector and collimator settings and exposure time at the isocenter, creating a treatment plan including determining shots to be delivered during treatment based on the treatment configuration, where each shot is modeled by a spatial dose volume distribution of radiation, the shape of the spatial distribution being dependent on the particular sector and collimator settings and exposure time, and selecting an optimal treatment plan that meets a criteria.
[0042] According to an embodiment of the present invention, the method further includes the steps of defining a set of beam directions; modeling the radiation dose profile to be delivered to the target as a number of beamlets, each having a beamlet intensity; setting a number of objectives reflecting criteria for the target; providing an optimization problem that manipulates the delivered radiation according to the objectives to create a fluence map, the fluence map defining the beamlet intensity for each beamlet; creating a treatment plan based on the fluence map and the criteria for the target; and selecting an optimal treatment plan that meets the criteria.
[0043] According to an embodiment of the invention, a method includes positioning a radiation source(s) relative to a patient, including generating a fixed isocenter position.
[0044] In an embodiment of the present invention, radiation source positions are generated as a set of contiguous points within the target volume based on basis functions, and the points are fixed during treatment planning.
[0045] According to embodiments of the present invention, objectives include delivered dose to the target, delivered dose to the boundary space surrounding the target, delivered dose to areas classified as organs at risk, and / or beam-on time penalization.
[0046] The optimized dose plan determined by the present invention can be transferred to a radiation therapy system for use in treating a patient. The dose plan determined by the present invention can also, or alternatively, be used as input in a treatment optimization procedure in which the number of shots, the location and beam-on time of each shot, and the shot size defined during the volume filling according to the present invention serve as a basis in optimizing the number of shots, the location and beam-on time of each shot, and the shot size.
[0047] According to yet another aspect of the invention there is provided a treatment planning computer architecture capable of implementing the method according to the invention.
[0048] In an embodiment of the present invention, the treatment planning computer structure can utilize the method according to the present invention and can be integrated into a system for delivering intensity modulated radiation therapy (IMRT) including a radiation source generating at least one radiation beam. A beam shaping device, for example a multi-leaf collimator or a cone collimator, can be arranged between the radiation source and the patient. The collimator is communicatively connected to the treatment planning computer structure and configured to modulate a plurality of beamlets to be delivered to the patient according to an optimal treatment plan, i.e. a fluence map based on the determined beam shape settings.
[0049] In an embodiment of the invention, the low-dose voxel set is determined based on predictions made using a machine learning algorithm trained on a database of treatment plans, such as a linear regression model, a logistic regression model, a support vector machine, a Gaussian process model, a random forest, a Markov random field, a conditional random field, an artificial neural network, or a deep learning network such as a convolutional neural network.
[0050] In a further embodiment of the present invention, a treatment planning computer structure can utilize the method according to the present invention and can be integrated into a radiotherapy system having a collimator body with several groups or sets of collimator passages, each set designed to provide a radiation beam of a respective specific cross section directed to a fixed focal spot. Suitably, the inlets of each set of collimator passages have a pattern that essentially corresponds to the pattern of the source on the source carrier arrangement. These sets of collimator passage inlets can be configured such that they can be changed from set to set, thereby changing the resulting beam cross section and the spatial dose distribution surrounding the focal spot. The collimator body is communicatively connected to the treatment planning computer structure for delivery to the patient according to an optimal treatment plan.
[0051] As will be appreciated by those skilled in the art, the steps of the method according to the present invention, and preferred embodiments thereof, are suitable for implementation as a computer program or as a computer readable medium.
[0052] Further objects and advantages of the present invention are explained below by means of exemplary embodiments. [Brief description of the drawings]
[0053] [Figure 1a] FIG. 1 is a perspective view of an assembly comprising a source carrier arrangement surrounding a collimator body in which the present invention may be used. [Figure 1b] FIG. 2 illustrates a radiation therapy device in which the assembly of FIG. 1 can be used. [Figure 2a] FIG. 1 illustrates a radiotherapy device, a Gamma Knife, in which the present invention may be used. [Figure 2b] FIG. 1 illustrates another radiation therapy device, a linear accelerator, in which the present invention may be used. [Diagram 3] 1 is a flow diagram illustrating one embodiment of the present invention. [Figure 4] FIG. 2 is a simplified two-dimensional diagram of the geometry of a backward programming problem according to another embodiment of the present invention. [Diagram 5] FIG. 1 illustrates one embodiment of a treatment planning computer architecture in accordance with the present invention. [Figure 6] 4 is a flow chart illustrating a further embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0054] Referring initially to Figures 1a and 1b, an exemplary radiation therapy device is shown in which a treatment plan developed using the present invention may be used to treat a patient.
[0055] Figure 1a shows a perspective view of an assembly comprising a source carrier arrangement 2 surrounding a collimator body 4. Both the source carrier arrangement 2 and the collimator body 4 have the shape of a truncated cone. The source carrier arrangement 2 comprises six segments 6 distributed along the annular circumference of the collimator body 4. Each segment 6 has a number of openings 8 in which a container containing a radioactive source such as cobalt is placed. The collimator body 4 is provided with a collimator passage or channel, the internal mouth 10 of which is shown in the figure.
[0056] Each segment 6 has two straight sides 12 and two curved sides 14a, 14b. One curved side 14a forms a longer arc and is located near the base of the cone, while the other curved side 14b forms a shorter arc. The segments 6 are linearly displaceable, i.e. they do not rotate around the collimator body 4, but instead are movable back and forth along an imaginary line drawn from the center of the shorter curved side 14b to the center of the longer curved side 14a. Such a translational displacement has the effect of a coordinate transformation such that the new axes are parallel to the old axes.
[0057] As can be seen from FIG. 1a, there are more internal openings 10 or holes of the collimator passages than there are openings 8 for receiving radioactive sources. In this particular case, there are three times as many collimator passages as there are openings for receiving radioactive sources, for example 180 openings and 540 collimator passages. The reason for this is that there are three different sizes of collimator passages in the collimator body 4, or passages directing radiation beams with three different diameters towards a focal point. The diameters can be, for example, 4, 8 and 16 mm. The three different types of collimator passages are each arranged in a pattern corresponding to the pattern of openings in the source carrier arrangement. The desired size or type of collimator passage is selected by linearly displacing a segment 6 of the source carrier arrangement along the collimator body so that it is aligned with the desired collimator passage.
[0058] In Fig. 1b, a radiotherapy system is shown, which includes a radiotherapy device 130 having a source carrier configuration as shown in Fig. 1a and a patient positioning unit 20. The radiotherapy unit 130 is thus provided with a radioactive source, a radioactive source holder, a collimator body and an external shielding element. The collimator body includes multiple collimator channels directed to a common focal point, as shown in Fig. 1a.
[0059] The patient positioning unit 20 comprises a rigid framework 22, a slidable or movable carriage 24 and a motor (not shown) for moving the carriage 24 relative to the framework 22. The carriage 24 is further provided with a patient bed 26 for generally supporting and moving a patient. At one end of the carriage 24, a fixation device 28 is provided for receiving and fixing a patient fixation unit or an interface unit. The coordinates of the fixation unit are defined by a fixed unit coordinate system, which is also used to define the contour of the treatment volume through a fixed relationship with the treatment volume. In operation, the fixation unit, and therefore the fixed unit coordinate system, is moved relative to a fixed radiation focal point such that the focal point is accurately positioned at the intended coordinates of the fixed unit coordinate system.
[0060] FIG. 2a shows a radiation therapy device 130, a gamma knife, in which the present invention can be used. A patient 202 can wear a coordinate frame 220 to keep a body part (e.g., head) of the patient undergoing surgery or radiation therapy stable. The coordinate frame 220 and a patient positioning system 222 can establish a spatial coordinate system, which can be used while imaging the patient or during radiosurgery. The radiation therapy device 130 can include a protective housing 214 for enclosing multiple radiation sources 212 for generating radiation beams (e.g., beamlets) through beam channels 216. The multiple beams may be configured to focus on an isocenter 218 from multiple different positions. Although the intensity of each individual radiation beam can be relatively low, when multiple doses from different radiation beams accumulate at the isocenter 218, the isocenter 218 can receive a relatively high level of radiation. In certain embodiments, the isocenter 218 can correspond to a target undergoing surgery or treatment, such as a tumor.
[0061] FIG. 2b shows another radiation therapy device 40, the linear accelerator 10, in which the present invention can be used. Using the linear accelerator 40, the patient 42 can be positioned on the patient table 43 to receive a radiation dose determined by the treatment plan. The linear accelerator 40 can include a radiation head 45 that generates a radiation beam 46. The entire radiation head 45 can be rotatable about a horizontal axis 47. In addition, under the patient table 43, a flat panel scintillator detector 44 can be provided that can rotate synchronously with the radiation head 45 around the isocenter 41. The intersection of the axis 47 and the center of the beam 46 generated by the radiation head 45 is usually called the "isocenter". The patient table 43 can be powered such that the patient 42 can be positioned such that the tumor site is at or near the isocenter 41. The radiation head 45 can rotate around the gantry 47 to provide the patient 42 with a number of different doses of radiation according to the treatment plan.
[0062] Now referring to FIG. 3, a general method 300 according to the present invention is described. The method can be used for treatment planning in, for example, brachytherapy, radiosurgery, and intensity modulated radiation therapy as well as intensity modulated carotid artery therapy. In a linear accelerator, an electron beam is generated by an electron accelerator including an electron gun, a waveguide, and a guiding magnet. The electron beam strikes a target made of a high atomic number material, thereby generating ionizing radiation. The ionizing radiation can be modeled as a number of beamlets, each having a beamlet intensity that can be modeled according to a fluence map. The fluence map is determined in an optimization. Moreover, the method can also be used for treatment planning in a radiation therapy system having a collimator body with several groups or sets of collimator passages, each set designed to provide a radiation beam of a respective specific cross-section directed to a fixed focal spot. Suitably, the entrances of each set of collimator passages have a pattern that essentially corresponds to the pattern of the source on the source carrier configuration. These sets of collimator passage entrances can be configured such that they can be changed from set to set, thereby changing the resulting beam cross-section and the spatial dose distribution surrounding the focal spot. The number of sets of collimator passages with different diameters may be more than two, for example three or four, or even more. A typical embodiment of the collimator includes eight sectors, each with four different states (beam off, 4 mm, 8 mm, and 16 mm). To vary the spatial distribution of the radiation around the focal point, the sectors can be adjusted individually, i.e. a different state can be selected for each sector.
[0063] The method 300 begins at step 310 by identifying a set of voxels that receive a dose above a predetermined threshold dose level, the set of voxels including voxels from at least one target volume.
[0064] Next, in step 320, a low-dose voxel set is determined. This can be done by applying a predefined dose distribution. In one embodiment of the present invention, this may include calculating a first shell surface outside the voxel set based on dose levels above a threshold dose level and calculating a second shell surface outside the voxel set based on dose levels below the threshold dose level. The low-dose voxel sets are then defined as the voxel sets between the first and second shell surfaces, respectively.
[0065] In another embodiment of the invention, this may preferably include the steps of calculating a first shell surface outside the voxel set based on a predetermined geometric shape, calculating a second shell surface outside the voxel set based on the predetermined geometric shape, and defining a low-dose voxel set as the voxel set between the first shell surface and the second shell surface.
[0066] In step 330, an objective function for voxels in the low-dose voxel set is provided, and voxels that receive a dose above a predetermined threshold are penalized such that the dose delivered to the low-dose voxel set is constrained.
[0067] The objective function is then used in an optimization problem to manipulate the delivered radiation in step 340. Thus, the optimization problem, which in a preferred embodiment is a convex optimization problem, manipulates the delivered radiation according to the objective, and a dose profile for a particular treatment configuration, including a beam shape setting of the radiation dose profile, is calculated using the optimization problem. A treatment plan can then be created that includes determining the radiation dose profile to be delivered during treatment based on the treatment configuration, where each radiation dose profile is modeled by a spatial dose volume distribution of radiation, where the shape of the spatial distribution depends on the beam shape setting, and an optimal treatment plan that meets the criteria can be selected.
[0068] The method according to the invention will now be described in more detail with particular reference to Fig. 4. It should be noted, however, that Fig. 4 shows the geometry of the inverse programming problem in a simplified two-dimensional view. Although the structure is shown as a ring, since the target volume is three-dimensional, the ring is consequently a shell or layer of spheres shape, encompassing the target volume (or tumor volume).
[0069] However, large amounts of V ad There is a wealth of clinical data showing that adverse cognitive effects can occur from relatively low doses of radiation to normal tissues, particularly when multiple targets are in close proximity to one another. The main problem is that the volume is ad The problem is that the set of voxels equal to depends on the dose distribution and varies during the optimization, which leads to a non-convex optimization problem that is difficult to solve in general. Therefore, to achieve a convex formulation it is necessary to have a fixed geometry for which the dose is penalized.
[0070] According to the present invention, a volumetric filling procedure or algorithm is applied, such as the filling algorithm used in Leksell GammaPlan. An example of a suitable filling algorithm is described in a commonly assigned, co-pending, unpublished patent application.
[0071] Based on the use of a filling algorithm, a fixed low-dose ring R of voxels as shown in FIG. 4 can be created in a volume surrounding at least one target (T1, T2, T3). The voxels of this fixed volume can then be penalized giving an efficient low-dose penalization. ad ,D ad ) is assumed to exist, where D ad is a volume at least V ad, which is the dose at which adverse effects become evident when the volume is greater than 13 cc and receives a dose greater than 10 Gy. See FIG. 4. For example, there is clinical evidence suggesting that the risk of radiation necrosis increases when the volume is greater than 13 cc and receives a dose greater than 10 Gy, but the user may specify other values. According to the present invention, such adverse effects can be significantly reduced or eliminated by introducing a penalty term into the optimization problem. As described above, a filling algorithm is used first. This is a preliminary step to optimization. In addition to providing the isocenter location, the filling algorithm also gives the shot collimator configuration for each isocenter. By setting a prescription dose that forces the weight to 1 for all shots, a realistic dose distribution can be calculated. To constrain the dose to large volumes, a low dose ring is created, and the algorithm for creating the low dose ring consists of the following steps: Thus, a filling algorithm is used to find the isocenter location and determine the shot collimator configuration for all isocenter locations. The prescription doses T1, T2, T3 for the target(s) are set. This gives the exposure times for all shots. The dose distribution is then calculated from the above configurations.
[0072] Then, the volume (after adding the target volume) is V ad Voxel SV within a 3D isodose volume with a dose distribution equal to ad Next, the ring of voxels R (R = 2δ r ) but for example, using the following distance model, see Figure 4, SV ad is defined as the set difference between the contraction and expansion of
number
number
number
number
[0073] Here, D F,ad is V ad is the isodose in Gy corresponding to 1 Gy. A "ring size" corresponding to 1 Gy is a reasonable choice to include various dose distributions without introducing too many voxels in the optimization. However, it should be noted that the distance model shown above is one particular example, and according to an embodiment of the present invention, the calculating step includes applying a more general distance model according to:
number
number
[0074] A penalty term is then added to the objective function that penalizes voxels whose dose exceeds the threshold dose. In an embodiment of the invention, it may be of the form:
number
[0075] A treatment planning computer architecture capable of implementing the method according to the invention will now be described with reference to Fig. 5. The treatment planning computer architecture 500 may include a modeling module 510 configured for modeling a patient volume as a three-dimensional voxel representation or for obtaining such a three-dimensional voxel representation. The volume includes a target volume to be treated during treatment of the patient in a radiation therapy unit.
[0076] The low-dose volume calculation module 520 calculates or determines the low-dose volumes by estimating a set of voxel volumes that will receive a dose higher than a predetermined boundary dose level, the voxel volume set including at least one target volume, determining the low-dose voxel volume set by applying a predetermined altered dose distribution, and applying an objective function to voxels in the low-dose voxel volume set, where voxels that receive a dose above a predetermined threshold are penalized such that the dose delivered to the low-dose voxel volume is suppressed. More specifically, the low-dose volume calculation module applies the method or procedure described above with reference to FIG.
[0077] Further, the calculation module 530 is configured to provide a convex optimization problem including an objective function for a low dose volume that generates a radiation dose profile to be delivered to the target and manipulates the radiation delivered according to the objective, and to use the convex optimization problem to calculate a dose profile for a specific treatment configuration including a beam shape setting for the radiation dose profile. The treatment planning module 540 is configured to create a treatment plan including determining a radiation dose profile to be delivered during treatment based on the treatment configuration, where each radiation dose profile is modeled by a spatial dose volume distribution of radiation represented by a three-dimensional voxel representation, and the shape of the spatial distribution depends on the beam shape setting. The optimization module 550 is configured to select an optimal treatment plan that meets clinical criteria. In an embodiment of the present invention, optimization, i.e., selection of an optimal treatment plan that meets clinical criteria, is performed, and then a treatment plan is created including determining a radiation dose profile to be delivered during treatment based on the treatment configuration.
[0078] In an embodiment of the present invention, the treatment planning computer structure 500 can utilize the method as described in Figures 3 and 5 and can be integrated into a system for delivering intensity modulated radiation therapy (IMRT), including a radiation source generating at least one radiation beam and a structure for generating a plurality of beamlets. A beam shaping device, for example a multi-leaf collimator or a cone collimator, can be placed between the radiation source and the patient. The collimator is communicatively connected to the treatment planning computer structure and has a plurality of leaves for modulating the plurality of beamlets to be delivered to the patient according to an optimal treatment plan, i.e. a fluence map based on the determined beam shape settings.
[0079] The method described herein according to the invention may further be used in combination with inner and outer shells that encompass the targets and are associated with each target, with the inner shell promoting selectivity and the outer shell promoting gradient rate, respectively. The frame description of each shell is formulated so that each voxel can be considered individually. According to one embodiment, the terms of the frame description are approximations of integrals, i.e. sums where these terms correspond to voxels at a distance r from the outer surface of the target(s) volume, and may be given by:
number
[0080] The objective function can be formulated voxel-wise for the target and two surrounding shells. According to an embodiment of the present invention, a minimal objective function for one target and therefore two shells, which has neither OAR nor beam-on-time penalization and can be easily generalized to two or more targets, can be formulated as follows:
number
number
[0081] According to an embodiment of the invention, the low-dose rings or volumes are generated using an iterative approach, which will now be described 600 with reference to Figure 6. First, in step 610, the outer dose distribution of each of the target(s) is estimated according to:
number
[0082] In step 620, an initial estimate of the low-dose volume is determined based on the combined dose distribution. In an example, the volume between two dose levels, for example these levels can be selected as 40%-60% relative to the lowest prescribed dose.
[0083] Next, in step 630, the dose is determined to be a threshold dose D lr A penalty term is added to the frame description that penalizes voxels that exceed
number
[0084] Next, in step 640, optimization is performed using either the weight settings used in the final optimization problem or the standardized weight settings.
[0085] It is then determined in step 650 whether the iterative process is finished, e.g., after a predetermined number of iterations. If so, in step 680, a final treatment plan optimization can be performed using the low-dose volume. If not, the procedure 600 proceeds to step 660, where an updated low-dose volume is defined as the volume between the two dose levels, which may be the same as in step 620, and the contributions to the frame description are modified accordingly. The objective function terms are modified in step 670 accordingly.
[0086] While illustrative embodiments of the present invention have been shown and described, it will be apparent to those skilled in the art that numerous changes, modifications, or variations can be made to the invention as described herein. Accordingly, it is to be understood that the above description of the invention and the accompanying drawings are to be considered non-limiting. [Explanation of symbols]
[0087] 2. Source carrier configuration 4 Collimator body 6 Segments 8 aperture 10 Internal port, linear accelerator 12 Straight side 14a Curved Side 14b Curved Side 20 Patient Positioning Unit 22 Rigid Framework 24 Movable carriage 26 Patient Beds 28 Fixation device 40 Linear accelerators, radiation therapy devices 41 Isocenter 42 patients 43 Patient Table 44 Flat Panel Scintillator Detector 45 Radiating Head 46 Radiation Beam 47 Horizontal axis, gantry 125 Iodine 130 Radiation therapy equipment, radiation therapy units 202 patients 212 Radiation Source 214 Protective Housing 216 Beam Channel 218 Isocenter 220 Coordinate Frame 222 Patient Positioning System 300 ways 310 Steps 320 steps 330 Steps 340 steps 500 Treatment planning computer structure 500 Treatment Planning Computer Structure 510 Modeling Module 520 Low Dose Volume Calculation Module 530 Calculation Module 540 Treatment Planning Module 550 Optimization Module 600 Procedures, Methods 610 Steps 620 steps 620 steps 630 steps 640 steps 650 steps 660 steps 670 steps 680 steps R Low Dose Ring T1 Prescription Dose T2 Prescription Dose T3 Prescription Dose
Claims
1. 1. A method for treatment planning for use in radiation therapy, implemented in a computerized architecture, in which radiation to be delivered is determined using an optimization problem, the optimization problem manipulating the delivered radiation according to an objective reflecting a criterion for a region of interest that includes at least one of at least one target, organ at risk, and / or healthy tissue to be treated during treatment of a patient, the at least one target, organ at risk, and / or healthy tissue being a three-dimensional voxel representation, the method comprising: estimating a set of voxels that will receive a dose above a predetermined threshold dose level, the set of voxels including voxels from at least one target volume; determining a low-dose voxel set, the low-dose voxel set representing a volume surrounding the voxel set receiving the high dose; providing a frame description for the voxels in the low-dose voxel set, the frame description including an objective function for the voxels in the low-dose voxel set, where voxels that receive a dose above the predetermined threshold dose level are penalized such that the dose delivered to the low-dose voxel set is constrained according to the objective function; using the frame description in the optimization problem that manipulates the delivered radiation; A method comprising:
2. 1. A method for treatment planning for use in a radiation therapy system implemented in a computer architecture, the radiation therapy system comprising a radiation therapy unit and capable of varying a delivered spatial dose by adjusting beam shape settings, the delivered radiation being determined using an optimization problem, the optimization problem manipulating the delivered radiation according to an objective reflecting a criterion for a region of interest comprising at least one of at least one target, organ at risk and / or healthy tissue to be treated during treatment of a patient, the at least one target, organ at risk and / or healthy tissue being a three-dimensional voxel representation, the method comprising: estimating a set of voxels that will receive a dose above a predetermined threshold dose level, the set of voxels including voxels from at least one target volume; determining a low-dose voxel set, the low-dose voxel set representing a volume surrounding the voxel set receiving the high dose; providing a frame description for the voxels in the low-dose voxel set, the frame description including an objective function for the voxels in the low-dose voxel set, where voxels that receive a dose above the predetermined threshold dose level are penalized such that the dose delivered to the low-dose voxel set is constrained according to the objective function; using the frame description in the optimization problem that manipulates the delivered radiation; A method comprising:
3. 10. The method of claim 1, wherein the method is for treatment planning in a radiation therapy system, the radiation therapy system comprising a radiation therapy unit having a fixed radiation focal spot.
4. determining a first estimate of the set of low-dose voxels; - performing at least one optimization with either the weighting settings to be used in the final optimization or standardized weighting settings, where the weightings may be different for each voxel; updating said low-dose voxel set, defined as the volume between two dose levels, based on the optimized dose distribution in the previous optimization step; modifying the objective function according to the updated low-dose voxel set; performing a final optimization using the updated low-dose voxel set from the prior optimization and the modified objective function; The method of claim 2 or 3, further comprising:
5. The method of claim 1 further comprising the step of estimating a dose distribution.
6. The method of claim 5 , wherein the dose distribution is estimated by estimating the dose distribution outside each target independently or by estimating a combined dose distribution outside the targets.
7. The method of claim 1 , further comprising determining the low-dose voxel set by applying a predetermined dose distribution.
8. determining a low-dose voxel set by applying a predetermined dose distribution calculating a first shell surface of an outer shell layer or sphere layer of the voxel set based on dose levels above the threshold dose level; calculating a second shell surface of an outer shell layer or sphere layer of the voxel set based on dose levels below the threshold dose level; defining the low-dose voxel set as the set of voxels between the first shell surface and the second shell surface; The method of claim 7, comprising:
9. The step of determining a set of low-dose voxels comprises: calculating a first shell surface outside the set of voxels based on a predetermined geometric shape; calculating a second shell surface outside the set of voxels based on a predetermined geometric shape; defining the low-dose voxel set as the set of voxels between the first shell surface and the second shell surface; The method of claim 1 , comprising:
10. 10. The method of claim 9, wherein calculating the first shell surface outside the set of voxels comprises uniformly shrinking an outer surface of the set of voxels that receives a dose above a predetermined threshold dose level.
11. 11. The method of claim 9 or 10, wherein calculating the second shell surface outside the set of voxels comprises uniformly expanding an outer surface of the set of voxels that receive a dose higher than a predetermined threshold dose level.
12. Finding an isocenter position; determining a shot collimator configuration for all isocenter positions; setting a dose to be delivered to the at least one target, including determining the duration of exposure for every shot to be delivered; calculating a dose distribution based on the determined shot configuration and the dose to be delivered; The method of claim 1 further comprising:
13. The method includes applying a distance model according to: [0010] where Q can be a constant or can include geometric and / or dose distribution factors, and V ad is the volume of the voxel set that receives a dose above a predefined threshold dose level, and δ r The method of claim 1 , wherein x corresponds to a scaled distance of an outer surface of the set of voxels.
14. The method includes applying a distance model according to: [0025] In the formula, D F,ad is the isodose within the set of voxels receiving dose above a given threshold dose level, δD is the given dose distribution variation, and V ad The method of claim 13 , wherein V is a set of voxels that receive a dose above a predetermined threshold dose level.
15. calculating a dose profile for a particular treatment configuration, including a beam shape setting for the radiation dose profile, using the optimization problem; generating a treatment plan including determining a radiation dose profile to be delivered during treatment based on the treatment configuration, each radiation dose profile being modeled by a spatial dose volume distribution of radiation, the shape of the spatial distribution being dependent on the beam shape setting; selecting an optimal treatment plan that meets said criteria; The method of claim 1 further comprising:
16. using said optimization problem to calculate a dose rate for a particular treatment configuration including sector and collimator settings and isocenter exposure time; generating a treatment plan, including determining shots to be delivered during treatment based on the treatment configuration, where each shot is modeled by a spatial dose volume distribution of radiation, the shape of the spatial distribution depending on the particular sector and collimator settings and exposure time; selecting an optimal treatment plan that meets said criteria; The method of claim 1 further comprising:
17. defining a set of beam directions; modeling a radiation dose profile to be delivered to the target as a plurality of beamlets, each beamlet having a beamlet intensity; establishing a number of objectives reflecting criteria for said target; providing an optimization problem that manipulates the delivered radiation in accordance with the objective to create a fluence map, the fluence map defining the beamlet intensities of each of the beamlets; generating a treatment plan based on the fluence map and criteria for the target; selecting an optimal treatment plan that meets said criteria; 20. The method of claim 16, further comprising:
18. The method of claim 1 , wherein determining a low-dose voxel set comprises determining the low-dose voxel set using a trained machine learning model.
19. 1. A treatment planning computer architecture for radiation therapy treatment planning, comprising: a radiation to be delivered determined using an optimization problem, the optimization problem manipulating the delivered radiation according to a frame description reflecting criteria for a region of interest including at least one of at least one target, organ at risk, and / or healthy tissue to be treated during treatment of a patient, the frame description including an objective function for a set of voxels, the at least one target, organ at risk, and / or healthy tissue being three-dimensional voxel representations; 1. A low dose volume calculation module, comprising: estimating a set of voxel volumes that will receive a dose higher than a predetermined boundary dose level, the set of voxel volumes including at least one target volume; determining a low-dose voxel volume set by applying a predetermined altered dose distribution, the low-dose voxel set representing a volume surrounding the set of voxels receiving the high dose; and a low-dose volume calculation module for calculating a low-dose volume by applying an objective function to the voxels in the low-dose voxel volume set, where voxels that receive a dose above a predetermined threshold are penalized such that the dose delivered to the low-dose voxel volume set is suppressed; a computation module configured to generate a radiation dose profile to be delivered to the target and manipulate the delivered radiation according to an objective, providing a convex optimization problem including the frame description for the low dose volume, and to compute a dose profile for a particular treatment configuration using the convex optimization problem; A treatment planning computer architecture comprising:
20. a treatment planning module configured to generate a treatment plan, including determining the radiation dose profile to be delivered during treatment based on the treatment configuration, wherein each radiation dose profile is modeled by a spatial dose volume distribution of radiation, the shape of the spatial distribution being dependent on a beam shape setting; and an optimization module configured to select an optimal treatment plan that meets the criteria; 20. The treatment planning computer architecture of claim 19, further comprising:
21. 20. A treatment planning computer architecture according to claim 19, wherein said low-dose volume calculation module is adapted to carry out the steps according to claim 2 or 3.
