System, method and computer program for determining a radiation therapy treatment plan for a radiation therapy system

Abstract

The present invention relates to a system for determining a radiotherapy plan for a radiotherapy system (100), comprising a multi-leaf collimator. The radiotherapy plan determination system (110) comprises: a treatment system characteristic providing unit (111), wherein the characteristics include possible leaf positions and possible radiation fluence values; a planning target providing unit (112), wherein the planning target indicates a desired therapeutic radiation dose distribution; an optimization function providing unit (113), wherein the optimization function indicates a deviation of the radiation dose distribution from the planning target and an uncertainty of the radiation dose distribution at the edge of the possible aperture; and a treatment plan optimization unit (114), which is adapted to determine a sequence of possible apertures and possible radiation fluence values, for which the optimization function is optimized. Thus, an optimal treatment plan can be provided for each individual patient.

Description

Technical Field

[0001] The present invention relates to a system, method and computer program for determining a radiation treatment plan for a radiation therapy system comprising a multi-leaf collimator (MLC). Background Art

[0002] Nowadays, for radiotherapy, radiation systems including multi-leaf collimators (MLCs) are commonly used. The MLC includes a plurality of movable leaves and allows the radiation beam provided by the radiotherapy system to be adapted to the individual shape and structure of the treatment area (e.g., a tumor that should be irradiated). In order to optimize the radiation dose distribution received by the patient during radiotherapy, a treatment plan optimization algorithm is used to adapt the treatment plan (i.e., the positions of the leaves of the MLC and the sequence of fluence values ​​that the radiotherapy system can provide) to the individual patient in the best possible way. However, in many cases, the same radiation dose distribution can be achieved using different treatment plans, i.e., using different leaf positions and sequences of fluence values. In these cases, the radiologist decides which treatment plan, i.e., which leaf position and sequence of fluence values, should be used for the individual patient based on his / her experience. However, this decision process does not always lead to the optimal treatment plan for the patient. Summary of the Invention

[0003] It is an object of the present invention to provide a system, a method and a computer program which allow an improved determination of a treatment plan such that an optimal treatment plan can be provided for each individual patient.

[0004] In a first aspect of the present invention, a radiotherapy treatment plan determination system is presented, wherein the radiotherapy treatment plan determination system is adapted to determine a radiotherapy treatment plan for a radiotherapy treatment system, the radiotherapy treatment system comprising a multi-leaf collimator (MLC), wherein the MLC comprises a plurality of movable leaves for shaping an aperture of the MLC such that a radiation beam passes through the shaped aperture before being provided to a patient, wherein the radiotherapy treatment plan determination system comprises a) a treatment system characteristic providing unit for providing characteristics of the radiotherapy treatment system, wherein the characteristics comprise possible leaf positions defining possible apertures of the MLC and possible radiation fluence values ​​that can be provided by the radiotherapy treatment system, and b) a planning target providing unit for providing a planning target, wherein the planning target refers to a desired radiation dose. indicating a desired therapeutic radiation dose distribution that should be provided to the patient, c) an optimization function providing unit for providing an optimization function, which indicates a deviation of the radiation dose distribution from a planning target, wherein the radiation dose distribution depends on a sequence of possible apertures and possible radiation fluence values, the sequence of possible apertures being defined by possible leaf positions, and wherein the optimization function further indicates an uncertainty of the radiation dose distribution at edges of the possible apertures, d) a treatment plan optimization unit for determining an optimized treatment plan, wherein the treatment plan optimization unit is adapted to determine a sequence of possible apertures and possible radiation fluence values, for which sequence the optimization function is optimized, wherein the sequence of optimized possible apertures and optimized possible fluence values ​​defines the optimized treatment plan.

[0005] Since the optimization function indicates the uncertainty of the radiation dose distribution at the edges of the possible apertures, the uncertainty of the radiation dose distribution provided by the radiotherapy system, for example due to scattered radiation at the edges of the aperture of the MLC or positioning inaccuracies, can be directly taken into account as part of the optimization process. As a result, a treatment plan can be determined that is not only optimized with respect to the desired dose distribution, but also provides the smallest possible uncertainty in the provided dose distribution. Since uncertainty is a measure of the possible deviation from the optimal treatment plan (i.e., the radiation dose distribution), a low uncertainty in the radiotherapy plan ensures that the radiation dose distribution provided to the patient follows the determined radiotherapy plan very accurately. Thus, treatment plan determination is improved so that a more optimal treatment plan can be provided for each individual patient.

[0006] A radiation therapy planning system is adapted to determine a radiation therapy plan for a radiation therapy system. The radiation therapy system may be any system used for radiation therapy, wherein radiation is to be delivered to a portion of a patient, such as a tumor. The delivered radiation may be any type of ionizing radiation used for medical treatment. Preferably, the radiation therapy system delivers therapeutic X-ray radiation to the patient. Alternatively, the radiation therapy system may be adapted to deliver protons during proton radiation therapy to the patient.

[0007] A radiotherapy system includes an MLC, wherein the MLC provides a plurality of movable blades that can be used to shape an aperture of the MLC. Radiation provided by the radiotherapy system is then shaped by the aperture before being delivered to the patient. In an embodiment, the blades of the MLC are provided as blade pairs that are arranged on either side of an axis in a plane formed by the blades of the MLC. However, other arrangements of the blades of the MLC are also contemplated.

[0008] The treatment system characteristic providing unit is adapted to provide characteristics of the treatment system. The treatment system characteristic providing unit may be a storage unit in which the characteristics of the treatment system are already stored and from which the characteristics can be retrieved, for example based on a list containing characteristics of a plurality of radiotherapy systems. Similarly, the treatment system characteristic providing unit may be a retrieval unit for retrieving characteristics, for example, from a radiotherapy system for which a radiotherapy treatment plan is to be determined, wherein the treatment system characteristic providing unit is then adapted to provide the received characteristics.

[0009] The provided characteristics refer to characteristics of the radiotherapy system that have an impact on the radiation dose distribution provided to the patient. In particular, the characteristics include possible leaf positions that define possible apertures of the MLC and possible radiation fluence values ​​that can be provided by the radiotherapy system. Possible leaf positions refer to positions of leaves of the MLC of the radiotherapy system that can be adopted by the leaves due to the construction of the MLC of the radiotherapy system. For example, possible leaf positions can be provided as position coordinates of each leaf, which position coordinates determine the positions that each leaf can adopt based on the construction of the MLC. Possible radiation fluence values ​​refer to fluence values ​​that can be provided by the radiotherapy system. For example, possible radiation fluence values ​​can refer to one or more x-ray radiation fluence values ​​that the radiotherapy system can provide due to its construction. In an embodiment, the radiotherapy system can be adapted to provide only one radiation fluence value, i.e., only a constant amount of radiation, wherein in this case, the treatment system characteristic providing unit is adapted to provide one possible radiation fluence value as part of the characteristics of the treatment system.

[0010] The planned target providing unit is adapted to provide the planned target. The planned target providing unit may be a storage unit in which the planned target is stored and from which the planned target can be retrieved. Similarly, the planned target providing unit may be a retrieval unit for retrieving the planned target from an input device, e.g., into which a user inputs the planned target, wherein the planned target providing unit is then adapted to provide the received planned target.

[0011] A planning target refers to a target that should be achieved by the radiotherapy of an individual patient. Specifically, a planning target indicates the desired therapeutic radiation dose distribution that should be provided to the patient. For example, a planning target may specify an area in the patient's body that should receive a high radiation dose or an area in the patient's body that should (if possible) only receive a low radiation dose or even an area that should not receive any radiation at all. Planning targets, such as areas that should receive different radiation doses, can be determined by a radiologist based on previously acquired images of the patient's area of ​​interest. For example, a radiologist can determine and select a tumor area that should receive a high radiation dose, a surrounding tissue area that should receive only a low radiation dose if possible, and an area including radiation-sensitive organs that should not receive any radiation at all. These planning targets then indicate the desired therapeutic radiation dose distribution. However, in other embodiments, a user, such as a radiologist, can directly provide the desired radiation dose distribution as a planning target to the planning target providing unit, for example via an input unit.

[0012] The optimization function providing unit is adapted to optimize the optimization function. The optimization function providing unit may be a storage unit in which the optimization function has been stored and from which the optimization function can be retrieved. For example, the storage unit may include different optimization functions, wherein the optimization function providing unit is adapted to select the optimization function, for example based on characteristics of the radiotherapy system. Similarly, the optimization function providing unit may be a retrieval unit for retrieving the optimization function from, for example, an input unit, which a user can use to select the optimization function, wherein the optimization function providing unit is then adapted to provide the received optimization function. The optimization function indicates the deviation of the radiation dose distribution from the plan target. For example, the optimization function may provide the difference between any radiation dose distribution and the plan target, for example by indicating areas where the plan target was not achieved in terms of the radiation dose distribution. In an embodiment, the optimization function may be calculated directly as the difference between the radiation dose distribution and a desired therapeutic radiation dose distribution, which desired therapeutic radiation dose distribution is provided as part of the plan target or is determined based on information provided by the plan target.

[0013] The radiation dose distribution depends on a sequence of possible apertures and possible radiation fluence values, the possible apertures being defined by possible positions. The sequence of possible apertures and possible radiation fluence values ​​can be defined by at least one possible aperture and at least one possible fluence value, but can also refer to a plurality of possible apertures and possible radiation fluence values ​​that should be provided to the patient in chronological order during radiotherapy. Preferably, all possible apertures of the sequence are associated with one possible fluence value. In other embodiments, in the sequence of possible apertures, each possible aperture of the sequence can also be associated with more than one fluence value. Each possible aperture associated with one or more possible radiation fluence values ​​defines a radiation dose distribution, so that the sequence of possible apertures and possible radiation fluence values ​​defines a sequence of partial radiation dose distributions, wherein the sum of all radiation dose distributions defines the radiation dose distribution received by the patient when the radiotherapy system provides radiation according to the sequence.

[0014] The optimization function also indicates the uncertainty of the radiation dose distribution at the edges of the possible apertures. For example, the uncertainty at the edges of the possible apertures may be caused by radiation scatter at these edges or by inaccuracies in the positioning of the blades that define the aperture. The uncertainty in the position of the edges of the possible apertures of the MLC directly leads to the uncertainty of the radiation dose distribution defined by the possible apertures. This uncertainty in the radiation dose distribution can be modeled for one or more edges (preferably all edges) of the possible apertures. The optimization function is then adjusted so that it takes into account these uncertainties in the radiation dose distribution. Preferably, the optimization function depends on these uncertainties.

[0015] The treatment plan optimization unit is adapted to determine an optimized treatment plan, wherein the optimized treatment plan is defined by a sequence of optimized possible apertures and optimized possible fluence values. In order to determine the optimized treatment plan, the treatment plan optimization unit is adapted to determine a sequence of possible apertures and possible radiation fluence values, for which sequence an optimization function is optimized. Specifically, the optimization function is optimized if the deviation between the radiation dose distribution and the plan target becomes very small, for example as small as possible. Based on the mathematical definition of the optimization function, minimization of the deviation can refer to minimization or maximization of the optimization function, i.e. determining a local or global extreme value of the optimization function. For example, the treatment plan optimization unit can be adapted to use an iterative method, such as a gradient descent method, or a direct method, such as direct machine parameter optimization, for finding the optimized treatment plan. Furthermore, the treatment plan optimization unit can be adapted to optimize the optimization function using a two-step optimization method, wherein in a first step, the fluence distribution is optimized, and in a second step, the optimized fluence distribution is converted into a sequence of possible apertures, i.e. possible leaf positions, and radiation fluence values. Alternatively, the treatment plan optimization unit can be adapted to optimize the optimization function in a single step, for example by using a direct machine parameter optimization algorithm, in which the sequence of aperture and radiation fluence values, i.e., the machine parameters, are optimized simultaneously. In this case, a nonlinear optimization method can be used, which can handle, for example, nonlinear constraints, to optimize the optimization function with respect to the objective. This method is preferably used if the radiation therapy is to be provided as a step-and-shoot protocol or a volumetric modulated radiation therapy protocol. Alternatively, the treatment plan optimization unit can also be adapted to use a column generation method.

[0016] Since the optimization function indicates the uncertainty of the radiation dose distribution at the edges of the possible apertures that define the radiation dose distribution, this uncertainty in the radiation dose distribution is also taken into account during the determination of the sequence of possible apertures and possible radiation fluence values ​​for which the optimization function is optimized, i.e. the optimized treatment plan is also optimized with respect to these uncertainties.

[0017] In an embodiment, the uncertainty of the radiation dose distribution is determined based on the possible blade positions of the MLC that define the possible apertures and on the possible radiation fluence values ​​on which the radiation dose distribution depends. Since the possible blade positions define the possible apertures and therefore also define the edges of the possible apertures, the uncertainty of the radiation dose distribution defined by the sequence of possible apertures can be determined very accurately based on the possible blade positions that define the possible apertures. Furthermore, in a preferred embodiment, the radiation dose distribution is determined based on the positions of the tips of the blades that define the possible apertures. For example, if the MLC includes a blade pair arranged along the axis of the MLC, the tip of each blade defines an edge of the aperture parallel to the axis of the MLC, i.e., an edge in the direction of movement of the blades of the MLC. Alternatively or additionally, the uncertainty of the radiation dose distribution can also be determined based on the edges between adjacent blades, wherein the blade positions of adjacent blades define the length of the edges between adjacent blades. Specifically, the edges between adjacent blades of the MLC that includes a blade pair arranged around the axis of the MLC define the edges of the aperture, which edges of the aperture are perpendicular to the axis of the MLC, i.e., perpendicular to the direction of movement of the blades of the MLC. Furthermore, when determining the uncertainty of the radiation dose distribution, it is also preferred to take into account the radiation fluence values ​​associated with the possible apertures. For example, the uncertainty of the radiation dose distribution defined by the possible apertures can be considered to be proportional to the radiation fluence values ​​associated with the possible apertures.

[0018] In a preferred embodiment, the uncertainty of a radiation dose distribution defined by a sequence of possible apertures and possible fluence values ​​is determined based on the uncertainty of each partial radiation dose distribution defined by each possible aperture associated with the possible fluence values ​​that are part of the sequence, the possible aperture defining the radiation dose distribution. For example, the uncertainty can be modeled for each partial radiation dose distribution, and then all uncertainties of all partial radiation dose distributions that affect the radiation dose distribution can be summed.

[0019] In an embodiment, the treatment plan optimization unit is adapted to determine a sequence of possible apertures and possible fluence values, the optimization function being optimized for the sequence such that the sequence of possible apertures and possible fluence values ​​is preferably an optimized treatment plan that results in a radiation dose distribution with a low uncertainty. Preferably, the optimized treatment plan is determined such that the uncertainty of the radiation dose distribution defined by the determined optimized treatment plan is lower than the uncertainty of the radiation dose distribution defined by another treatment plan, wherein the two radiation dose distributions are identical. For example, if during optimization, two treatment plans are found that provide the same optimal dose distribution to the patient but include different sequences of possible apertures and possible fluence values, the optimization function is optimized such that the plan that provides the radiation dose distribution with a low uncertainty is selected as the optimized treatment plan from the two plans. In an embodiment, the optimization function can be optimized such that the uncertainty of the radiation dose distribution of the optimized treatment plan is as small as possible. In an embodiment, a weight is predefined and provided in the optimization function for the uncertainty of the radiation dose distribution such that the treatment plan with a low uncertainty is weighted higher during the optimization process of the optimization function than the treatment plan with a higher uncertainty. Additionally or alternatively, the treatment plan optimization unit may be adapted to determine a sequence of possible apertures and possible fluence values ​​for which the optimization function is optimized such that the sequence of possible apertures and possible fluence values ​​is preferred as an optimized treatment plan that results in a homogeneous uncertainty in the radiation dose distribution. Homogeneous uncertainty refers to an uncertainty distribution in which there is no accumulation of uncertainties at different points. For example, an uncertainty may be considered homogeneous if the variance of the uncertainty is below a predetermined threshold, wherein the variance V may be determined, for example, by using the mathematical term V = ∫(U(x) - avg(U(x))) 2 dx, where U is the uncertainty. Alternatively, if the term ∫U(x) 2 If dx is below a predetermined threshold, the uncertainty can be considered to be homogeneous. The threshold can be determined based on an expected uncertainty distribution or based on general theoretical or experimental data, with respect to which general theoretical or experimental data, the heterogeneity of the uncertainty is acceptable for a specific radiotherapy. In an embodiment, the uncertainty of the radiation dose distribution is modeled based on an uncertainty function centered on at least one edge of each possible aperture, wherein the uncertainty function of the edge includes a width corresponding to the expected uncertainty of the corresponding edge. The uncertainty function can be any function suitable for modeling the uncertainty of the radiation dose distribution caused by the edges of the possible apertures that define the radiation dose distribution. In a preferred embodiment, the uncertainty function refers to a Gaussian function. However, the uncertainty function can also refer to another probability function, such as a corresponding scaled and normalized basis function of a cubic spline or an asymmetric function (such as a gamma distribution).

[0020] The uncertainty function includes a width corresponding to the expected uncertainty of the corresponding edge, with the uncertainty centered on the edge. The expected uncertainty can be determined, for example, based on known construction margins of the MLC, the scattering properties of the MLC blades, calibration measurements performed on the MLC, the positioning accuracy of the MLC blades, etc. To determine the width, for example, a standard body film can be used and subjected to a radiotherapy plan, where the difference between the simulated dose and the measured dose for different blade positions can be determined. Based on these measurements, the width of the uncertainty of the corresponding edge of a given MLC can be determined.

[0021] In a preferred embodiment, the uncertainty function for each edge is weighted with the possible fluence values ​​associated with the possible apertures to which the edge belongs. Weighting the uncertainty function for each edge based on the possible fluence values ​​associated with the edge (i.e., the possible apertures to which the edge belongs) allows taking into account the dependence of the uncertainty on the fluence value associated with the corresponding aperture.

[0022] In an embodiment, the uncertainty of the radiation dose distribution in the direction of motion of the blades of the MLC blade pair is determined using the following formula:

[0023] U=∫(∑ i w i (e i (xx l，i )+e i (x r，i -x))) 2 dx,

[0024] where w i corresponds to the possible fluence values ​​associated with the possible apertures, e i (x) is the uncertainty function that defines the distribution of uncertainty, x l,i and x r,i refers to the possible left and right blade positions for a pair of blades in the MLC, and x runs over the aperture size in the x-direction, which is defined as the direction of motion of the blades of the blade pair, i.e., the direction in which the blade positions can change. The above function involves integration over x; however, the x-direction can be in any direction. The above equation provides the uncertainty for the blade pair in the x-direction, where smaller openings between the blades of the blade pair include higher uncertainty.

[0025] Furthermore, in embodiments, the uncertainty function described above can also be extended to account for edges in other directions, such as edges perpendicular to the x-direction (i.e., along the y-direction). These uncertainties may include characteristics other than the uncertainty in the x-direction, such as due to other structural constraints of the MLC in this direction. Preferably, the uncertainty of the aperture in the y-direction is determined such that an aperture with a rough edge has a higher uncertainty than an aperture with a more rounded opening area.

[0026] It is also considered that the uncertainty of the radiation dose distribution in the y direction can be expressed, for example, as explained below. The uncertainty of the lower left edge of the left blade in row i can be expressed as

[0027]

[0028] where y l,i is the position of the lower edge of the left leaf of a pair of leaves in row i toward row i-1, e i (y) refers to the uncertainty function that defines the distribution of uncertainty in the y direction of the lower edge of the blade, and Y is the uncertainty function e i (y) is a non-zero domain. For other edges of row i, an equivalent formula can be given, namely:

[0029]

[0030]

[0031]

[0032] where for all edges of a leaf pair, the uncertainty function e i The uncertainty in the x- and y-directions can then be determined by summing all occurrences for all leaf pairs, for example weighting the terms by their corresponding fluence values.

[0033] In an embodiment, the optimization function is expressed as follows:

[0034] O×D+λU,

[0035] Where D refers to the deviation and λ refers to a weight used to weight the effect of the uncertainty U during the optimization process of the optimization function O. The weight λ can be determined empirically, for example by measurements, and can depend on the radiotherapy protocol to be used. The weight can be provided to the user, for example by providing it in a user interface on a display, and the user can adjust the weight during the planning process of the treatment plan, for example to achieve a desired compromise between the uncertainty of the treatment plan (i.e., the quality) and the satisfaction of other goals of the treatment plan.

[0036] In an embodiment, a radiotherapy system is configured to deliver a radiation beam from multiple directions, wherein a treatment system characteristic providing unit is adapted to provide possible beam directions of the radiotherapy system as a characteristic of the radiotherapy system, wherein the radiation dose distribution further depends on a sequence of possible radiation directions, and wherein the uncertainty is determined based on an uncertainty determined for each partial radiation dose distribution, the partial radiation dose distribution being defined by possible directions, possible apertures, and possible fluence values ​​that are part of a sequence defining the radiation dose distribution. For example, the radiotherapy system can be adapted to deliver radiotherapy from different directions by moving a radiation beam generating unit, the radiation beam generating unit being adapted to generate radiation beams in different directions around a patient. The possible beam directions can then be defined by possible positions and possible angles of the radiation beam generating unit at which the radiation beam generating unit can deliver the radiation beam to the patient. In this embodiment, the radiotherapy plan is defined by a sequence of possible positions, possible apertures, and possible fluence values. Based on the possible apertures and possible fluence values ​​associated with the possible directions, a partial radiation dose distribution can be determined for each possible direction that is part of the sequence of possible directions defining the radiation dose distribution. Furthermore, for each of these partial radiation dose distributions, an uncertainty may be determined based on the associated possible aperture and the associated fluence value, and the uncertainty of the radiation dose distribution may be determined based on the uncertainty of each partial radiation dose distribution.

[0037] In an embodiment, each radiation direction of the sequence of possible radiation directions is associated with at least one possible aperture and at least one possible fluence value of a radiation dose distribution defining the radiation direction, wherein the uncertainty is determined based on a sum of the uncertainties of each partial radiation dose distribution associated with each possible direction of the sequence of possible directions. Specifically, in this embodiment, the radiation therapy system is adapted to provide radiation therapy as a sequence of multiple discrete beam directions according to a static intensity modulated protocol, wherein each beam direction is associated with one or more aperture and fluence values. In these cases, the uncertainty may be determined based on a sum of the uncertainties of each partial radiation dose distribution associated with each possible direction of the sequence of possible directions defining the radiation dose distribution.

[0038] In another embodiment, the radiotherapy system is adapted to continuously change between possible beam directions, leaf positions and fluence values ​​while generating a radiation beam, wherein the uncertainty of the radiation dose distribution caused by the sequence of continuously changing possible beam directions, leaf positions and fluence values ​​is estimated based on summing uncertainties, which are determined for partial radiation dose distributions, which are determined for multiple directions of the continuously changing sequence of possible directions. The uncertainties determined for the multiple directions can be propagated into the patient before summing. Preferably, the radiotherapy system is adapted to provide radiotherapy according to a volumetric intensity modulated radiation therapy protocol so that the leaf positions and fluence values ​​can also be continuously changed during the rotation of the radiation beam, i.e., during the continuous change of the beam direction. In this case, the radiation dose distribution is obtained by determining partial radiation dose distributions for a set of discrete directions, which are obtained by sampling a sufficiently dense continuous motion (e.g., every four degrees) and then summing these partial radiation dose distributions. The uncertainty of the radiation dose distribution provided by the consecutive sequences according to the VIRT protocol can then be determined accordingly, i.e. it can be estimated based on summing the uncertainties determined for some partial radiation dose distributions provided from multiple beam directions and by summing the uncertainties over all beam directions.

[0039] In an embodiment, the uncertainty is determined by also taking into account the uncertainty caused by the arclet approximation used to determine the radiation dose distribution from a sequence of continuously changing possible beam directions, leaf positions and fluence values. For example, a detailed explanation of the small arc approximation for determining the radiation dose distribution when radiotherapy is provided according to a volumetric modulated radiation therapy protocol can be found in the following articles: S. Webb et al., Physics in Medicine & Biology, Vol. 54, pp. 4345-4360 (2009), “Some considerations concerning volume-modulated arc therapy: a stepping stone towards a general theory”, V. Feygelman et al., Journal of Applied Clinical Medical Physics, Vol. 11, pp. 99-116 (2010), “Initial dosimetric evaluation of SmartArc-a novel VMAT treatment planning module implemented in a multi-vendor delivery chain” and J. Park et al., The British Journal of Radiology, Vol. 88, p. 1049 (2015), “The effect of MLC speed and acceleration on the plan delivery accuracy of VMAT”.

[0040] For situations where the blades must move rapidly during the delivery of radiation therapy, the arcuate approximation can introduce further uncertainty into the radiation dose distribution. Determining the uncertainty of the radiation dose distribution based on the uncertainty caused by the arcuate approximation can lead to further optimization of the radiation therapy plan for volumetric intensity modulated radiation therapy. For example, the uncertainty caused by the arcuate approximation can be calculated based on the difference between the radiation dose distribution calculated using the arcuate approximation using a first angular spacing (e.g., a four-degree spacing) and the radiation dose distribution estimated using the arcuate approximation using a second angular spacing (e.g., a two-degree spacing). In this case, the uncertainty describes the uncertainty of approximating a continuous integral by summing over discrete angular positions, as with the arcuate approximation. One possibility for determining uncertainty from this effect is to construct a set of uncertainty functions that have an adaptable width for each interval between consecutive discrete sampling points (i.e., between different angular positions). The determined uncertainty function can then be associated with the center of the interval spanning from one sampling point to the next. This allows the uncertainty of the arcuate approximation to be determined based on the above-described determination of the uncertainty of the MLC aperture. In this case, the weighting can be based on, for example, a linear combination of the fluence values ​​of adjacent MLC apertures.

[0041] In another aspect of the present invention, a radiotherapy system for providing radiotherapy to a patient is presented, wherein the radiotherapy system comprises: a) a radiation beam generation unit adapted to generate a radiation beam with at least one possible fluence value, wherein the radiation beam generation unit is further adapted to provide the radiation beam to the patient, b) an MLC comprising a plurality of movable leaves movable to a plurality of possible leaf positions for shaping an aperture of the MLC so that the radiation beam is shaped by the aperture before being provided to the patient, c) a radiotherapy plan determination system according to claim 1 for determining an optimized radiotherapy plan comprising a sequence of optimized possible leaf positions and optimized possible fluence values, and d) a radiotherapy control unit for controlling the radiation beam generation unit and the MLC so that the treatment plan is provided to the patient.

[0042] In another aspect of the present invention, a radiotherapy plan determination method is presented, wherein the radiotherapy plan determination method is suitable for determining a radiotherapy plan for a radiotherapy system, the radiotherapy system comprising an MLC, wherein the MLC comprises a plurality of movable leaves for shaping an aperture of the MLC so that the radiation beam is shaped through the aperture before being provided to a patient, wherein the radiotherapy plan determination method comprises a) providing characteristics of the treatment system, wherein the characteristics comprise possible leaf positions defining possible apertures of the MLC and possible radiation fluence values ​​that can be provided by the radiotherapy system, b) providing a planning target, wherein the plan target indicates a desired therapeutic radiation dose distribution that should be provided to the patient, c) providing an optimization function, wherein the optimization function indicates a deviation of the radiation dose distribution from the planning target, wherein the radiation dose distribution depends on a sequence of possible apertures and possible radiation fluence values, the possible aperture being defined by the leaf positions, and wherein the optimization function further indicates an uncertainty of the radiation dose distribution at edges of the possible apertures, and d) determining an optimized treatment plan by determining a sequence of possible apertures and possible radiation fluence values, the optimization function being optimized for the sequence, wherein the sequence of optimized possible apertures and optimized possible fluence values ​​defines the optimized treatment plan.

[0043] In another aspect of the present invention, a computer program for determining a radiation therapy plan for a radiation therapy system including an MLC is presented, wherein the computer program includes program code means for causing the radiation therapy plan determination system to perform the radiation therapy plan determination method when the system executes the computer program.

[0044] It shall be understood that the radiation therapy planning determination system according to claim 1, the radiation therapy system according to claim 13, the radiation therapy planning determination method according to claim 14 and the computer program according to claim 15 have similar and / or identical preferred embodiments, in particular as defined in the dependent claims.

[0045] It shall be understood that a preferred embodiment of the invention can also be any combination of the dependent claims or the above-described embodiments with the corresponding independent claim.

[0046] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In the following figure:

[0048] Figure 1 An embodiment of a radiation therapy system comprising a radiation therapy planning determination system according to the present invention is schematically and exemplarily shown for providing radiation therapy to a patient.

[0049] Figure 2 The basic principle of determining the uncertainty of the radiation dose distribution is shown schematically and exemplarily,

[0050] Figure 3 The principle of determining the uncertainty of the edge of the MLC aperture is schematically and exemplarily shown, and

[0051] Figure 4 A flow chart exemplarily illustrating an embodiment of a method for determining a radiation therapy plan is shown. DETAILED DESCRIPTION

[0052] Figure 1 An embodiment of a radiotherapy system including a radiotherapy planning system according to the present invention is schematically and exemplarily shown. In the following embodiment, the radiotherapy system 100 includes a radiation beam generation unit 101 adapted to generate a radiation beam that is provided to a patient 122 lying on a patient examination couch 121. The radiation beam generation unit 101 is adapted to generate a radiation beam having one possible fluence value. Preferably, the radiation beam generation unit 101 is adapted to generate a radiation beam having a plurality of different fluence values.

[0053] Furthermore, the radiation therapy system 100 includes an MLC 102 comprising a plurality of movable leaves that are movable to a plurality of possible leaf positions for shaping the aperture of the MLC 102. For example, Figure 3 A schematic example of MLC 102 is given in FIG. Figure 3 Multiple blades, such as blade pairs 311 and 312, and an aperture 340 formed by the blades are shown. The MLC 102 is provided by the radiotherapy system 100 so as to be positioned between the radiation beam generating unit 101 and the patient 122 lying on the patient couch 121. The MLC 102 is adapted to shape the radiation beam provided by the radiation beam generating unit 101 according to the aperture formed by the MLC 102 before the radiation beam is provided to the patient 122. Based on the radiation generated by the radiation beam generating unit 101, the MLC 102 is adjusted so that the radiation cannot pass through areas where the blades of the MLC 102 are positioned within the radiation beam. Therefore, the radiation beam can only pass through the aperture of the MLC 102.

[0054] In this embodiment, the radiotherapy system 100 further comprises a radiotherapy control unit 103 adapted to control the radiation beam generation unit 101 and the MLC 102 according to a radiotherapy plan. Specifically, the radiotherapy plan comprises a sequence of possible leaf positions of the MLC 102 and possible radiation fluence values ​​that can be generated by the radiation beam generation unit 101. The radiotherapy control unit 103 is then adapted to provide the radiotherapy plan to the patient 122 by controlling the radiation beam generation unit 101 to generate a radiation beam according to the sequence of fluence values ​​and by controlling the MLC 102 according to the sequence of apertures or leaf positions according to the radiotherapy plan.

[0055] In this embodiment, the radiotherapy system 100 comprises a radiotherapy planning determination system 110. The radiotherapy planning determination system 110 is adapted to provide an optimized radiotherapy plan to the radiotherapy control unit 103, wherein the radiotherapy control unit 103 is then adapted to control the radiation beam generating unit 101 and the MLC 102 according to the optimized radiotherapy plan.

[0056] The radiation therapy plan determination system 110 includes a treatment system characteristic providing unit 111 , a plan target providing unit 112 , an optimization function providing unit 113 , and a treatment plan optimizing unit 114 .

[0057] The treatment system characteristic providing unit 111 is adapted to provide characteristics of the radiotherapy system 100. Specifically, the treatment system characteristic providing unit 111 is adapted to provide possible leaf positions defining possible apertures of the MLC 102 and possible radiation fluence values ​​of the radiation generation unit 101 as characteristics of the radiotherapy system 100. The treatment system characteristic providing unit 111 may provide the possible leaf positions, for example, as possible coordinates of the tips of the leaves of the MLC 102. Furthermore, the treatment system characteristic providing unit 111 may provide the possible radiation fluence values ​​as a list of possible radiation fluence values ​​or a range of possible radiation beam fluence values ​​that may be generated by the radiation beam generation unit 101.

[0058] The treatment system characteristic providing unit 111 may also be adapted to provide additional characteristics of the radiation therapy system 100. For example, if the radiation therapy system 100 is adapted to provide radiation also from different directions, the treatment system characteristic providing unit 111 may be adapted to provide possible beam directions of the radiation therapy system 100 as characteristics.

[0059] In this embodiment, the treatment system characteristics providing unit 111 is adapted to receive the treatment system characteristics from a storage unit, where the radiation therapy system characteristics are stored, for example manually or in a list of characteristics. However, in other embodiments, the treatment system characteristics may also be provided by a user to the input unit, and then received by the treatment system characteristics providing unit 111 from the input unit.

[0060] The planning target providing unit 112 is adapted to provide planning targets for the radiation treatment that should be provided to the patient 122. For example, the planning target providing unit 112 can be adapted to communicate with the display unit 104 and / or the input unit 105 (e.g., a keyboard or target) for receiving planning targets from a user (e.g., a radiologist). The planning targets indicate a desired therapeutic radiation dose distribution that should be provided to the patient 122. For example, the radiologist can identify a tumor region on a computed tomography image of the patient 122 that should receive a predetermined radiation dose during the radiation treatment. Furthermore, the radiologist can identify a region of healthy tissue surrounding the tumor that should receive as little radiation as possible and also provide a radiation dose threshold for this region that should not be exceeded. Additionally, the radiologist can provide regions, including organs that should not receive any radiation at all, such as portions of the aorta or brain. The planning targets provided by the radiologist indicate a desired therapeutic radiation dose distribution that should be provided to the patient 122. However, due to the specific configuration of the radiation therapy system 100, it may not be possible to provide such a completely ideal desired radiation dose distribution to the patient 122, so a treatment plan must be determined to provide a radiation dose distribution that achieves the plan goals as best as possible for the patient 122. In some cases, the radiologist may also provide weights for the plan goals, which indicate the importance of the corresponding plan goals, and the weights may be used to influence the optimization process during the determination of the optimized treatment plan.

[0061] The optimization function providing unit 113 is adapted to provide an optimization function indicating the deviation of the radiation dose distribution from the planned target. In addition, the optimization function also indicates the uncertainty of the radiation dose distribution at the edges of the possible apertures, for example due to scattered radiation or uncertainty in the positions of the blades of the MLC 102. These uncertainties of the radiation dose distribution at the edges of the possible apertures can be calculated based on the possible blade positions of the MLC 102 and the possible radiation fluence values ​​defining the radiation dose distribution. Figure 2 and Figure 3 An example based on the general principle of considering uncertainty at the edges of the MLC 102 is explained.

[0062] Figure 2Two illustrative examples 210 and 220 are shown, wherein different aperture sequences are used to provide radiation dose distributions 211 and 221 in these examples 210 and 220. The illustrated graphs for the two examples 210 and 220 include a y-axis 201 representing the fluence value 211 received by the patient 122 and an x-axis 202 representing the x-position of the two leaves forming the corresponding aperture, the radiation dose distribution of which is shown. The radiation dose distributions 211 and 221 of the two examples 220 and 222 are similar and are formed by a sequence of two apertures (e.g., apertures 212 and 213), wherein each aperture is associated with a fluence value.

[0063] In a first example 210, a radiation dose distribution 211 is achieved through a first aperture 213, which includes a blade position zero for the left blade and a blade position four for the right blade, and a second aperture 212, which includes a blade position zero for the left blade and a blade position one for the right blade. Both apertures 213 and 212 are associated with the same fluence value. For this case, the uncertainty of the radiation dose distribution at the edges of the apertures is exemplarily represented by curves 214, 215, and 216 along the x-axis. At each position in the radiation dose distribution 211 where the edge of a blade in the sequence of apertures 212 and 213 is found in the x-direction, the uncertainty is provided according to uncertainty functions 215 and 216. At position zero, uncertainty function 214 is twice the uncertainty functions 215 and 216 because, at this point, both apertures 212 and 213 of the aperture sequence include edges in the x-direction, such that the uncertainty functions at this position combine, in particular, sum.

[0064] In the second example 220, a radiation dose distribution 221 is implemented by providing a sequence of apertures 222 and 223. Aperture 223 is defined by a position of one for the left blade and a position of four for the right blade of aperture 223. Aperture 222 is defined by a position of zero for the left blade and a position of one for the right blade, and is associated with a fluence value that is twice that of aperture 223. As can be seen by the uncertainties 224, 225, and 226 exemplarily represented on the x-axis 202, the uncertainty of this radiation dose distribution is quite different from the uncertainty of the first radiation dose distribution 211. Specifically, uncertainty 224 is twice the uncertainty 226, even though only one edge, that of aperture 222, is located at position zero in the sequence. However, because the fluence value of aperture 222 is twice the fluence value of aperture 223, the uncertainty associated with aperture 222 is also twice the uncertainty associated with aperture 223. This also results in an uncertainty 225 for which it must be taken into account that position one in the sequence also has two edges, resulting in an uncertainty 225 that is three times higher than the uncertainty 226 .

[0065] As is clear from this schematic example, although both the aperture and fluence value sequences result in the same radiation dose distribution 211, 221, the uncertainty of the radiation dose distribution in the second example 220 (i.e., the second sequence) is much higher than that in the first example 210 (i.e., the first sequence). Based on this principle, it is clear that if the radiation treatment plan according to the first example 210 is selected as the optimized radiation treatment plan, the uncertainty of the radiation dose distribution received by the patient 122 during the provision of the radiation treatment plan can be reduced.

[0066] As an example, the principle can be expressed mathematically as follows. For this mathematical example, considering only the edges of the aperture in the x direction, the possible MLC aperture i can be described by the following formula

[0067] b i (x)=H(xx l，i )·H(x r，j -x),

[0068] where H(x) is the Heavyside step function, and x l,i and x r,i Refers to the possible aperture b i (x) is the possible left and right blade positions of a pair of blades of the MLC 102. In this example, the planning target can be considered in the form of a desired radiation dose distribution f(x). In this case, the function D indicating the deviation of the radiation dose distribution from the planning target can be expressed as:

[0069] D=∫f(x)-∑ i w i b i (x)) 2 dx w i ≥0,

[0070] where w i Refers to the possible aperture b i (x) the associated possible fluence values. The sum of i represents the sum of all partial radiation dose distributions defined by the possible apertures and fluence values ​​that are part of the sequence defining the radiation dose distribution.

[0071] In this example, the uncertainty U can be obtained by using the uncertainty function e for each edge i (x). Then, the uncertainty can be expressed as follows, i.e., modeled:

[0072] U=∫(∑ i w i (e i (xx l，i )+e i (xr，i -x))) 2 dx.

[0073] In a preferred example, the uncertainty function can be selected as a Gaussian function. The optimization function can then be provided as O=D+λU, where λ represents a weight for weighting the effect of the uncertainty U during the optimization of the optimization function O. In this mathematical formula, the optimization function can be minimized to provide an optimized treatment plan.

[0074] During the minimization of this optimization function, the uncertainty at the edges of the aperture is taken into account and leads to a preference for radiotherapy plans, i.e., a sequence of possible apertures and possible fluence values, including as little uncertainty as possible. In particular, returning Figure 2 , in which case optimizing such an optimization function as provided above would result in a radiation therapy plan that complies with the first example 210 but not the second example 220.

[0075] Figure 3 An MLC 102 is shown as an example, including blade pairs arranged along the axis of the MLC 102, such as the blade pair formed by blades 312 and 311, which form an aperture 340. In this example, the uncertainty calculated according to the above formula refers to the edge area in the x-direction, represented by the number 320. However, the formula given above can also be extended to consider the edge area 330 between the two apertures in the y-direction. The uncertainty in this case can be expressed in terms of the uncertainty of the edge in the x-direction.

[0076] The treatment plan optimization unit 114 is then adapted to optimize the optimization function, i.e., determine a sequence of possible apertures and possible radiation fluence values ​​for which the optimization function is optimized. The treatment plan optimization unit 114 may optimize the optimization function according to a known optimization algorithm (e.g., an iterative optimization algorithm). By taking into account the uncertainty of the radiation dose distribution at the edges of the possible apertures, the optimized treatment plan determined by the treatment plan optimization unit 114 will include the smallest possible uncertainty in the radiation dose distribution received by the patient 122.

[0077] Figure 4A flow chart illustrating an embodiment of a method for determining a radiation therapy treatment plan is shown. The radiation therapy treatment plan determination method 400 includes a first step 410 of providing characteristics of the radiation therapy system 100. The characteristics include possible leaf positions defining possible apertures of the MLC 102 and possible radiation fluence values ​​that can be provided by the radiation therapy system 100 (e.g., the radiation beam generation unit 101). In a second step 420, a planning target is provided, wherein the planning target indicates a desired therapeutic radiation dose distribution that should be provided to the patient 122. In a step 430, an optimization function is provided. The optimization function can be provided according to the principles described above and indicates deviations of the radiation dose distribution from the planning target and uncertainties in the radiation dose distribution at the edges of the possible apertures. In a final step 440, an optimized treatment plan is determined by determining a sequence of possible apertures and possible radiation fluence values ​​for which the optimization function is optimized. The sequence of optimized possible apertures and optimized possible fluence values ​​then defines an optimized treatment plan that can be provided to the patient 122 using the radiation therapy system 100.

[0078] Although in the above-described embodiment, the radiation beam is provided to the patient 122 from only one direction, in other embodiments, the radiation therapy system 100 may be adapted to provide the radiation beam to the patient 122 from different directions. For example, the radiation therapy system 100 may be adapted to provide radiation therapy according to a static intensity modulation protocol, wherein a sequence of radiation beam directions with associated possible apertures and associated possible radiation fluence values ​​is provided to the patient 122 as a radiation therapy plan. In this case, an optimization function may be determined, for example, based on the exemplary function above, as where j refers to the different beam directions of a sequence, i.e., a radiotherapy treatment plan according to a static IMRT protocol. Thus, the uncertainty of a radiation dose distribution defined by a sequence of possible beam directions, apertures, and fluence values ​​can be determined as the sum of the uncertainties of the partial radiation dose distributions defined by the aperture and fluence values ​​associated with the possible beam directions of the sequence.

[0079] Furthermore, the radiotherapy system 100 may be adapted to provide radiotherapy in the form of volumetric intensity modulated radiotherapy, wherein in this form of treatment the radiotherapy plan is provided as a continuous sequence of beam directions and associated possible apertures and possible radiation values, i.e. radiation is continuously delivered to the patient 122 during movement of the radiation beam generating unit 101 and the MLC 102 around the patient 122, and thus also during changes from one aperture of the sequence and one fluence value of the sequence to another aperture of the sequence or another fluence value of the sequence. In this case, one possibility for calculating the uncertainty is to calculate the uncertainty from the fluence plane (i.e., the fluence plane representing the Figure 3The uncertainty is propagated into the patient's body (the two-dimensional space of the MLC shown in ), and by integrating the uncertainty over the patient's body, the uncertainty of the different beam directions of the sequence is accumulated. The resulting uncertainty can then be added to the deviation D to provide the optimization function O in the same manner as described above.

[0080] In this embodiment, when determining uncertainties, additional uncertainties caused by the approximation used to determine the radiation dose distribution of the volumetric modulated radiation therapy protocol can be taken into account. For example, these uncertainties can be caused by the movement of the blades, and in particular, due to the construction limitations of the MLC, the exact movement and speed of the rapidly moving blades include some uncertainty. These uncertainties can be determined, for example, in the case of the small arc approximation, by calculating a first radiation dose distribution using a first angular spacing in the small arc approximation, and then calculating a second radiation dose distribution by assuming a second angular spacing in the small arc approximation and determining the difference between the first radiation dose distribution and the second radiation dose distribution. This difference indicates the uncertainty introduced by using the small arc approximation to calculate the radiation dose distribution. For example, the total radiation dose distribution can be represented by two slightly different approximations:

[0081]

[0082] where d 2i (where i = 0, ... n) refers to the partial radiation dose distribution determined for the first angular interval, d j (where j = 0, ... 2n) refers to the partial radiation dose distribution of the same radiotherapy plan determined using a second angular spacing that is finer than the first angular spacing, and w 2i and w j refers to the corresponding fluence value. The difference between these approximations of the total radiation dose distribution can be used as a measure of the uncertainty due to the finite approximation. Taking the square of the difference and multiplying it by some appropriate scaling factor can be used as a three-dimensional uncertainty function in the patient domain, which, as described above, can be used to determine the uncertainty by integrating over the patient domain.

[0083] The MLC, and in particular the position of its blades, is often only approximately described, often significantly affecting the simulated radiation dose distribution for radiotherapy plan optimization. In particular, blade tips can contribute to uncertainty, for example due to scattered radiation and positioning inaccuracies. This effect is amplified if two MLC apertures are adjacent, such as in a sequence defining the apertures of a radiotherapy plan. This uncertainty effect can be reduced by preferring a sequence consisting of one large MLC aperture shape over a sequence consisting of two smaller MLC aperture shapes that would theoretically produce the same radiation dose distribution.

[0084] Current radiotherapy planning optimization algorithms do not account for the uncertainty in the interpretation of radiation dose distributions described above, but instead use empirical rules that, for example, favor plans with fewer monitoring units to generate high-quality plans. This makes it difficult to design algorithms that generate radiotherapy plans that define sequences of MLC aperture shapes that meet the extensive empirical expectations of clinicians and dosimetrists for the accuracy of good radiotherapy plans.

[0085] The basic idea of ​​the present invention is to use the uncertainty of the radiation dose distribution as an additional criterion during the optimization of a radiotherapy plan, in addition to the simulated radiation dose distribution, i.e., the radiation dose distribution that should be optimized according to the planning objectives. For each blade tip of an MLC, the uncertainty can be modeled by a Gaussian-type function that is weighted with the fluence value associated with the corresponding MLC aperture centered at the blade tip and with a width corresponding to the expected uncertainty. In order to use the uncertainty, the objective function to be minimized is penalized by adding a measure of the uncertainty. For example, when the fluence map is optimized by discrete MLC apertures during the leaf sequencing process to approximate a continuous radiation dose distribution, the result is penalized with the weighted mean squared dose uncertainty, i.e., the sequence including the MLC aperture is preferred, resulting in a smaller uncertainty or a more uniform uncertainty distribution. Similarly, in the direct machine parameter optimization of a treatment plan leading to the optimization of a static intensity modulation protocol, a weighted measure of the uncertainty in the fluence plane can be added to the objective function to be minimized, i.e., the optimization function. For volumetric intensity modulated radiation therapy, dose uncertainty can be projected into a volume representing the patient's body and accumulated, for example similar to the dose itself, where the objective function, i.e., the optimization function, is penalized by a weighted measure of the uncertainty accumulated over the body volume and / or target and organs at risk.

[0086] For a static intensity modulated protocol, for example, a tumor is irradiated with different radiation beam directions, wherein each beam direction can be associated with a plurality of MLC apertures with associated fluence values. In one embodiment, a radiation dose distribution in a patient's body can be determined given machine parameters (i.e., possible leaf positions of possible MLC apertures and possible fluence values ​​associated with the MLC apertures) and planning objectives (e.g., minimum radiation dose in a tumor or maximum radiation dose in an organ at risk) using a direct machine parameter optimization (DMPO) optimization function, wherein an objective function, i.e., an optimization function, is minimized according to the machine parameters. The objective function can quantify the deviation of the planned target of the desired radiation dose distribution from the simulated radiation dose distribution. Uncertainty can be taken into account by minimizing the optimization function, which is also defined by the sum of the uncertainties of all beam directions and includes weights for weighting the uncertainties.

[0087] In volumetric intensity modulated radiation therapy (VMAT), a linear accelerator, i.e., the radiation beam generating unit, rotates around the patient while the leaves in the collimator move and continuously delivers radiation to a target, such as a tumor. In this case, the radiation dose distribution uncertainty can be calculated in a different way than in static intensity modulated protocols, in which radiation is delivered from the same direction through multiple MLC apertures. In this case, the uncertainty can be propagated from the fluence plane into the patient's body and accumulated in different beam directions, just as the dose inside the patient's body can be calculated in this case. A measure of uncertainty can then be calculated by integrating the uncertainty of the irradiated part of the patient's body, the target region, and / or the region corresponding to the organ at risk. Likewise, the resulting uncertainty can then be added to the objective function to account for the uncertainty of the radiation treatment plan generated by the minimization.

[0088] In the case of VMAT, additional uncertainty in the radiation dose distribution can be taken into account, which uncertainty is caused by, for example, the small arc approximation. For example, in the case of rapidly moving leaves, such an approximation can lead to inaccuracies in the process of estimating the radiation dose distribution. For example, the difference between the dose calculation for a four-degree angular spacing and a two-degree angular spacing can be used to quantify the uncertainty of the irradiated part of the patient's body, the target area and / or the area corresponding to the risk organ and used as a measure of uncertainty. Typically, information about the uncertainty can be displayed by a user interface (i.e., a display) to support dosimetrists, physicists and clinicians in plan generation and quality assurance.

[0089] Although in the above-described embodiments the radiation therapy planning determination system is part of the radiation therapy system, in other embodiments the radiation therapy planning determination system may be a stand-alone system or a system connected to a plurality of different radiation therapy systems.

[0090] Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

[0091] In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality.

[0092] A single unit or device may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0093] Programs (e.g., providing radiation therapy system characteristics, planning goals, and objective functions) or programs (e.g., determining an optimized treatment plan to be performed by one or more units or devices) may be executed by any number of units or devices. For example, these programs may be executed by a single device. These programs may be implemented as program code means of a computer program and / or dedicated hardware.

[0094] The computer program may be stored / distributed on suitable media such as optical storage media or solid-state media provided together with or as part of other hardware, but may also be distributed in other forms such as via the Internet or other wireless communication systems.

[0095] Any reference signs in the claims should not be construed as limiting the scope.

[0096] The present invention relates to a system for determining a radiation treatment plan for a radiation therapy system, comprising a multi-leaf collimator. The radiation treatment plan determination system comprises: a treatment system characteristic providing unit, wherein the characteristics include possible leaf positions and possible radiation fluence values; a planning target providing unit, wherein the planning target indicates a desired therapeutic radiation dose distribution; an optimization function providing unit, wherein the optimization function indicates deviations of the radiation dose distribution from the planning target and uncertainties of the radiation dose distribution at edges of possible apertures; and a treatment plan optimization unit adapted to determine a sequence of possible apertures and possible radiation fluence values, for which the optimization function is optimized. Thus, an optimal treatment plan can be provided for each individual patient.

Claims

1. A radiation therapy plan determination system for determining a radiation therapy plan for a radiation therapy system (100), the radiation therapy system (100) comprising a multi-leaf collimator (MLC) (102), wherein the MLC (102) comprises a plurality of movable leaves for shaping an aperture of the MLC (102) such that a radiation beam passes through the aperture before being provided to a patient (122), wherein the radiation therapy plan determination system (110) comprises: a treatment system characteristics providing unit (111) for providing characteristics of the radiotherapy system (100), wherein the characteristics include possible leaf positions defining possible apertures of the MLC (102) and possible radiation fluence values ​​that can be provided by the radiotherapy system (100), - a planning target providing unit (112) for providing a planning target, wherein the planning target indicates a desired therapeutic radiation dose distribution that should be provided to the patient (122), an optimization function providing unit (113) for providing an optimization function indicating a deviation of a radiation dose distribution from the planning target, wherein the radiation dose distribution depends on a sequence of possible apertures and possible radiation fluence values, the possible apertures being defined by the possible leaf positions, and wherein the optimization function further indicates an uncertainty of the radiation dose distribution at edges of the possible apertures, - a treatment plan optimization unit (114) for determining an optimized treatment plan, wherein the treatment plan optimization unit (114) is adapted to determine a sequence of possible apertures and possible radiation fluence values ​​for which the optimization function is optimized, wherein the sequence of optimized possible apertures and optimized possible fluence values ​​defines the optimized treatment plan.

2. The radiotherapy planning determination system according to claim 1, wherein the uncertainty of the radiation dose distribution is determined based on the possible leaf positions of the MLC (102) that define the possible apertures and based on the possible radiation fluence values ​​on which the radiation dose distribution depends.

3. The radiation therapy plan determination system according to claim 2, wherein the treatment plan optimization unit (114) is adapted to determine the optimized treatment plan so that the uncertainty of a first radiation dose distribution defined by the optimized treatment plan is smaller than the uncertainty of a second radiation dose distribution defined by another treatment plan, wherein The first radiation dose distribution is the same as the second radiation dose distribution.

4. The radiation therapy planning determination system of claim 1 , wherein the uncertainty for the radiation dose distribution is modeled based on an uncertainty function centered about at least one edge of each possible aperture, wherein the uncertainty function for an edge includes a width corresponding to an expected uncertainty of the corresponding edge. 5 . The radiation therapy planning determination system of claim 4 , wherein the uncertainty function of each edge is weighted using possible fluence values ​​associated with the possible apertures to which the edge belongs.

6. The radiation therapy planning determination system according to claim 1, wherein the uncertainty of the radiation dose distribution in the movement direction of the blade of the blade pair of the MLC is determined using the following formula: U=∫(∑ i w i (e i (x-x l，i )+e i (x r，i -x))) 2 dx, in, w i Corresponding to the possible fluence values ​​associated with the possible apertures, e i (x) refers to the uncertainty function of the distribution defining the uncertainty, and x l，i and x r，i refers to the possible left and right blade positions of a pair of blades of the MLC (102), and x runs over the aperture size in the x-direction, which is defined as the direction of motion of the blades of the blade pair. 7 . The radiation therapy planning determination system according to claim 4 , wherein the uncertainty function is a Gaussian function.

8. The radiation therapy planning determination system according to claim 7, wherein the optimization function is defined as: O=D+λU, where D refers to the deviation and λ is a weight used to weight the influence of the uncertainty U during the optimization of the optimization function O.

9. A radiotherapy planning determination system according to any one of claims 1 to 6, wherein the radiotherapy system (100) is configured to provide the radiation beam from a plurality of directions, wherein the treatment system characteristic providing unit (111) is suitable for providing possible beam directions of the radiotherapy system (100) as characteristics of the radiotherapy system (100), wherein the radiation dose distribution also depends on a sequence of possible radiation directions, and wherein the uncertainty is determined based on the uncertainty determined for each partial radiation dose distribution, the partial radiation dose distribution being defined by the possible directions, possible apertures and possible fluence values ​​being part of the sequence defining the radiation dose distribution.

10. The radiation therapy planning determination system of claim 9 , wherein each radiation direction in a sequence of possible radiation directions is associated with at least one possible aperture and at least one possible fluence value defining a radiation dose distribution for the radiation direction, wherein the uncertainty is determined based on a sum of the uncertainties of each partial radiation dose distribution associated with each possible direction in the sequence of possible directions.

11. A radiotherapy planning determination system according to any one of claims 1 to 6, wherein the radiotherapy system (100) is adapted to continuously change between possible beam directions, leaf positions and fluence values ​​while generating the radiation beam, wherein the uncertainty for the radiation dose distribution caused by the sequence of continuously changing possible beam directions, leaf positions and fluence values ​​is estimated based on summing uncertainties determined for partial radiation dose distributions, which are determined for multiple directions in the sequence of continuously changing possible directions.

12. The radiation therapy planning determination system of claim 11, wherein the uncertainty is determined further taking into account uncertainty caused by a small arc approximation used to determine the radiation dose distribution from the sequence of continuously changing possible beam directions, leaf positions, and fluence values.

13. A radiation therapy system for providing radiation therapy to a patient (122), wherein the radiation therapy system (100) comprises: a radiation beam generating unit (101) adapted to generate a radiation beam having at least one possible fluence value, wherein the radiation beam generating unit (101) is further adapted to provide the radiation beam to a patient (122), - an MLC (102) comprising a plurality of movable blades movable to a plurality of possible blade positions for shaping an aperture of the MLC (102) such that the radiation beam is shaped by the aperture before being provided to the patient (122), - The radiation therapy plan determination system (110) according to claim 1, configured to determine an optimized radiation therapy plan, the optimized radiation therapy plan comprising a sequence of optimized possible leaf positions and optimized possible fluence values, and - a radiotherapy control unit (103) for controlling the radiation beam generating unit (101) and the MLC (102) so that the treatment plan is provided to the patient (122).

14. A method for determining a radiation therapy plan for a radiation therapy system (100), the radiation therapy system (100) comprising an MLC (102), wherein the MLC (102) comprises a plurality of movable leaves, the plurality of movable leaves being configured to shape an aperture of the MLC (102) such that a radiation beam is shaped by the aperture before being provided to a patient (122), wherein the method comprises: - providing characteristics of the treatment system, wherein the characteristics include possible leaf positions defining possible apertures of the MLC (102) and possible radiation fluence values ​​that can be provided by the radiotherapy system (100), - providing a planning target, wherein the planning target indicates a desired therapeutic radiation dose distribution that should be provided to the patient (122), - providing an optimization function indicating a deviation of a radiation dose distribution from the planning target, wherein the radiation dose distribution depends on a sequence of possible apertures and possible radiation fluence values, the possible apertures being defined by the leaf positions, and wherein the optimization function further indicates an uncertainty of the radiation dose distribution at edges of the possible apertures, - determining an optimized treatment plan by determining a sequence of possible apertures and possible radiation fluence values ​​for which the optimization function is optimized, wherein the sequence of optimized possible apertures and optimized possible fluence values ​​defines the optimized treatment plan.

15. A computer program for determining a radiotherapy plan for a radiotherapy system (100), the radiotherapy system (100) comprising an MLC (102), wherein the computer program comprises program code means for causing the radiotherapy plan determination system (110) according to claim 1 to perform the steps of the radiotherapy plan determination method according to claim 14 when the computer program is executed by the system.

Images