JP6588158B2 - Dose error distribution calculation device and particle beam therapy device equipped with dose error distribution calculation device - Google Patents

Description

  The present invention relates to a particle beam treatment for irradiating a diseased part such as a tumor with a particle beam (particle beam) of protons, carbon ions, etc., and treating the affected part with a predetermined dose according to the three-dimensional shape of the affected part. The present invention relates to a dose distribution calculation device that calculates a dose distribution by distribution measurement and a particle beam therapy system that includes the dose distribution calculation device.

  Particle beam therapy uses a device such as an accelerator to accelerate charged particles such as protons or carbon ions to several hundred mega-electron volts, and irradiates the patient to give a dose to the tumor in the body to treat cancer. It is a method to do. At this time, it is important to form a dose distribution instructed by the doctor, that is, a dose distribution as close as possible to the target distribution for the tumor. In many cases, the target distribution is such that the dose is as uniform as possible within the tumor and that the dose is as low as possible outside the tumor than in the tumor. However, this is not always the case. For example, a target distribution in which the dose is not uniform within the tumor may be set in preference to lowering the dose outside the tumor. In addition, IMPT (Intensity Modulated Particle Therapy) which makes the total applied dose more ideal distribution by combining the particle beam irradiation from multiple angles and optimizing the dose and dose distribution for each angle. In general, the dose distribution from a single angle is generally not uniform within the tumor.

  In general, when an object (including a human body) is irradiated with a particle beam accelerated by an accelerator, the three-dimensional dose distribution in the object has a characteristic that the dose has a maximum peak at one point. This maximum dose peak is called the Bragg peak. In addition, when there is a maximum dose peak at one point in the three-dimensional space, the peak position is defined as the “irradiation position” of the particle beam. In order to form the target distribution three-dimensionally using the particle beam having the peak structure as described above, some device is required.

  One of the methods for forming the target distribution is a scanning irradiation method. In order to use this method, first, a mechanism for arbitrarily deflecting the particle beam in two directions perpendicular to the Z direction, that is, the traveling direction of the particle beam, that is, the X and Y directions, using an electromagnet or the like is used. . Furthermore, the function which adjusts arbitrarily the position where a Bragg peak is formed in a Z direction by adjustment of particle energy is required. In general, a particle beam generating and transporting apparatus that transports and blocks a particle beam includes an accelerator that accelerates the particle beam, and the accelerator also has an energy adjustment function. A plurality of irradiation positions (also referred to as spots) are set in the tumor, and the particle beam is sequentially irradiated to each irradiation position using the above two mechanisms. The balance of the dose to be applied to each irradiation position is adjusted and determined in advance, and the respective dose distributions applied to the irradiation positions are added together, thereby forming a target distribution as a result.

  In the scanning irradiation method, since there are various uncertainties in the actual irradiation, the target dose distribution may not be the target distribution even though the target distribution should be obtained in the calculation. As uncertain factors, there are, for example, the time change of the particle beam amount, the time change and hysteresis of the magnetic field of the scanning electromagnet, the sensitivity variation of the dose monitor, the signal delay and noise of the control device, and the like. These effects may cause the actual dose distribution to differ from the calculated value.

  In order to eliminate the uncertainties, after planning the particle beam therapy and before actually irradiating the patient with the beam, beam irradiation is performed on the phantom (patient substitute) under the same conditions as possible. In general, an operation of measuring an absolute dose value (absolute dose value) and a dose distribution and confirming whether or not it matches a plan is performed (for example, see Non-Patent Document 1). This operation is called patient QA (Quality Assurance). In general, water in a water tank is often used for a phantom, and a dose is measured using a dosimetry device installed in the water. Considering the purpose of the patient QA, it is desirable to confirm not only the absolute dose value at the center of the tumor but also the dose distribution around it, and for that purpose, it is desirable to measure the dose at a plurality of measurement points.

T. Inaniwa, et al., “Development of treatment planning for scanning irradiation at HIMAC”, Nuclear Instruments and Methods in Physics Research B 266 (2008) 2194-2198

  When measuring a dose distribution in particle beam scanning irradiation, when a dosimeter using an ordinary ionization chamber is used, measurement can be performed only at one point for one scanning irradiation. Therefore, when measuring doses at a plurality of points, there has been a problem that scanning irradiation must be performed as many times as the number of measurement points, which takes time. In general particle beam treatment facilities, an upper limit of the amount of beam irradiation that can be irradiated in one week is set. Since it is necessary to perform the patient QA for each patient, there is a problem that the number of patients that can be accepted by the treatment facility, that is, the number of patients that can be treated decreases when the time and the number of irradiations for the patient QA increase.

  As a simple method for solving the above problem, a method of measuring multiple doses at a time can be considered. For example, it is possible to measure a dose distribution in a two-dimensional plane at a time by using a radiation sensitive film. However, this method has problems such as variations in lots in film production and dependency of dose and film sensitivity on the quality of radiation, and generally has a problem that measurement accuracy is lower than that of an ionization chamber. In addition, as another method for measuring multi-point doses at a time, a large number of small ionization chambers can be arranged. However, this method has a problem that it is difficult to make the arrangement interval of the ionization chambers smaller than about 1 cm, and it is difficult to work accuracy and wiring. In addition, there is a possibility that the particle beam scattered by hitting the electrode of the ionization chamber may affect the measurement value of the other ionization chamber, and the measurement accuracy is lowered.

  The present invention has been made in order to solve the above-described problems, and can reduce the measurement time without reducing the measurement accuracy of the dose distribution in the patient QA subjected to particle beam scanning irradiation. An object of the present invention is to provide a particle beam therapy system including a distribution calculation device and a dose distribution calculation device.

The dose error distribution calculation device according to the present invention includes a treatment plan information storage unit for storing treatment plan information, an error information storage unit for storing error information, and the treatment plan information stored in the treatment plan information storage unit. A dose error distribution calculation unit that calculates a dose error distribution according to the error information stored in the error information storage unit for irradiation based on the information, and the treatment plan information is determined by the treatment plan The position of the generated spot and the energy and beam amount of the particle beam that irradiates each spot, and the error information includes an operation error of the apparatus, the position of the spot that can be generated by the movement of the irradiation target, the energy, The dose error distribution calculation unit includes at least one error of the beam amount, and the dose error distribution calculation unit is based on a tendency of the spot position, the energy, and the beam amount error. Can, create an error scenario is a combination of values of said error, and calculating the corresponding dose distribution to the error scenario, the variation of each of the error scenario characterized in that calculated as distribution of the dose error is there.

  According to this invention, for the irradiation based on the treatment plan information, by calculating the dose error distribution according to the error information, the measurement time can be shortened without reducing the measurement accuracy of the dose distribution. Can be planned.

It is a schematic block diagram of the particle beam therapy system provided with the dose distribution calculating apparatus in Embodiment 1 of this invention. It is a figure which is a block diagram of the dose distribution calculating apparatus in Embodiment 1 of this invention. It is a figure explaining the example of the total dose distribution measured by the dose distribution calculating apparatus in Embodiment 1 of this invention, and a dose dose evaluation point. It is a display example which displays the dose error distribution calculated by the dose distribution calculation apparatus in Embodiment 1 of this invention. It is a display example which displays the dose error distribution calculated by the dose distribution calculation apparatus in Embodiment 1 of this invention. It is a display example which displays the dose error distribution calculated by the dose distribution calculation apparatus in Embodiment 1 of this invention. It is a figure explaining the concept of the error scenario used with the dose distribution calculating apparatus in Embodiment 1 of this invention. It is a flowchart of the dose error distribution calculation using the dose distribution calculation apparatus in Embodiment 1 of this invention. It is an image figure of the error scenario corresponding to each step of the dose error distribution calculation flow using the dose distribution calculation apparatus in Embodiment 1 of this invention. It is an image figure of the error scenario corresponding to each step of the dose error distribution calculation flow using the dose distribution calculation apparatus in Embodiment 1 of this invention. It is an image figure of the error scenario corresponding to each step of the dose error distribution calculation flow using the dose distribution calculation apparatus in Embodiment 1 of this invention. It is an image figure of the error scenario corresponding to each step of the dose error distribution calculation flow using the dose distribution calculation apparatus in Embodiment 1 of this invention. It is an image figure of the error scenario corresponding to each step of the dose error distribution calculation flow using the dose distribution calculation apparatus in Embodiment 1 of this invention. It is a display example which displays the dose error distribution calculated by the dose distribution calculation apparatus in Embodiment 4 of this invention. It is a figure explaining the average autocorrelation function calculated with the dose distribution calculating apparatus in Embodiment 5 of this invention.

Embodiment 1 FIG.
FIG. 1 is a schematic configuration diagram of a particle beam therapy system 100 including a dose distribution calculation device 1 according to Embodiment 1 of the present invention. As shown in FIG. 1, a particle beam treatment apparatus 100 includes a particle beam generation apparatus 10 that generates a particle beam 2 having energy necessary for treatment, a particle beam irradiation apparatus 30 provided with a dose distribution calculation apparatus 1, and a particle The beam transport device 20 transports the particle beam 2 from the beam generator 1 to the particle beam irradiation device 30.

  The particle beam generator 10 includes a control unit (not shown) that generates a particle beam 2 having energy necessary for treatment and controls the start and stop of extraction of the particle beam 2. The particle beam irradiation apparatus 30 deflects the particle beam 2 in two directions perpendicular to the z direction which is the beam traveling direction, that is, the x direction and the y direction, and can scan by scanning irradiation of the particle beam 2 at the patient position. A scanning device 3 and a controller (not shown) that controls scanning of the particle beam 2 by the scanning device 3 are provided. In addition, the particle beam irradiation device 30 includes a dose measurement device 7 that measures a dose value that the particle beam 2 scanned by the scanning device 3 is irradiated to each irradiation position of the treatment target (patient), the x-direction scanning electromagnet 4, A position monitor (not shown) for detecting beam information for calculating the passing position (center of gravity position) and size of the beam through which the particle beam 2 scanned by the y-direction scanning electromagnet 5 passes is provided.

  The dose distribution calculation device 1 includes a treatment plan information storage unit 11, an error information storage unit 12, and a dose error distribution calculation unit 13. The treatment plan information storage unit 11 stores treatment plan information including the number and position of spots for scanning irradiation determined by the treatment plan in particle beam treatment, and the energy and beam amount of the particle beam irradiated to each spot. . The error information storage unit 12 stores information on at least one error among the operation error of the device included in the particle beam therapy system or the spot position, energy, and beam irradiation amount that may be generated by the movement of the irradiation target. The dose error distribution calculation unit 13 calculates a dose error distribution for irradiation based on the treatment plan information in consideration of the influence of the error on the dose distribution according to the error information.

Next, the total dose to be given to the tumor volume (tumor region) by scanning irradiation in the treatment plan will be described. In scanning irradiation, by providing a plurality of spots in a tumor volume (tumor region) and irradiating each spot with an appropriate amount of the particle beam 2, for example, as shown in FIG. Form. The spot number is j, the dose evaluation point number in the phantom 9 is i, and the dose given to the i-th dose evaluation point pi when the j-th spot is irradiated with a unit beam amount is di _{, j} . , Where _j is the beam amount to be given to the j-th spot and n is the total number of spots, the total dose D _i given to the i-th dose evaluation point pi when irradiation is completed for all spots is: It can be expressed as equation (1).

A step of calculating the beam amount w _j to be applied to the optimum spot before irradiation is necessary so that the total dose D _{i at} each dose evaluation point pi is as close as possible to the target dose distribution. This process is called a treatment plan. As appropriate, the beam amount _{w j} is referred to as a spot beam weight _{w j.}

FIG. 3 is an example in which the number and position of spots and the spot beam amount w _j to be determined in the treatment plan are determined. The vertical axis in FIG. 3 is the dose, and the horizontal axis is the position in the z direction. FIG. 3 shows a one-dimensional example of the spot arrangement and the dose distribution in the z-axis direction (beam traveling direction) for simplicity. FIG. 3 shows four spots sp1, sp2, sp3, sp4 and 13 dose evaluation points p1, p2, p3, p4, p5, p6, p7, p8, p9, p10, p11, p12, p13. Indicated. The dose distribution 26 is a dose distribution according to the amount of the spot beam irradiated to the spot sp1. Similarly, the dose distributions 27, 28, and 29 are dose distributions depending on the amount of spot beams irradiated to the spots sp2, sp3, and sp4, respectively. The total dose distribution 25 is a dose distribution obtained by adding the dose distributions 26, 27, 28, and 29. There are 8 dose evaluation points in the tumor, which are dose evaluation points p3 to p10. There are five dose evaluation points outside the tumor, which are dose evaluation points p1, p2, and p11 to p13.

As shown in FIG. 3, the total dose distribution 25 can be high inside the tumor and low outside the tumor by appropriately determining the beam amount w _j given to the spots sp1 to sp4. In FIG. 3, the number of spots is 4 and the dose evaluation points are 13. However, more spots and dose evaluation points are generally arranged at short intervals according to the tumor size. Further, in FIG. 3, for simplicity, the spot arrangement and the dose distribution are displayed only in one dimension in the z-axis direction, but in reality, the spot is three-dimensional including the x-axis direction and the y-axis direction according to the tumor shape. Placed in. Since the dose distribution needs to be calculated in three dimensions according to the actual tumor shape, the dose evaluation points are also arranged in three dimensions.

  In general, the spot positions in the x-axis direction and the y-axis direction that are perpendicular to the beam traveling direction (z-axis direction) are determined by the beam kick angle, and the beam kick angle depends on the magnetic field strength formed by the scanning device 3. It is decided. The spot position in the z-axis direction, which is the beam traveling direction, is determined depending on the beam energy of the particle beam 2. Therefore, the particle beam irradiation device 30 of the particle beam therapy system 100 adjusts the spot position by adjusting the magnetic field intensity of the scanning device 3 according to the beam energy of the particle beam 2.

  In equation (1), the total dose distribution at the i-th dose evaluation point pi was obtained by adding the dose distribution for each spot. The dose distribution after the irradiation of the particle beam 2 with respect to the same object can be added by time, and can be calculated as in equation (2) as in equation (1). It is.

Here, Formula (2) is a case where the total irradiation time is divided into m time intervals. k is the number of the time interval. At the i-th dose evaluation point pi, when the beam amount irradiated in the k-th time interval is w _k , and the average beam position staying in the k-th time interval is irradiated with a unit beam particle beam The dose (unit particle dose) given to the i-th dose evaluation point pi is defined as _{di, k} . If the time interval is made sufficiently short, this equation (2) can reproduce the dose distribution with high accuracy. Here, it is desirable that the time interval is about the same as or shorter than the required time per spot, and for example, about several tens of microseconds to about 1 millisecond is preferable. Beam amount w _k and the unit particle dose at the same time interval d _i, and multiplying the _k w _{k d} i, _k is the time interval dose.

When the dose evaluation point in Equation (2) is three-dimensional, the dose d _{i, k} can be obtained as follows. It is known that the three-dimensional dose distribution d (x, y, z) can be approximated by the product of the dose distribution in the z direction, the dose distribution in the x direction, and the dose distribution in the y direction. Inaniwa et al. (Non-Patent Document 1) shows a three-dimensional dose distribution d (x, y, z) for one beam in each of the z direction, the x direction, and the y direction as shown in Equation (3). A method for factoring the distribution is introduced.

The left side of Equation (3) is matched with the notation in Equation (1), and the dose d _{i, j} given to the i-th dose evaluation point pi when the particle beam is irradiated to the j-th spot by the unit beam amount It can be expressed as (4).

Here, x _i , y _i , and z _i are the x, y, and z coordinates of the i-th dose evaluation point pi, respectively. Further, x _j and y _j are the x and y coordinates of the j-th spot, respectively, and E _j is the energy of the particle beam when irradiating the j-th spot.

Similarly, when considering the dose distribution divided into time intervals as in equation (2), the dose d given to the i-th dose evaluation point pi when the particle beam is irradiated by the unit beam amount in the k-th time interval. _{i and k} can be expressed as shown in Equation (5).

Here, x _i , y _i , and z _i are the x, y, and z coordinates of the i-th dose evaluation point pi, respectively. Further, x _k and y _k are the x and y coordinates of the beam position in the kth time interval, respectively, and E _k is the energy of the particle beam in the kth time interval.

Regarding the component d _z (z, E) in the z-axis direction of the equations (4) and (5), a calculation based on the theory known as the Bragg equation is possible, but the water phantom (phantom 9) and the dose in advance We think that it is most convenient to actually measure using a meter and create a database. In the case of pre-measurement, the distribution can be obtained by placing water in a water phantom, placing the dosimeter, and moving the position of the dosimeter in the z-axis direction while irradiating the particle beam 2. If the measurement for obtaining the proportionality coefficient C (E) is carried out before carrying out this measurement, the irradiation beam amount w and the QA ionization chamber at that time are arranged upstream. The dose d in the water phantom can be obtained. The dose distribution d _z (z, E) with respect to the unit beam amount can be known by obtaining the ratio.

Regarding the x-axis direction component d _x (x _i −x _j , z _i , E _j ) of Expression (4) and the x-axis direction component d _x (x _i −x _k , z _i , E _k ) of Expression (5) Can be calculated by the multiple scattering theory by Mollier, Fermi-Eyes, Highland et al. Moreover, it is also possible to actually measure in advance using a water phantom (phantom 9) and a dosimeter in advance and create a database. Compared to the measurement of the dose distribution d _z (z, E), the dose distribution changes depending on both x and z, so it is necessary to carry out measurements for all x and z. It is. Therefore, by using a Monte Carlo simulation tool known as Genet 4 or the like, the dose per unit beam amount at an arbitrary position in the water phantom (phantom 9) can be calculated. Specifically, when the Monte Carlo simulation is executed, the shape of an object such as the phantom 9, the energy of the particle beam 2, the generation position and direction of ionization, the electromagnet of the scanning device 3 (x-direction scanning electromagnet 4, y By inputting information such as the beam center axis position deflected by the direction scanning electromagnet 5), the dose per unit beam amount at an arbitrary position in the water phantom (phantom 9) can be calculated. Therefore, if the Monte Carlo simulation is executed, the x-axis direction component d _x (x−x ₀ , z, E) and d _x (x, z, E) can be obtained more efficiently than actually measured. y-direction component, i.e. the same for equation _d y in _{(4) (y i -y j} , z i, E j) d y of or formula _{(5) (y i -y k} , z i, E k) is there.

  When the Monte Carlo simulation tool is used, not only the dose distribution in the one-dimensional direction but also the three-dimensional dose distribution d (x, y, z) can be directly obtained, and this is calculated in advance and d (x , Y, z) is also possible as a database. However, a large amount of memory capacity is required to store the three-dimensional dose distribution in the storage device. Therefore, the data storage is performed in any form in consideration of the performance of the storage device and the required data accuracy. It is necessary to consider whether to do it.

  Next, the flow of dose error distribution calculation using the dose distribution calculation apparatus 1 according to Embodiment 1 of the present invention will be described. The basic flow is the same whether the dose distribution is considered for each spot as in equation (1) or for each time interval as in equation (2). Only the case of thinking for each spot will be described.

In order to calculate the dose error distribution by the dose error distribution calculation unit 13, it is first necessary to calculate the dose distribution in the presence of the irradiation error. In Expression (1) and Expression (4), a beam amount error Δw _j and a spot position error with respect to the beam amount w _j , the spot position x _j , y _j , and the particle beam energy E _j irradiated to the j-th spot, respectively. Consider a case where Δx _j , Δy _j and particle beam energy error ΔE _j occur. At this time, the dose d _{i, j} given to the i-th dose evaluation point pi when the particle beam is irradiated to the j-th spot by a unit beam amount, and the i-th dose evaluation when all the spots are summed Errors Δd _{i, j} and ΔD _i occur in the dose D _{i applied} to the point pi, respectively, and can be expressed as in the equations (6) and (7).

Here, the beam amount error Δw _j , the spot position error Δx _j , Δy _j , and the particle beam energy error ΔE _j are listed as the types of errors, but in the present embodiment, all these errors must be considered. is not. For example, if the accuracy of the particle beam energy E _j is guaranteed as a characteristic of the apparatus and the error ΔE _j does not exist or is negligibly small, it can be ignored and calculated as in equation (8). I do not care.

In addition to the beam amount error Δw _j , the spot position error Δx _j , Δy _j , and the particle beam energy error ΔE _j , there are an infinite number of error factors. You can take it in.

There are various factors that cause the errors Δw _j , Δx _j , Δy _j , and ΔE _j . For example, at the time of irradiation of a particle beam by a general particle beam therapy system, the beam amount w _j of the spot is managed by a dose monitor installed on the beam line, and the irradiation beam amount to the spot is planned in advance. When the amount is reached, control is performed such as temporarily stopping irradiation of the particle beam by the accelerator, or moving the beam position to the next spot by a scanning electromagnet. However, there is always a time lag between when the dose monitor is irradiated with the beam and when the dose monitor detects it and outputs a signal, and between when the dose monitor outputs a signal and when the accelerator or irradiation magnet operates. Therefore, there is a possibility that a beam amount error Δw _j may occur due to the existence.

The spot positions x _j and y _j are controlled by the current value flowing through the scanning electromagnet, but the current value may fluctuate due to the current control error of the electromagnet power source and electromagnetic noise, and the spot position errors Δx _j and Δy _j may occur. is there. Alternatively, spot position errors Δx _j and Δy _j may also occur when the distance between the electromagnet position and the phantom position changes due to thermal expansion or contraction due to temperature change. In order to reduce the position error, feedback control may be performed to correct the spot position based on the output of the beam position monitor installed on the beam line. Depending on the accuracy of the monitor, spot position errors Δx _j and Δy _j may occur.

In addition, for example, in particle beam therapy to an affected area that moves by respiration such as lungs and liver, a gate called a respiration gate is set, and irradiation of a particle beam is performed only when the affected area is within a predetermined range. Respiratory synchronized irradiation may be selected. In that case, QA using a respiratory synchronization phantom that can input movements in three axial directions may be performed as verification before treatment. At this time, the position of the particle beam itself does not change, but the relative position of the particle beam as viewed from the affected area to be irradiated can be different from the planned position when there is no movement of the affected area. Therefore, it can be considered that a spot position error occurs. The magnitudes of the spot position errors Δx _j and Δy _j at this time are determined in accordance with the magnitude of the movement of the affected area within the set respiration gate.

Although the particle beam energy E _j is controlled by the particle accelerator, a beam energy error ΔE _j may be generated due to errors in the magnetic field intensity generated by the beam deflecting electromagnet constituting the accelerator and the electric field intensity and frequency of the high frequency acceleration cavity. There is sex. Further, for example, when the particle accelerator is a cyclotron accelerator and an energy selection system (ESS) is adopted, the beam energy error ΔE _j is caused by the magnetic field strength error of the deflection electromagnet constituting the ESS, the position error of the energy selection slit, or the like. May occur.

There are various methods for estimating the magnitude of these errors, but the simplest method is to measure directly. For example, as a method of measuring the beam amount error Δw _j , a measurement start trigger that operates in accordance with the irradiation start timing of the spot and a measurement end trigger that operates in accordance with the end start timing are set for the dose monitor, It is obtained by calculating the difference between the beam amount measured during this time and the beam amount specified by the treatment plan. Such a measurement can be repeated a plurality of times to grasp the error tendency. For example, it is possible to calculate an average value, variance, standard deviation, and the like of the error Δw _j .

Regarding the spot position, errors Δx _j and Δy _j can be obtained by actually measuring the position with a dosimeter or a gafchromic film and obtaining a difference from the planned position. Regarding energy, an error ΔE _j can be obtained by actually measuring the energy by using a scintillation detector or the like and obtaining the difference from the planned energy. For these errors, by repeating the measurement a plurality of times, it is possible to grasp trends such as the average value, variance, and standard deviation of the errors.

  These error trends may vary depending on the planned beam amount, planned spot position, planned energy, and other conditions. We must understand the difference in trends. In addition, the trend may change depending on various conditions such as season, time, room temperature, and atmospheric pressure. When the error tendency changes depending on the condition, it is necessary to select an appropriate error tendency according to the condition at that time in the operations described in the following paragraphs.

  There are two main methods for calculating the dose error distribution. The former is a method that assumes independence and linearity of error factors, and the latter is a method that uses the concept of an “error scenario”. In the present embodiment, the former will be described.

By the trend analysis of the error [Delta] w _j, it is possible to define a probability error [Delta] w _j follow distribution function _ρ Δwj (u). That is, a function ρ _Δwj (u) is set such that the probability that the magnitude of the error Δw _j is in the range between u and u + du is ρ _Δwj (u) du. The probability distribution function ρ _Δwj (u) is normalized so as to satisfy Expression (9). Similar probability distribution functions ρ _Δxj (u), ρ _Δyj (u), and ρ _ΔEj (u) can be defined for other errors.

From these probability distribution functions and ΔD _i obtained by Equation (6), the expected value E (ΔD _i ), variance V (ΔD _i ), and standard deviation σ (ΔD _i ) of dose error ΔD _i are respectively expressed by equations (10), Formula (11), and Formula (12).

Here, the integration operators ∫... ∫ correspond to the integration variables u _{w, j} , u _{x, j} , u _{y, j} , u _{E, j} , respectively. Since spot numbers j exist from 1 to n, there are 4n integration operators in total, and the integration ranges are all [−∞, ∞].

As described above, the dose error distribution calculation unit 13 responds to information on at least one of the operation error of the device included in the particle beam therapy system, the spot position generated by the movement of the irradiation target, the energy, and the beam irradiation amount. By calculating the dose error distribution in consideration of the effect of the error on the dose distribution, the measurement time can be shortened without reducing the measurement accuracy of the dose distribution. In addition, by displaying the information regarding the dose error ΔD _i calculated as described above to the user using a display unit such as a display, the user can easily make a policy for measuring the QA dose using a phantom. Become.

FIG. 4 shows a display example on the display 15. The dose evaluation point i is plotted on the horizontal axis, the dose is plotted on the vertical axis, the value 16 (D _i + E (ΔD _i )) obtained by adding the expected dose error to the planned dose, and the dose error standard deviation (σ (ΔD _i) )) And values 18 and 19 (D _i + E (ΔD _i ) + σ (ΔD _i ), D _i + E (ΔD _i ) −σ (ΔD _i )) are plotted. By looking at this figure, for example, the user performs a close measurement by reducing the interval between the dose measurement points in the vicinity of the portion where the dose error standard deviation is large, and conversely, the dose is measured near the portion where the dose error standard deviation is small. It is possible to make an efficient dosimetry plan such as increasing the interval between measurement points to shorten the measurement time.

FIG. 4 shows an example in which a distribution having a width of ± σ (ΔDi) is displayed. This is, for example, D _i + E (ΔD _i ) + 2σ (ΔD _i ) and D _i + E (ΔD _i ) −2σ (ΔD). _{As in i} ), it may be a distribution having a width twice as large as ± σ (ΔD _i ), or a distribution having a width three times as large.

  The dose distribution calculation device 1 may search for a dose evaluation point with the largest dose error standard deviation by the dose error distribution calculation unit 13 and display it on the display 35 of the display unit 14 as the maximum error point 22 as shown in FIG. . Further, the dose distribution calculation device 1 searches for a dose evaluation point where the dose error standard deviation exceeds the predetermined reference ranges 31 and 32 by the dose error distribution calculation unit 13, and sets the existence range as a reference as shown in FIG. You may display on the display 45 of the display part 14 as the excess area | regions 33a and 33b. With these displays, the user can more quickly grasp the point or range where the dose error standard deviation is large.

  As described above, in the particle beam therapy system 100 including the dose distribution calculation device 1 according to Embodiment 1 of the present invention, the dose error distribution calculation unit 13 is based on the treatment plan information stored in the treatment plan information storage unit 11. Depending on the information of at least one of the operation error of the device included in the particle beam therapy system stored in the error information storage unit 12, the spot position generated by the movement of the irradiation target, the energy, and the beam irradiation amount with respect to the irradiation Thus, since the dose error distribution is calculated in consideration of the influence of the error on the dose distribution, the measurement time can be shortened without reducing the measurement accuracy of the dose distribution.

Embodiment 2. FIG.
In the first embodiment, the method for assuming the independence and linearity of error factors has been described. In the second embodiment, a method for calculating a dose error distribution using the concept of “error scenario” will be described. The configuration of the particle beam therapy system including the dose distribution calculation apparatus according to the second embodiment is the same as that of the particle beam therapy system 100 according to the first embodiment, and a description thereof is omitted.

FIG. 7 shows the concept of the error scenario in the present embodiment. First, the size of each error Δw _j , Δx _j , Δy _j , ΔE _j is determined for all spots from j = 1 to j = n. A combination of these series of error values is called one “error scenario”.

  The method for determining each error in creating an error scenario is basically the same concept as in the first embodiment, and uses the probability distribution function based on the error tendency investigated in advance, and the dose error distribution calculation unit 13. Randomly determined using random numbers generated by. At this time, if all errors are determined independently and randomly, the same result as in the first embodiment can be expected. However, in the method using the error scenario, the correlation between two or more errors can be considered here. .

For example, if the beam energy error ΔE _j is a positive value, that is, if the beam energy is larger than the planned value, the spot position can be shifted toward the center because the beam is difficult to bend even if the intensity of the magnetic field generated by the scanning electromagnet is the same. High nature. Furthermore, at this time, it cannot be said that the position error Δx _j and the energy error ΔE _j are independent, and the probability distribution function of the position error Δx _j varies depending on the energy error ΔE _j . Since the relationship between the current value flowing through the electromagnet, the beam energy, and the beam kick angle is known in advance, after determining ΔE _j using random numbers based on the probability distribution function of the energy error, spots generated by the energy error The position shift amount is obtained, and Δx _j determined by using another random number based on the probability distribution function under the spot position error is changed by the shift amount, thereby determining an error scenario considering the correlation. be able to.

In general scanning irradiation, the beam stays in one spot on the order of milliseconds to microseconds. However, the time variation of the spot position error is not much due to the prior trend survey. unchanged at an early period, for example, if a fluctuation period of about 100 ms, the difference between the position error [Delta] x j _{+ 1} in the position error [Delta] x _j and the adjacent spots in a certain spot, will not move much. In such a case, the position errors Δx _j and Δx _{j + 1} cannot be said to be independent. In this case, after determining Δx _j using random numbers based on the probability distribution function before, Δx _{j + 1} is selected from a predetermined range from Δx _j that has already been determined, and the probability distribution function is corrected. A method is conceivable. Alternatively, initially advance decided independently and [Delta] x _j and [Delta] x j _{+ 1,} if the difference between the two becomes more than a certain discard the error scenario, independently re [Delta] x _j and the [Delta] x j _{+ 1} with another random number A method is also conceivable in which this is determined and this is repeated until the difference between the two is within a certain range.

As described above, a plurality of error scenarios are generated by changing random numbers. It is considered that the number of error scenarios to be generated is preferably about 1,000 to 1,000,000 depending on conditions. Dose error ΔD _i is calculated based on Equation (6) and Equation (7) for each error scenario. When the number of error scenarios is N, the error scenario number is s, and the dose error corresponding to the s-th error scenario is expressed as ΔD _{i, s} , the expected value of dose error E (ΔD _i ) and variance V (ΔD _i ) , Standard deviation σ (ΔD _i ) can be obtained as shown in equations (13), (14), and (15), respectively.

As described above, the dose error distribution calculation unit 13 prepares a plurality of error scenarios as a set of error values based on the error tendency of the spot position, energy, and beam amount, and each dose corresponding to each error scenario. By calculating the distribution and calculating the variation for each error scenario at each calculation point as a dose error, even if two or more error factors have a correlation and the effect is large, an accurate dose error distribution Can be calculated. In addition, the information regarding the dose error ΔD _i calculated as described above is displayed to the user on a display unit such as a display by the same display method as in the first embodiment, so that the user can perform QA using a phantom. It becomes easy to make a dosimetry policy.

  As described above, in the particle beam therapy system 100 including the dose distribution calculation apparatus 1 according to the second embodiment of the present invention, the dose error distribution calculation unit 13 is based on the error tendency of the spot position, the energy amount, and the beam amount. Because we prepared multiple error scenarios that are error value sets, calculated each dose distribution corresponding to each error scenario, and calculated the variation for each error scenario at each calculation point as dose error, Not only can the measurement time be shortened, but an accurate dose error distribution can be calculated.

Embodiment 3 FIG.
In the first and second embodiments, the method of performing the effective QA measurement by performing the dose error distribution calculation before the dose measurement has been described. However, in the third embodiment, the dose error distribution calculation and the dose measurement are alternately performed. A method for efficiently performing QA measurement by repeating the above will be described. The configuration of the particle beam therapy system including the dose distribution calculation apparatus according to the third embodiment is the same as that of the particle beam therapy system 100 according to the first embodiment, and a description thereof is omitted.

In Embodiments 1 and 2, the dose evaluation point i is expressed in the form of dose values and dose error values for discrete dose evaluation points i, such as the dose D _i and the dose error ΔD _i , but the dose evaluation point i is sufficiently small. By taking an interval or interpolating between adjacent dose evaluation points, it is possible to easily express the dose at continuous evaluation points (x, y, z) as well as a dose error. That is, D _i can be rewritten as D (x, y, z) and ΔD _i can be rewritten as ΔD (x, y, z). The following paragraphs will be explained on the premise of notation by continuous evaluation points.

  FIG. 8 is a flow chart of dose error distribution calculation using the dose distribution calculation device 1 according to Embodiment 3 of the present invention. A flow in which the dose error distribution calculation and the dose measurement are alternately repeated according to the third embodiment will be described with reference to FIG. 9 to 13 are image diagrams of error scenarios corresponding to the steps in the flowchart of FIG.

  First, in step S801, the dose distribution calculation device 1 uses the dose error distribution calculation unit 13 to generate a plurality of error scenarios 41 in the same manner as described in the second embodiment (see FIG. 9). Subsequently, in step S802, a plurality of dose distributions corresponding to the plurality of error scenarios are calculated, their expected values and standard deviations are calculated, and a maximum error occurrence position 43, which is a coordinate at which the standard deviation of the dose error is the largest. Is specified (see FIG. 10).

  Next, in step S803, the dose error distribution calculation unit 13 performs dose measurement at the coordinates, and after confirming the measurement result 44, information on the measurement result, that is, dose measurement position and measurement dose value, arrangement position error and measurement. The dose value error is input to the error information storage unit 12 (see FIG. 11). In the dose measurement, the dose measuring device 7 is arranged at the coordinates in the water phantom 9 and the particle beam 2 is actually irradiated. At this time, it is assumed that the arrangement position error and the measurement dose error of the dose measurement device 7 are known in advance as specifications of the dose measurement device arrangement jig and the dose measurement device.

  Subsequently, in step S804, the dose error distribution calculation unit 13 classifies all error scenarios into two groups, one that does not match the measurement result and one that does not match the measurement result, and discards the error scenario belonging to the group that does not match (FIG. 12). Here, as a criterion for judgment, the dose error distribution calculation unit 13 compares a plurality of dose distributions corresponding to a plurality of error scenarios with the input measurement information, and arranges them for the measurement position and the dose measurement value. When the dose distribution falls within the allowable range of the position error and the dose measurement value error, the error scenario corresponding to the dose distribution is set to “adapt to the measurement result”. Conversely, if the dose distribution does not fall within the allowable range of the placement position error and dose measurement value error for the measurement position and dose measurement value, the error scenario corresponding to the dose distribution is determined to be “not suitable for the measurement result”. .

  Finally, in step S805, only the error scenario belonging to the matching group in which the dose distribution falls within the allowable range 46 of the placement position error and the dose measurement value error with respect to the measurement position and the dose measurement value is used. The expected value, variance, and standard deviation are recalculated and redisplayed on the display (see FIG. 13). By doing in this way, the user can know the information of the dose error distribution that matches the measurement result and reflect it in the second and subsequent dose measurements.

  The dose distribution calculation device 1 further classifies the remaining error scenarios into two groups, one that matches the second dose measurement result and one that does not match, by the dose error distribution calculation unit 13, and the error belonging to the group that does not match The scenario is discarded and only the error scenario belonging to the matching group is used, and the expected value, variance, and standard deviation of the dose error are recalculated and redisplayed on the display. The user may continue this operation until the end condition is satisfied. For example, when the dose distribution falls within a predetermined tolerance for all error scenarios, or the standard deviation of the dose error falls within a predetermined tolerance for all coordinates. Or when the dose error distribution calculation and dose measurement are repeated a predetermined number of times, or when the user decides to stop repeating for other reasons.

  As described above, in the particle beam therapy system 100 including the dose distribution calculation apparatus 1 according to Embodiment 3 of the present invention, the measurement position, the measured dose value, the measurement position error in the dose measurement at the time of QA measurement using a phantom, The measurement dose value error is input to the error information storage unit 12, and the dose error distribution calculation unit 13 inputs the information on the measurement position and the measurement dose value input to the error information storage unit 12, and the error scenario corresponding to each error scenario. The dose distribution is collated and classified into the error scenario that matches the measurement position and the measured dose value within the allowable range of the measurement position error and the measured dose value error, and the error scenario that does not match, leaving only the compatible error scenario. Since the dose error is calculated again based on the variation for each error scenario, not only can the measurement time be shortened, It is possible to improve the accuracy of QA measured by the measurement is repeated a plurality of times.

Embodiment 4 FIG.
In the fourth embodiment, a method of classifying and selecting error scenarios into a plurality of groups based on measurement results will be described. The configuration of the particle beam therapy system including the dose distribution calculation apparatus according to the fourth embodiment is the same as that of the particle beam therapy system 100 according to the first embodiment, and a description thereof is omitted.

  First, the dose distribution calculation device 1 performs a dose error distribution calculation by the dose error distribution calculation unit 13. After determining a plurality of dose measurement coordinates, the dose error distribution calculation unit 13 performs dose measurement using the dose measurement device 7 at each coordinate, and inputs the result to the error information storage unit 12. Subsequently, in the dose distribution calculation device 1, the dose error distribution calculation unit 13 calculates a parameter called “reliability” for each error scenario based on the measurement result stored in the error information storage unit 12.

“Reliability” is a parameter having a characteristic that the dose distribution corresponding to the error scenario is closer to the measurement result, and an example of its definition will be described here. s is an error scenario number, t is a dose measurement coordinate number, D _s (x, y, z) is a dose distribution corresponding to the sth error scenario, and (x _{m, t} , y _m ) is a tth dose measurement coordinate. _{, T 1} , z _{m, t} ), where D _{m, t} is the measured dose value at the t-th dose measurement coordinate, the dose distribution D _s (x, y, z) and the measurement result (D _{m, t} , x _{m , T 1} , y _{m, t 1} , z _{m, t 2} ) and “distance” 1 _{s, t} are defined as in Expression (16).

ここで、σ_Ｄならびにσ_ｒは線量と座標の単位系を揃えるための基準定数であり、例えば測定線量値誤差と位置誤差を用いることが推奨される。

Here, σ _D and σ _r are reference constants for aligning the dose and coordinate unit systems, and it is recommended to use, for example, a measured dose value error and a position error.

  The reliability Rs of the s-th error scenario is expressed by the following equation (17) as the reciprocal of the square root of the sum of squared distance values for all measurement points (from t = 1 to m) with respect to the s-th error scenario. Can be defined.

信頼度の値が大きいほど、その誤差シナリオに対応する線量分布と線量測定結果との平均距離が短く、すなわち測定結果に近い誤差シナリオであると言うことが出来る。

It can be said that the larger the reliability value, the shorter the average distance between the dose distribution corresponding to the error scenario and the dose measurement result, that is, an error scenario closer to the measurement result.

  Based on the reliability value, the dose distribution calculation device 1 can divide the error scenarios into a plurality of groups by the dose error distribution calculation unit 13. For example, among a plurality of initially prepared error scenarios, a half is group A and the other half is group B in descending order of reliability. In this case, by discarding the error scenario belonging to group B and recalculating the expected value, variance, and standard deviation of the dose error only in the error scenario belonging to group A, and displaying them on the display unit 14 as shown in FIG. The user can know the dose error distribution reflecting the measurement result. Here, other methods for classifying the error scenarios into two groups are conceivable. For example, 60% may be group A and the remaining 40% may be group B in descending order of reliability, and a threshold value of reliability is determined in advance. Things may be group B.

Further, the grouping by reliability does not necessarily need to be two groups. For example, it is possible to assign one third to group A, the next highest third to group B, and the remaining third to group C in descending order of reliability. At this time, the set I is defined as the group A only, the set II is defined as the union of the groups A and B, and the set III is defined as the union of the groups A, B, and C. Recalculate the expected dose error, variance, and standard deviation for the dose distribution corresponding to the scenario. Using the recalculated results, a graph of D _i + E (ΔD _i ) + σ (ΔD _i ) and D _i + E (ΔD _i ) −σ (ΔD _i ) as shown in FIG. 55. By doing so, the user can more visually understand the range of dose reliability.

It is also possible to recalculate the dose error distribution using the reliability parameter without grouping. When calculating the variance and expected value of the dose error is to add an error scenario weight coefficient W _s in accordance with the reliability of the error scenario, it is possible to reflect darker more results reliable error scenarios. In this case, the expected value E (ΔDi), variance V (ΔDi), and standard deviation σ (ΔDi) of the dose error can be obtained as shown in Equation (18), Equation (19), and Equation (20), respectively.

Here, how to determine the error scenario weight coefficient W _s is variously considered, but it is important that there is a relationship of an error scenario weight coefficient increases as the confidence is also high. The simplest way to decide is to make the value of the error scenario weight coefficient the same as the reliability, that is, to define it as equation (21).

  As described above, in the particle beam therapy system 100 including the dose distribution calculation apparatus 1 according to the fourth embodiment of the present invention, the measurement position, the measured dose value, the measurement position error in the dose measurement at the time of QA measurement using the phantom, The measured dose value error is input to the error information storage unit 12, and the dose error distribution calculation unit 13 trusts each error scenario based on the information on the measurement position and the measured dose value input to the error information storage unit 12. The error scenario is classified into at least two groups based on the calculated reliability, and the dose error distribution according to the variation in dose distribution corresponding to the error scenario belonging to at least one of the classified groups Since not only the measurement time can be shortened, but multiple error scenarios are classified into groups and selected. It is possible to improve the accuracy of QA measurements Te.

Embodiment 5. FIG.
In the fifth embodiment, as in the second embodiment, the dose error distribution calculation apparatus generates a plurality of error scenarios, sets a measurement interval based on an average autocorrelation function, and calculates a dose distribution. explain. The configuration of the particle beam therapy system including the dose distribution calculation apparatus according to the fifth embodiment is the same as that of the particle beam therapy system 100 according to the first embodiment, and the description thereof is omitted.

For the dose error ΔD _s (x, y, z) for one error scenario s, the autocorrelation function I _s (τ _x , τ _y , τ _z ) can be expressed as in equation (22).

  Here, it is desirable that the integration range of the three-dimensional integral 取 る is ideally set from minus infinity to infinity for all of x, y, and z. However, in reality, dose distribution in the infinite range should be calculated. Therefore, the range may be limited to an important part, and for example, a range in which the dose distribution must be confirmed in the region of interest, that is, the patient QA may be set as the integration range. In order to simplify the explanation, here, Equation (23), which is an autocorrelation function focusing only on the x direction, is considered.

このとき、全ての誤差シナリオに対する平均の自己相関関数Ｉ_{ＡＶＥＲＡＧＥ}（τ_ｘ）は、式（２４）のように計算できる。

At this time, the average autocorrelation function I _AVERAGE (τ _x ) for all error scenarios can be calculated as shown in Equation (24).

Generally, the function D _s (x, y, z) does not become a periodic function because it includes an error factor, and the average autocorrelation function I _AVERAGE (τ _x ) is τ _x = 0 as shown in FIG. At the maximum. If the value of the average autocorrelation function I _{AVERAGE (τ x)} is positive, the dose error value at a point x, the dose error value in tau _x apart points x + tau _x therefrom, is somewhat closer to the value It means that the possibility is high. And the larger the value of the average autocorrelation function I _AVERAGE (τ _x ), the more likely the dose error values at the two points are closer, so the significance of measuring both of the two points in the QA measurement is less. Become. Therefore, if the interval between the dose measurement points is determined based on the average autocorrelation function I _AVERAGE (τ _x ), efficient QA measurement can be performed. Here, determining the interval between the dose measurement points based on the average autocorrelation function is, for example, setting the half width (Half Width at Half Maximum) 51 of the autocorrelation function as the measurement interval.

Similar calculations can be performed for the y and z directions, and average autocorrelation functions I _AVERAGE (τ _y ) and I _AVERAGE (τ _z ) in the x and y directions can be obtained, and their half-value half widths can also be obtained. Is possible. In QA measurement, a user can set measurement intervals based on respective average autocorrelation functions in the x, y, and z directions, and can arrange measurement points in three dimensions.

  As described above, in the particle beam therapy system 100 provided with the dose distribution calculation apparatus 1 according to Embodiment 5 of the present invention, the dose error distribution calculation unit 13 causes the autocorrelation function for the dose distribution corresponding to each error scenario. Since the value of the dose measurement interval is calculated based on the autocorrelation function, efficient QA measurement can be performed.

  It should be noted that the present invention can be freely combined with each other within the scope of the invention, and each embodiment can be appropriately modified or omitted.

  DESCRIPTION OF SYMBOLS 1 Dose error distribution calculation apparatus, 2 Particle beam, 11 Treatment plan information storage part, 12 Error information storage part, 13 Dose error distribution calculation part, 100 Particle beam therapy apparatus.

Claims (9)

A treatment plan information storage unit for storing treatment plan information;
An error information storage unit for storing error information;
A dose error distribution calculation unit that calculates a dose error distribution according to the error information stored in the error information storage unit for irradiation based on the treatment plan information stored in the treatment plan information storage unit ; , equipped with a,
The treatment plan information includes the position of the spot determined by the treatment plan, and the energy and beam amount of the particle beam irradiated to each spot,
The error information includes at least one error among an operation error of the device, a position of the spot that can be generated by movement of an irradiation target, the energy, and the beam amount,
The dose error distribution calculation unit creates an error scenario that is a combination of the error values based on the tendency of the spot position, energy, and beam amount error, and calculates a dose distribution corresponding to the error scenario and, the linear amount error distribution calculation device you and calculates the variance for each error scenario as distribution of the dose error. The dose error distribution calculation device according to claim 1 , wherein the dose error distribution calculation unit searches for a point having the largest dose error from a region designated in advance. 3. The dose according to claim 1, wherein the dose error distribution calculation unit searches a range in which the magnitude of the dose error is larger than a predetermined allowable range from a predetermined region. Error distribution calculation device. The dose error distribution calculation unit inputs a dose measurement position, a dose value, a position error, and a dose value error to the error information storage unit, and corresponds to the position, the dose value information, and the error scenario. The calculated dose error distribution is collated, and the position and the dose scenario are classified into the error scenario that matches the position error and the dose value error within the allowable range of the position error and the error scenario that does not match, and the fit dose error distribution calculation device according to any one of claims 1 to 3, characterized in that the operation again as the distribution of the dose error based on a variation of each of the matching error scenario only error scenarios that. The dose error distribution calculation unit inputs a dose measurement position and a dose value to the error information storage unit, and calculates a reliability for the error scenario based on the position and the dose value information, in accordance with the reliability that the calculated dose error distribution calculation device according to any one of claims 1 to 3, and calculates the distribution of the dose error. The dose error distribution calculation unit classifies the error scenarios into at least two groups based on the calculated reliability, and based on variations in dose distribution corresponding to the error scenarios belonging to at least one of the classified groups. The dose error distribution calculation device according to claim 5 , wherein the dose error distribution is calculated. The dose error distribution calculation unit sets a weighting coefficient for each error scenario based on the reliability that the calculated, in accordance with the weighting factor, according to claim 5, characterized in that for calculating the distribution of the dose error The dose error distribution calculation device described in 1. The dose error distribution calculation unit, based on the autocorrelation function for the dose distribution corresponding to the error scenario, set the dose measurement interval, one of claims 1 to 3, characterized by calculating the distribution of the dose error or dose error distribution calculation device according to item 1. A particle beam therapy system comprising the dose error distribution calculation device according to any one of claims 1 to 8 .

Images