JP2025037011A - Dose distribution calculation system, program, and dose distribution calculation method - Google Patents

Abstract

To provide a dose distribution calculation system, a program and a dose distribution calculation method capable of obtaining a calculation result of highly accurate dose distribution in a shorter time compared to conventional arts.SOLUTION: A dose distribution calculation system 113 calculates a transmittance of a multi-leaf collimator for each transported particle, and calculates dose distribution using the transmittance. When calculating the transmittance, the dose distribution calculation system sets a high importance condition based on an aperture shape of the multi-leaf collimator or a particle production process, determines whether transported particles satisfy the high importance condition, and for transported particles that are determined not to satisfy the high importance condition, increases the weight of the transported particles, and reduces the transported particles so that an expectation value of the weight of all transported particles is preserved.SELECTED DRAWING: Figure 5

Description

The present invention relates to a dose distribution calculation system, program, and dose distribution calculation method that are suitable for use in a radiation therapy system for irradiating radiation to an affected area such as a tumor to treat the area.

Non-Patent Document 1 describes a multi-leaf collimator particle transport model using a semi-analytical computational method for use in Monte Carlo dose distribution calculations that is accurate, fast, and efficient, and therefore applicable to iterative intensity modulated radiation therapy dose distribution calculations.

J. V. Siebers et al., “A method for photon beam Monte Carlo multileaf collimator particle transport”, Phys. Med. Biol. 47, 3225-3249 (2002).

Radiation therapy is a localized treatment that kills cancer cells by irradiating the cancer lesion with radiation and damaging the DNA of the cancer cells.

X-rays are the predominant type of radiation used, and treatment is carried out using high-precision irradiation methods that irradiate radiation from multiple directions and improve dose concentration, such as Intensity Modulated Radiation Therapy (IMRT), Volumetric Modulated Arc Therapy (VMAT), and Dynamic Wave Arc (DWA), which performs continuous non-coplanar irradiation with a wave-like trajectory.

In VMAT, continuous rotational irradiation is performed while continuously modulating three parameters: gantry angle, multi-leaf collimator (MLC) aperture shape, and dose rate, making it possible to deliver a high dose to the tumor while reducing the dose to normal tissue.

In radiation therapy, the dose distribution inside the patient's body is simulated using treatment planning software based on the patient's CT images and the prescribed dose set by the doctor, and irradiation parameters such as the treatment device's equipment parameters (such as the collimator opening shape), irradiation angle, irradiation amount per irradiation angle, and energy are determined.

Here, the determined irradiation parameters and the dose distribution on which the irradiation parameters are calculated are called a treatment plan, and the irradiation device irradiates the beam based on the treatment plan to form the desired dose distribution in the affected area.

To ensure the soundness of the treatment plan used in treatment, it is checked whether the device can irradiate the dose distribution according to the treatment plan before irradiating the patient. This is called patient QA. The most common method used for patient QA is to actually irradiate a homogeneous phantom such as a solid phantom and measure the dose distribution.

Through patient QA, clinical staff will (1) verify the calculation accuracy of the dose distribution calculation algorithm in the treatment planning software, (2) verify the effect of irradiation errors of the irradiation device on the dose distribution within a homogeneous phantom, and (3) confirm the validity of device operation in relation to the treatment plan.

Items (1) and (2) are verified by comparing the actual measurement results with the calculation results of the treatment planning software. Each treatment facility sets its own patient QA evaluation items and judgment criteria, and once it has been confirmed that each evaluation item meets the judgment criteria, it can be used as a treatment plan. If the criteria are not met, patient QA is re-performed and the treatment plan is re-created.

Radiation therapy involves irradiating the affected area once a day for several days to several weeks. As a result, the tumor may expand or shrink during the treatment period, and the internal structure around the tumor may change.

When such changes occur, there is an increasing need for adaptive treatment, which recreates the treatment plan and uses the recreated treatment plan to perform treatment. One type of adaptive treatment in which the treatment plan is recreated immediately before treatment while the patient is lying in the treatment bed is called online adaptive treatment (OART).

OART is a treatment method that optimizes the radiation dose to accommodate the patient's body shape, which varies from treatment day to treatment day, such as the condition of the nasal cavity or intestines. By optimizing the dose distribution for each treatment day, the area outside the target that is irradiated as a margin can be made smaller, making it possible to reduce damage to normal tissue.

However, with OART, the number of processes carried out during treatment, such as recreating treatment plans, increases significantly, which presents various development challenges, such as increased treatment time. One of the challenges is patient QA during OART.

In OART, patient QA of the replanned treatment plan must be performed quickly while the patient is lying on the bed, making it difficult to perform the actual measurement-based patient QA that has been performed up until now. For this reason, in recent years, a method has been proposed for patient QA in which dose distribution calculations are performed using high-precision dose distribution calculation methods such as the Monte Carlo method to check whether the dose distribution can be formed as planned.

In the Monte Carlo method, random numbers are used to probabilistically reproduce the behavior of individual radiation, and this is repeated many times (several million times or more) to estimate the average behavior of radiation and calculate the dose distribution. If a Monte Carlo simulation is performed in an integrated manner from the stage where electrons accelerated in an accelerator generate X-rays, to the stage where X-rays pass through the MLC to form an irradiation field, to the stage where X-rays that have passed through the MLC deliver a dose inside the patient, it would take an enormous amount of calculation time.

For this reason, when calculating the dose distribution in X-ray therapy, it is common to calculate the dose distribution in two steps: (A) the process up to the formation of the irradiation field by the MLC (transport calculation within the irradiation device), and (B) the process in which the X-rays deliver the dose inside the patient's body (internal transport calculation).

This invention relates to (A) transport within the irradiation device. In VMAT irradiation, etc., the leaf shape changes in a complex manner according to the rotation of the gantry, so in order to efficiently perform transport calculations for the MLC irradiation field formation process, it is common to simulate the irradiation field formation process using analytical or semi-analytical methods.

Non-Patent Document 1 shows a calculation method for the irradiation field calculation process using a semi-analytical method. In the semi-analytical method, particle type information (particle type, position, energy, and traveling direction information) upstream of the MLC is generated in advance by calculating or modeling using Monte Carlo simulation, and this particle type information is used as input to calculate the path length within the MLC for each particle. The transmittance calculated from the calculated path length is set as the particle weight, thereby taking into account the effect of the MLC shape.

The number of particles used in Monte Carlo simulations ranges from tens to hundreds of millions, and in semi-analytical methods, the calculation of the MLC path length is repeated several times for each particle, and the transmittance is calculated for each particle. Therefore, calculations for particles passing through the closed part of the MLC take longer than calculations for particles passing through the MLC opening.

In high-precision irradiation, such as VMAT and DWA, the dose distribution is formed by overlapping irradiation fields irradiated from multiple irradiation directions, so the importance of the low-dose region of the MLC closure increases. If the calculation of particles passing through the closure is ignored, there was a concern that the calculation accuracy would decrease. In particular, in small radiation field cases, the opening area is small, so the calculation cost of the transmittance calculation increases, which was an issue.

The present invention provides a dose distribution calculation system, program, and dose distribution calculation method that can obtain highly accurate dose distribution calculation results in a shorter time than conventional methods.

The present invention includes multiple means for solving the above problems, and one example is a dose distribution calculation system that calculates the transmittance of a multi-leaf collimator for each transport particle and calculates a dose distribution using the transmittance. When calculating the transmittance, a high importance condition is set based on the aperture shape of the multi-leaf collimator or the particle generation process, and it is determined whether the transport particle satisfies the high importance condition. For transport particles that are determined not to satisfy the high importance condition, the weight of the transport particle is increased and the transport particle is reduced so that the expected value of the weight of all transport particles is preserved.

According to the present invention, it is possible to obtain highly accurate dose distribution calculation results in a shorter time than in the past. Problems, configurations, and effects other than those described above will become clear from the explanation of the following examples.

1 is an overall configuration diagram of a radiation therapy system according to an embodiment of the present invention. FIG. 13 is a diagram showing a flow of a calculation process of a dose distribution using the dose distribution evaluation system of the present embodiment. FIG. 13 is a diagram showing a flow of the MLC transport calculation process when the dose distribution evaluation system of this embodiment is used. 13 is a diagram showing a process flow for performing dispersion reduction processing of transported particles based on the MLC aperture shape when using the dose distribution evaluation system of this embodiment. FIG. 11 is an explanatory diagram of a process for generating a high importance region when a rectangle circumscribing an opening of an MLC is used as a reference opening region. 13 is a diagram for explaining the process of generating a high importance region of transport particles when a rectangle circumscribing the opening of the MLC between two control points is used as the reference opening region. FIG. 13 is a diagram showing a process flow for generating a high importance region of transport particles when the collimator opening shape is used as the reference opening region between two control points. FIG. 13 is a diagram for explaining the process of generating a high importance region of transport particles when the collimator opening shape is used as the reference opening region between two control points. FIG. 13 is a diagram showing a process flow for performing a dispersion reduction process of transported particles based on the generation process of transported photons when using the dose distribution evaluation system of this embodiment. FIG. 10 is a comparison diagram of fluence distribution at a downstream position of the MLC with and without the dispersion reduction treatment of transported particles in this embodiment.

The following describes an embodiment of the dose distribution calculation system, program, and dose distribution calculation method of the present invention with reference to Figs. 1 to 10. Note that in the drawings used in this specification, identical or corresponding components are denoted by the same or similar reference numerals, and repeated explanations of these components may be omitted.

The present invention can be suitably applied to an X-ray irradiation system that uses X-rays as the radiation used, but can also be suitably applied to a radiation irradiation system that uses neutrons other than X-rays. In the following embodiment, an X-ray irradiation system will be used as an example.

First, the overall configuration of the X-ray therapy system of the present invention will be described using Figure 1. Figure 1 is a diagram showing an outline of the overall configuration of the X-ray therapy system.

The X-ray therapy system shown in FIG. 1 includes an X-ray irradiation system 110, a data server 111, a treatment planning device 112, and a dose distribution calculation system 113.

The X-ray irradiation system 110 includes an X-ray irradiation device 100, a radiation irradiation control device 104, a communication device 105, a storage device 106, and an input device 107. The X-ray irradiation device 100 includes a ring-type gantry 101, an irradiation nozzle 102, and a bed 103.

In the X-ray irradiation system 110, when an extraction start signal is output from the radiation irradiation control device 104, the linear accelerator in the irradiation nozzle 102 accelerates the electron beam and generates X-rays by irradiating the accelerated electron beam onto tungsten.

The irradiation nozzle 102 is equipped with a collimator (hereafter, multi-leaf collimator (MLC)) consisting of multiple plate-shaped shields (hereafter, leaves) arranged on the left and right, and the generated X-rays are shaped into the desired distribution by arbitrarily changing the position of each leaf based on the instruction value from the radiation irradiation control device 104.

In addition, a dose monitor that measures the amount of X-ray irradiation is provided inside the irradiation nozzle 102, and the detected measurement value is output to the radiation irradiation control device 104 and used for control during radiation irradiation.

The bed on which the irradiation target A is placed is called the bed 103. The bed 103 can be moved in the directions of three orthogonal axes based on instructions from the radiation irradiation control device 104, and can also be rotated independently around each axis. Through these movements and rotations, the position of the irradiation target A can be moved to the desired position.

The irradiation nozzle 102 is mounted on a ring-type gantry 101, and the gantry rotation angle and ring rotation angle are set in response to instructions from a radiation irradiation control device 104, making it possible to irradiate X-rays from any angle onto an irradiation target A on a bed 103. The center of rotation of the gantry and ring is called the isocenter.

The radiation irradiation control device 104 is connected to the ring-type gantry 101, the irradiation nozzle 102, the bed 103, the communication device 105, the memory device 106, the input device 107, etc., and controls the operation of the equipment in the irradiation nozzle 102, the ring-type gantry 101, the bed 103, etc.

The communication device 105 is connected to the data server 111 via a network, and obtains treatment plan data from the data server 111, which contains irradiation parameters (gantry angle, irradiation dose, leaf position information, etc.) created by the treatment planning device 112 via the network before irradiation, and stores the irradiation parameters in the storage device 106.

The input device 107 is connected to the radiation irradiation control device 104, and displays information on a monitor based on signals acquired from the radiation irradiation control device 104. It also receives input signals from a medical professional who operates the X-ray irradiation system 110, and transmits various control signals to the radiation irradiation control device 104. When the radiation irradiation control device 104 receives an instruction to start irradiation of radiation via the input device 107, it starts irradiation based on the irradiation parameters stored in the storage device 106.

The dose distribution calculation system 113 is a system that calculates the transmittance of the multi-leaf collimator for each transported particle and calculates the dose distribution using the transmittance, and is composed of an MLC transport calculation unit 120, a dose distribution calculation unit 121, a display unit 122, an input unit 123, and a memory unit 124. The dose distribution calculation system 113 is connected to the data server 111 via a network, and acquires treatment plan information for calculating the dose distribution.

The MLC transport calculation unit 120 has a transmittance calculation unit 120A and a dispersion reduction unit 120B.

The transmittance calculation unit 120A receives input of particle species information on the upper surface of the MLC stored in the memory unit 124, calculates the path length within the MLC for each particle, and sets the transmittance calculated from the path length as the particle weight to generate particle species data downstream of the MLC. Here, the particle species data includes position information, travel vector, energy, particle species (X-ray, electron, positron), and statistical weight for each particle/photon. This transmittance calculation unit 120A preferably executes the procedure for calculating the transmittance of the multi-leaf collimator for each transported particle.

In order to reduce the calculation cost of the transmittance calculation unit 120A, the variance reduction unit 120B sets a high importance condition based on the aperture shape of the multi-leaf collimator or the particle generation process when calculating the transmittance, determines whether the transported particles satisfy the high importance condition, and for transported particles determined not to satisfy the high importance condition, increases the weight of the transported particles and reduces the number of transported particles so that the expected value of the weight of all transported particles is preserved. In this way, by reducing the transported particles that do not satisfy the high importance condition, the number of particles for which transport calculations are performed within the MLC is reduced, and the calculation load of the transmittance calculation is reduced.

Here, the "high importance condition" in the present invention refers to, for example, a rectangular area (irradiation field size) circumscribing the opening shape of the MLC, and an area obtained by expanding the irradiation field size, and whether or not the transported particles satisfy the high importance condition can be determined by whether or not the transported particles pass through the high importance area.

The dose distribution calculation unit 121 receives as input the particle species data downstream of the MLC generated by the MLC transport calculation unit 120 and the transmittance, and calculates the dose distribution inside the patient's body or in a numerical phantom. Here, the particle species data includes position information for each particle/photon, a travel vector, energy, particle species (X-ray, electron, positron), and statistical weight. This dose distribution calculation unit 121 preferably executes the procedure for calculating the dose distribution using the transmittance.

The display unit 122 is a display that displays the dose distribution calculated by the dose distribution calculation unit 121.

The radiation irradiation control device 104 and the dose distribution calculation system 113 each have a central processing unit (CPU) and a memory connected to the CPU.

The control processes for the operations to be executed may be integrated into one program, or may be divided into multiple programs, or may be a combination of these.

Some or all of the programs held by each device may be implemented using dedicated hardware, or may be modularized. Furthermore, the various programs may be installed on each device via a program distribution server or external storage media, or existing devices may be updated.

In addition, each device may be an independent device connected to a wired or wireless network, or two or more devices may be integrated together.

Next, the flow of calculating the dose distribution inside the patient's body using the dose distribution calculation system 113 of this embodiment will be described. The overall flow of the dose distribution calculation is shown in FIG. 2, and the flow of the transport calculation part in the MLC using the variance reduction unit 120B, which is a characteristic process of the present invention, is shown in FIG. 3. The dose distribution can be obtained during online adaptive treatment or when calculating the dose distribution during treatment planning, but is not limited to these timings.

As shown in FIG. 2, first, the dose distribution calculation system 113 acquires treatment plan data for verifying the dose distribution from the data server 111 (S201). At this time, the treatment plan data to be read by the dose distribution calculation system 113 can be selected by the operator from the treatment plan information 111A in the data server 111 using the input unit 123, or the dose distribution calculation system 113 can automatically search the treatment plan information 111A in the data server 111 for the treatment plan data read by the radiation irradiation control device 104 and read it.

Next, the MLC transport calculation unit 120 of the dose distribution calculation system 113 extracts plan parameters to be used for dose distribution calculation and MLC transport calculation from the treatment plan data read in step S201, and stores them in the memory unit 124 (S202).

Next, the MLC transport calculation unit 120 calculates the total number of samples at the MLC upstream position using the plan parameters of the treatment plan data read in step S202 (S203).

The number of samples here is the number of particles used in dose distribution calculations, and since the X-ray dose in the X-ray irradiation device 100 is determined by the MU value of the dose monitor installed in the irradiation nozzle, the number of particles per MU value is often used. Here, an example will be explained using the number of samples [particles/(MU/Gy)] calculated per unit dose at the normalization point.

The MLC transport calculation unit 120 acquires dose value and MU value information at the normalization point from the treatment plan information, and calculates the MU value per unit dose [MU/Gy] at the normalization point. The number of particles per MU value per unit dose [particles/(MU/Gy)] is stored in advance in the memory unit 124, and the total number of samples is calculated by multiplying the MU value per unit dose at the normalization point by the number of particles per MU value per unit dose. Here, the number of particles per MU value per unit dose [particles/(MU/Gy)] specified by the operator using the input unit 123 can also be used.

Next, the MLC transport calculation unit 120 reads particle species data upstream of the MLC and performs transport calculations within the MLC to generate particle species data downstream of the MLC (S204).

After the transport calculation is completed, the dose distribution calculation unit 121 of the dose distribution calculation system 113 reads the particle species data downstream of the MLC generated by the MLC transport calculation unit 120, and performs dose distribution calculation (S205). In the dose distribution calculation, assuming that the statistical weight of the input particle k is s _k and the dose that the particle k imparts to the voxel i is d _i,k , the dose D _i of the voxel i is calculated by the following formula (1).

where C is a correction factor to convert the dose calculated using the number of calculation samples to the dose given using the actual number of particles.

The present invention is characterized by the calculation method of the MLC transport calculation portion in step S204. During the MLC transport calculation, the variance reduction unit 120B sets high importance conditions, which mean conditions under which the importance of transported particles in dose evaluation is increased, for the transported particles, and for transported particles that do not satisfy the high importance conditions, increases the weights of the transported particles and reduces the particles while preserving the expected value of the weights of all transported particles. This makes it possible to reduce the calculation load of the transmittance calculation unit 120A. Figure 3 shows a flow diagram of the MLC transport calculation using the variance reduction unit 120B.

As shown in FIG. 3, first, the MLC transport calculation unit 120 reads the system data of the multi-leaf collimator used in the X-ray irradiation device 100 and constructs a calculation system (step S301).

A multi-leaf collimator has multiple plate-shaped shields (hereafter referred to as leaves) that can be driven in one axial direction arranged in a row. The direction in which the leaves are driven is called the leaf drive direction, and the direction in which the leaves are arranged is called the leaf arrangement direction. In the leaf arrangement direction, a frame is placed on the outside of the leaves, and its role is to shield the radiation outside the irradiation field.

Next, the MLC transport calculation unit 120 reads the MLC upstream particle species information database 125 stored in the memory unit 124, and obtains particle data upstream of the MLC that will be input for the MLC transport calculation (step S302).

Next, the transmittance calculation unit 120A of the MLC transport calculation unit 120 reads particles one by one from the particle species data upstream of the MLC read in step S302 (step S303) and starts the transmittance calculation process. Next, the variance reduction unit 120B performs variance reduction processing to reduce the calculation load of the transmittance calculation process (step S304).

The flow of the variance reduction process performed by the variance reduction unit 120B will be described below with reference to FIG. 4. Here, in order to reduce the calculation load in small radiation field cases, the high importance condition is whether or not the particle passes through a high importance area determined based on the aperture shape of the multi-leaf collimator, and an example is shown in which an increase and decrease are performed on transport particles that are determined to pass outside the high importance area.

As shown in FIG. 4, first, a reference aperture area is determined based on the treatment plan information read in step S202 (step S401), and a high importance area is set based on the reference aperture area (step S402). A schematic diagram of the high importance area setting and light reduction is shown in FIG. 5.

Here, we will explain an example in which the reference aperture region is defined as a rectangle that circumscribes the MLC aperture shape.

When acquiring the aperture shape information of the MLC, it is necessary to first determine only the leaves that are being driven. Therefore, the x-direction position x _i,L of the i-th left leaf and the x-direction position x _i,R of the i-th right leaf are referenced, and compared with the standby position x _standby,L of the left leaf and the standby position x _standby,R of the right leaf, respectively, to determine the driven leaf, and calculate the y-direction length of the reference aperture area.

Next, calculation of the reference aperture area length in the x direction will be described. As shown in Fig. 5, the negative side position of the reference aperture area in the x direction through which the beam 505 passes is determined by the left leaf forming the MLC aperture shape 501, and the positive side position is determined by the right leaf. Therefore, the reference aperture area 502 in the x direction is defined by calculating the minimum position xmin _,L of the left leaf and the maximum position xmax _,R of the right leaf among the driven leaves.

Here, we show an example in which the reference opening region 502 is expanded by an arbitrary length δL to set a high importance region 503.

In X-ray therapy, the 50% dose end position of the dose distribution generated from each irradiation direction is basically determined by the collimator opening position. Therefore, if you want to emphasize the peripheral area of the dose distribution (50%-20% dose area) in the dose evaluation, you need to set an area that is an extension of the standard opening area as the high importance area.

Next, it is determined whether the transport ray passes through a high importance region (step S403). Here, an example is shown in which the passage of a high importance region is determined at the isocenter position.

In an X-ray treatment device, assuming that the beam from the X-ray source spreads out in a fan beam, if the source-to-isocenter distance is SAD and the source-to-collimator center distance is l, the high importance area at the isocenter position is the high importance area at the leaf position multiplied by SAD/l.

The variance reduction unit 120B reads the particle type information of the transport ray irradiated upstream of the MLC, and determines whether the ray passes through a high importance area based on the transport particle's progress vector ( _vx , _vy , _vz ) and position information ( _x0 , _y0 , _z0 ), under the assumption that the transport ray travels straight from the upstream of the MLC.

If it is determined in step S403 that the ray passes through the high importance region, the flow of Fig. 4 is terminated and the process returns to the flow of Fig. 3, where the transmittance calculation for each ray is performed (step S305). On the other hand, if it is determined that the ray does not pass through the high importance region, a weight wgeom that is higher than the initial weight _w0 of the ray is set as the weight of the _ray (step S404).

Next, a process of reducing the number of transport rays is performed so that the expected value of the weights of all transport rays is preserved (step S405). Therefore, for particles that do not pass through high importance regions, a uniform random number r is generated for each transport ray so that transport particles are reduced with a probability of _w0 / _wgeom , and particles that satisfy r> _w0 / _wgeom are eliminated.

This makes it possible to reduce the number of transport rays that pass outside of high-importance areas while maintaining the expected value of the weights of all transport rays. This allows computational resources to be concentrated in high-importance areas, making it possible to calculate transmittance efficiently.

Returning to FIG. 3, it is checked whether the transport ray has been eliminated by the variance reduction process (step S305). If the transport ray has been eliminated, the transport for this transport ray ends and the process moves to the processing for the next transport ray (step S309). On the other hand, if a transport ray exists, the path length of the MLC for each transport ray is calculated and the transmittance is calculated (step S306).

Returning to FIG. 2, the statistically weighted particle species data created in step S204 is read, and the dose distribution is calculated using the Monte Carlo method (step S205). Therefore, if there are many particles with small statistical weights in the particle species data read in step S205, the efficiency of the dose distribution calculation is significantly reduced.

Therefore, to improve the calculation efficiency of step S205, a step (step S307) is provided to reduce low-weight particles using the Russian roulette method, which is one of the variance reduction methods, as shown in Figure 3.

In the Russian roulette method, a cutoff value w is set, and for particles below _{w ,} uniform random numbers are used to eliminate particles with a probability of w/ _w . On the other hand, for surviving particles with statistical weights below _w , the cutoff value _w is set as the statistical weight.

Then, the coordinate values of the particles at the downstream position of the MLC are calculated, and then written and saved as particle species data downstream of the MLC (step S308). Then, it is determined whether there are any transport particles to be calculated next (step S309), and if so, the process returns to step S303, and if not, the flow of FIG. 3 ends.

If it is necessary to take into account the effects of the rotation of the treatment device (gantry angle information, irradiation ring angle information, couch angle information), it is possible to reflect this by performing coordinate conversion of the progress vector information and position information of the particle species data using a rotation matrix that represents the device rotation before writing the particle species data downstream of the MLC.

So far, we have explained how to set high importance areas and reduce variance using the example of discrete X-ray irradiation for each irradiation port.

Next, we will explain how to set high importance areas when the multi-leaf collimator shape is changed between control points without interrupting radiation irradiation, i.e., in the case of VMAT irradiation, in which irradiation is performed while continuously changing the MLC aperture shape.

In the VMAT irradiation conditions, an irradiation control point is set for each irradiation angle, and the leaf position and cumulative irradiation amount are recorded for each irradiation control point. Therefore, in the transport calculation of VMAT irradiation, the irradiation amount and the number of transported rays are determined for each irradiation control point from the cumulative irradiation amount, and the transport calculation is performed for each control point. Here, a method for setting a high importance area is explained using an example of MLC transport calculation between irradiation control point A at gantry angle θ _A and irradiation control point B at gantry angle θ _B.

Figure 6 shows an explanatory diagram of how to set the MLC opening shape and high importance area at control point A and control point B.

As shown in Figure 6, similar to the case of discrete X-ray irradiation, the reference aperture area 602 of control point A is represented by a rectangle circumscribing the MLC aperture shape 601 of control point A, and the reference aperture area 612 of control point B is represented by a rectangle circumscribing the MLC aperture shape 611 of control point B. When considering the MLC transport of particles irradiated between control points A and B, each leaf moves within the reference aperture areas of control points A and B, so the reference aperture area (= high importance area) between control points A and B is represented by the union of the two reference aperture areas.

This union can be set by taking the union of the reference opening shapes at each control point A and control point B and then expanding the specified area, or it can be set by expanding the reference opening shapes at each control point A and control point B by a specified area and then taking the union of the reference opening shapes expanded by the specified area, and it may be selected appropriately.

Figure 6 shows an example in which the reference opening area 622 between control points A and B is more simply represented as a rectangle circumscribing the two opening areas.

Next, similarly to the case of discrete irradiation of X-rays, a high importance region 621 is set based on the reference opening region. Here, an example is shown in which the reference opening region is expanded by an arbitrary length δL to create a high importance region. Here, the same expansion width δL is set in the y direction and the x direction, but different expansion widths δL _x and δL _y may be set in the x direction and the y direction, respectively.

So far, we have described an example in which a rectangle circumscribing the MLC opening shape is defined as the reference opening area, and a high importance area is set based on the reference opening area, but the method for setting the high importance area is not limited to this setting method as long as it is a method that uses the MLC opening shape.

Next, the procedure for setting a high importance region will be described by defining the MLC aperture shape itself as a reference aperture region and expanding the reference aperture region by arbitrary lengths δL _x and δL _y in the x and y directions. Here, the procedure is applied to VMAT irradiation, and an MLC transport calculation between irradiation control point A at gantry angle θ _A and irradiation control point B at gantry angle θ _B is taken as an example. A flow diagram of the procedure for setting a high importance region is shown in FIG. 7, and an explanatory diagram is shown in FIG. 8.

First, as shown in FIG. 7, the leaf positions of irradiation control point A and irradiation control point B are read (step S701). In FIG. 7, the leaf positions of control points A and B are represented by dashed and dotted lines.

Next, for the right leaf, the maximum leaf position xL _max,i at each of the control points A and B is defined as the widest leaf position, and for the left leaf position, the minimum leaf position xL _min,i at each of the control points A and B is defined as the widest leaf position, and the opening area defined by these leaf positions is defined as the reference opening area (step S702).

Next, a region expanded from the reference opening region by ±δL _y in the y direction and ±δL _x in the x direction is generated, and this is set as a high importance region.

First, an area obtained by expanding the reference opening area in the y direction is calculated using the expansion width ±δL _y (step S703). The minimum value of the left leaf position and the maximum value of the right leaf position of the leaves existing in the range of the expansion width ±δL _y for each leaf are calculated to form the reference opening area expanded in the y direction. The left leaf position lpos _i and the right leaf position rpos _i of leaf i of the reference opening area expanded in the Y direction are defined by the following equations (2) and (3).

Here, n _down,i and n _up,i represent the numbers of leaves forming the reference aperture region that exist within the range of −δL _y and +δL _y from the i-th leaf.

Next, in order to expand the reference opening region in the x direction, δL _x is subtracted from the left positions of all leaves forming the reference opening region expanded in the y direction, and δL _x is added to the right positions (step S704). Finally, the reference opening region expanded in this way is set as a high importance region (step S705). An example of the set high importance region is shown in FIG.

As shown in FIG. 8, it can be seen that the opening position 803 of the left leaf and the opening position 805 of the right leaf at control point A, and the opening position 804 of the left leaf and the opening position 806 of the right leaf at control point B are included in the high importance area between the left leaf 801, which forms the high importance area between control points A and B, and the right leaf 802, which forms the high importance area between control points A and B.

In addition, in this variance reduction process, there are calculation parameters such as the expansion width δL from the reference opening region, which is a setting parameter of the high importance region, and the weight w _geom of the particles irradiated outside the high importance region, etc. Therefore, the values of these parameters may be set by the operator using the GUI, or a parameter file in which the calculation parameters are recorded may be recorded in the storage unit 124, and each calculation parameter may be read during the variance reduction process.

It is also possible to display the extent of the enlargement width δL on a screen such as that shown in FIG. 5 or FIG. 10 described below.

So far, we have discussed methods to reduce the calculation cost far outside the irradiation field, but another factor that increases the calculation load in MLC transport calculations is the consideration of Compton photons.

Compton photons are photons generated within the MLC by Compton scattering. Compared to primary photons, they are scattered at larger angles and are more likely to pass through multiple MLC leaves. This requires that the MLC passing length be calculated using multiple leaves, which increases the calculation time. On the other hand, since the Compton scattering component mainly contributes to the low-dose range of 20% or less, its importance from the perspective of dose distribution verification is lower than that of primary photons.

Therefore, the high importance condition is whether the transport particle is a primary photon or a Compton photon, and the weight can be increased and particles can be reduced for transport particles that are determined to be Compton photons.

In other words, by introducing a similar method to the Compton scattering generation process, it is possible to reduce the number of Compton photons generated and the computational load of the MLC transport calculation code. Therefore, the fact that it is a primary photon is considered a high importance condition, and the process of reducing Compton photons is explained below.

An example of variance reduction processing for Compton photons is shown in Figure 9.

As shown in FIG. 9, first, it is determined whether the transport photon is a primary photon or a Compton photon (step S901). If it is determined to be a primary photon, the process proceeds directly to a step of calculating the transmittance of the light beam. On the other hand, if it is determined to be a Compton photon, the process proceeds to a process of reducing the number of transport photons.

Here, the reduction process of Compton photons will be described using the weight w _c0 of a Compton photon generated in the MLC by a certain primary photon (in the case of Compton photons, this weight w _c0 differs for each Compton photon) and the threshold weight w _compton of the Compton photon.

If the Compton photon weight w _c0 is equal to or greater than the threshold weight w _compton of the variance reduction process (w _c0 ≧ w _compton ), the particle is transported with the original weight w _c,j (Yes in step S902). On the other hand, if w _c0 < w _compton , the corresponding Compton particle is reduced (step S903) so that the expected value of the Compton photon weight is preserved, and then w _c0 = w _compton is set as the Compton photon weight (step S904).

This makes it possible to reduce Compton photons while maintaining the expected value of the weight of all transport rays.

It is possible to use the high importance condition of determining whether or not the photon is a Compton photon that is independent of the irradiation field as described above, and the high importance condition of the high importance area, which has a larger shortening effect because the smaller the irradiation field, the greater the number of particles outside the area (i.e., the number of particles to be reduced), so the smaller the irradiation field.

Figures 10(a) and (b) show a comparison of the fluence distribution calculation results downstream of the MLC with and without dispersion reduction processing when dispersion reduction processing is performed based on the MLC aperture shape and the generation process of transported photons.

As shown in Figure 10, it can be confirmed that the calculation results of Geant4, a general-purpose particle simulation code, are reproduced very well, regardless of whether or not the variance reduction method is used. This shows that by using the variance reduction process, it is possible to speed up particle transmittance calculations and perform high-speed dose distribution calculations without causing a significant change in calculation accuracy.

Next, we will explain the effects of this embodiment.

The dose distribution calculation system 113 of the present embodiment described above is a dose distribution calculation system that calculates the transmittance of a multi-leaf collimator for each transport particle and calculates a dose distribution using the transmittance. When calculating the transmittance, a high importance condition is set based on the aperture shape of the multi-leaf collimator or the particle generation process, and it is determined whether or not a transport particle satisfies the high importance condition. For transport particles that are determined not to satisfy the high importance condition, the weight of the transport particle is increased and the number of transport particles is reduced so that the expected value of the weight of all transport particles is preserved.

According to the present invention, since low-dose regions are no longer ignored, it is possible to provide highly accurate dose distribution calculation results quickly.

The high importance condition is whether or not the particle passes through a high importance area determined based on the aperture shape of the multi-leaf collimator. Increases and reductions are performed on transported particles that are determined to pass outside the high importance area. The smaller the radiation field, the greater the number of particles outside the area (i.e., particles to be reduced), and the greater the reduction effect. This makes it possible to greatly speed up transmittance calculations in small radiation field cases, for example.

Furthermore, a high-importance condition is whether the transport particle is a primary photon or a Compton photon, and by implementing increases and decreases for transport particles determined to be Compton photons, it becomes possible to take into account the effect of the dose in areas other than the irradiation field, thereby improving the accuracy of dose distribution calculations outside the irradiation field.

In addition, when the multi-leaf collimator shape is changed between control points without interrupting radiation irradiation, the high importance area is set by taking the union of the reference aperture shapes at each control point and then expanding the specified area, or when the multi-leaf collimator shape is changed between control points without interrupting radiation irradiation, the high importance area is set by expanding the reference aperture shape at each control point by a specified area and then summing the reference aperture shapes expanded by the specified area, making it possible to speed up transmittance calculations without reducing accuracy under complex conditions where the multi-leaf collimator shape is changed between control points without interrupting radiation irradiation.

Furthermore, by calculating the dose distribution during online adaptive treatment, highly accurate irradiation can be achieved in a short time, reducing the burden on patients compared to conventional methods.

In addition, by calculating the dose distribution when planning treatment, the burden on doctors before treatment begins can be reduced.

＜Other＞
The present invention is not limited to the above-mentioned embodiment, and various modifications and applications are possible. The above-mentioned embodiment has been described in detail to explain the present invention in an easily understandable manner, and the present invention is not necessarily limited to having all of the configurations described.

The present invention may take the following forms:

(1) A dose distribution calculation system that calculates the transmittance of a multi-leaf collimator for each transport particle and calculates a dose distribution using the transmittance, and when calculating the transmittance, sets a high importance condition based on the aperture shape of the multi-leaf collimator or the particle generation process, determines whether the transport particle satisfies the high importance condition, and for the transport particle that is determined not to satisfy the high importance condition, increases the weight of the transport particle and reduces the transport particle so that the expected value of the weight of all transport particles is preserved.

(2) In the dose distribution calculation system described in (1), the high importance condition is whether or not the transported particles pass through a high importance area determined based on the aperture shape of the multi-leaf collimator, and the increase and reduction are implemented for the transported particles that are determined to pass outside the high importance area.

(3) In the dose distribution calculation system described in (1), the high importance condition is whether the transport particle is a primary photon or a Compton photon, and the increase and decrease are performed on the transport particle determined to be a Compton photon.

(4) In the dose distribution calculation system described in (2), when the shape of the multi-leaf collimator is changed between control points without interrupting radiation irradiation, the high importance area is set by taking the union of the reference aperture shapes at each control point and then expanding the specified area.

(5) In the dose distribution calculation system described in (2), when the shape of the multi-leaf collimator is changed between control points without interrupting radiation irradiation, the high importance area is set by expanding the reference opening shape by a specified area at each control point, and then taking the union of the reference opening shapes expanded by the specified area.

(6) In the dose distribution calculation system described in any one of (1) to (5), the dose distribution is calculated during online adaptive treatment.

(7) In the dose distribution calculation system described in any one of (1) to (5), the dose distribution is calculated when calculating the dose distribution during treatment planning.

A...Irradiation target 100...X-ray irradiation device 101...Ring-type gantry 102...Irradiation nozzle 103...Bed 104...Radiation irradiation control device 105...Communication device 106...Storage device 107...Input device 110...X-ray irradiation system 111...Data server 111A...Treatment plan information 112...Treatment planning device 113...Dose distribution calculation system 120...MLC transport calculation unit 120A...Transmittance calculation unit 120B...Dispersion reduction unit 121...Dose distribution calculation unit 122...Display unit 123...Input unit 124...Storage unit 125...Particle species information database 501...MLC aperture shape 502...Reference aperture area 503...High importance area 504...Bi MLC leaf 505 viewed from beam irradiation direction...Beam 601...MLC aperture shape 602 of control point A...Reference aperture area 611 of control point A...MLC aperture shape 612 of control point B...Reference aperture area 621 of control point B...High importance area 622 between control points A and B...Reference aperture area 801 between control points A and B...Left leaf 802 forming the high importance area between control points A and B...Right leaf 803 forming the high importance area between control points A and B...Opening position 804 of left leaf at control point A...Opening position 805 of left leaf at control point B...Opening position 806 of right leaf at control point B

Claims (9)

A dose distribution calculation system that calculates a transmittance of a multi-leaf collimator for each transport particle and calculates a dose distribution using the transmittance,
When calculating the transmittance, a high importance condition is set based on an aperture shape of the multi-leaf collimator or a particle generation process;
determining whether the transport particle satisfies the high importance condition;
for the transport particles determined not to satisfy the high importance condition, increasing the weight of the transport particles and reducing the transport particles so that an expected value of the weight of all transport particles is preserved. 2. The dose distribution calculation system according to claim 1,
the high importance condition being whether or not the beam passes through a high importance region that is determined based on an aperture shape of the multi-leaf collimator;
and performing said increase and said decrease on said transport particles determined to pass outside said high importance region. 2. The dose distribution calculation system according to claim 1,
the high importance condition being whether the transport particle is a primary photon or a Compton photon;
performing said increase and said decrease on said transport particles determined to be said Compton photons. 3. The dose distribution calculation system according to claim 2,
When the shape of a multi-leaf collimator is changed between control points without interrupting radiation irradiation, the high importance area is set by taking the union of the reference aperture shapes at each control point and then expanding the union by a predetermined area. 3. The dose distribution calculation system according to claim 2,
A dose distribution calculation system in which, when the shape of a multi-leaf collimator is changed between control points without interrupting radiation irradiation, the high importance area is set by expanding a reference aperture shape by a specified area at each control point, and then setting the union of the reference aperture shapes expanded by the specified area. 2. The dose distribution calculation system according to claim 1,
A dose distribution calculation system for determining the dose distribution during online adaptive treatment. 2. The dose distribution calculation system according to claim 1,
A dose distribution calculation system that calculates the dose distribution when calculating the dose distribution at the time of treatment planning. A procedure for calculating the transmittance of a multi-leaf collimator for each transport particle;
A program for causing a calculation system to execute a procedure for calculating a dose distribution using the transmittance,
setting a high importance condition based on an aperture shape of the multi-leaf collimator or a particle generation process when calculating the transmittance;
determining whether the transport particle satisfies the high importance condition;
a program for increasing the weight of the transport particle determined not to satisfy the high importance condition and reducing the transport particle so that an expected value of the weight of all the transport particles is preserved. A dose distribution calculation method for calculating a transmittance of a multi-leaf collimator for each transport particle and calculating a dose distribution using the transmittance, comprising:
When calculating the transmittance, a high importance condition is set based on an aperture shape of the multi-leaf collimator or a particle generation process;
determining whether the transport particle satisfies the high importance condition;
for the transport particles determined not to satisfy the high importance condition, increasing the weight of the transport particles and reducing the transport particles so that an expected value of the weight of all transport particles is preserved.

Images