RU2840334C2 - Radioactive irradiation system and method of its control - Google Patents

Abstract

FIELD: medicine.

SUBSTANCE: invention relates to a radioactive irradiation system and a method for controlling a radioactive irradiation system and can be used in radiation therapy. System comprises a beam irradiation device generating a therapeutic beam and emitting a therapeutic beam to an irradiated body to form an irradiated area, a treatment plan module which simulates and calculates a dose in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area, which generates a treatment plan. Treatment plan module is configured to filter elements that affect the simulation results and calculation of neutrons and photons in the application scenes of the radioactive irradiation system by taking into account the weight proportions of the plurality of elements in the human body and the intensities of the plurality of elements reacting with neutrons and photons, so that during simulation only filtered elements are simulated. System also comprises a control module which controls the radiation of the beam irradiation device in accordance with the treatment plan.

EFFECT: shorter time for calculating radiation dose distribution due to increased speed of calculating the radiation exposure plan module.

8 cl, 8 dwg, 4 tbl

Description

AREA OF TECHNOLOGY

[0001] One aspect of the invention relates to a radioactive beam irradiation system, and another aspect of the invention relates to a method for controlling the radioactive beam irradiation system.

LEVEL OF TECHNOLOGY

[0002] With the development of nuclear technology, radiotherapy such as linear accelerator, electron beam or the like, has become one of the main means of cancer treatment. However, traditional photon or electron therapy is limited by the physical conditions of the radioactive rays themselves, and thus will also harm a large number of normal tissues along the path of the beam when destroying tumor cells. In addition, due to the different sensitivity of tumor cells to radioactive rays, traditional radiation therapy usually has a weak curative effect on radioresistant malignant tumors (such as glioblastoma multiforme and melanoma) with radioresistance.

[0003] In order to reduce the radiation damage to normal tissues around tumors, the concept of targeted therapy in chemotherapy is applied to radiotherapy. As for tumor cells with high radioresistance, radiation sources with high relative biological effectiveness (RBE), such as proton therapy, heavy particle therapy, neutron capture therapy, etc., are also being actively developed. Here, neutron capture therapy combines the above two concepts, such as boron neutron capture therapy (BNCT), provides a better choice for cancer treatment than traditional radioactive rays by specifically aggregating boron-containing drugs in tumor cells in combination with precise beam regulation and control.

[0004] In BNCT, two highly charged particles ^{of 4} He and ⁷ Li are produced by using the characteristic of a boron-containing drug ( ¹⁰ B) having a high capture section for a thermal neutron, and by means of neutron capture of ¹⁰ B (n, α) ⁷ Li and a nuclear fission reaction, and these two particles have a total range approximately equivalent to the size of a cell, so that radiation damage to the body can be limited to the cell level, and when the boron-containing drug is selectively aggregated in the tumor cell, the purpose of locally killing the tumor cell can be achieved under the condition of not inducing too much damage to normal tissues, in combination with a suitable neutron radiation source.

[0005] In order for radiation particles to kill cancer cells as much as possible and reduce damage to normal cells, CT or PET scanning is usually performed before treating a patient, the material information of human body tissues is obtained from the scanning result, a calculation model is established according to the material information and the radiation source, the process of transporting radiation particles in the human body is simulated, the distribution of the dose of radiation particles in the human body is finally obtained, and then a scheme with the optimal dose distribution to the patient is selected as the treatment regimen for the patient.

[0006] At present, the dose calculation module in the radiation therapy plan system obtains the dose distribution mainly by simulating radiation particles using the Monte Carlo method. The photon and electron motion processes need to be simulated for conventional radiation therapy, and the neutron and photon motion processes need to be simulated for radiation therapy. At present, the Monte Carlo method is the most accurate dose calculation method, but it takes a long time to calculate and has a large memory consumption.

[0007] At present, the most commonly used general Monte Carlo programs for calculation in radiation therapy are MCNP and Geant 4, here MCNP is primarily used for reactor design and calculation, Geant 4 is primarily used for high-energy physical calculation. These programs do not take into account the physical scenes of radiation therapy at the beginning of their development, and therefore do not perform special optimization with respect to the calculation area of radiation therapy. The preparation of a radiation therapy plan usually needs to be completed within a certain time, for example, radiation therapy requires the radiation therapy plan system to give a treatment plan within one hour, here the dose calculation process takes up most of the time formulated by the radiation therapy plan. Therefore, it is necessary to optimize the dose calculation method to shorten the time specified in the radiation therapy plan.

ESSENCE OF THE INVENTION

[0008] In order to overcome the disadvantages of the related art, a first aspect of the invention provides a radioactive beam irradiation system comprising a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a treatment beam and irradiates an irradiated body with the treatment beam to form an irradiated area. The treatment plan module performs modeling and dose calculation in accordance with the parameters of the treatment beam and medical imaging data of the irradiated area and generates a treatment plan, filters elements affecting the results of modeling and calculation of neutrons and photons in scenes of application of the radioactive beam irradiation system by taking into account the weight proportions of a plurality of elements in the human body and the intensities of reactions of a plurality of elements with neutrons and photons, so that only the filtered elements are modeled during the modeling. The control module controls the radiation of the beam irradiation device in accordance with the treatment plan.

[0009] In other embodiments of the invention, the elements filtered by the treatment plan module may be selected from one or more of H, He, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, Ar, K and Ca.

[0010] More specifically, the elements filtered by the treatment plan module may be H, He, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, Ar, K, and Ca.

[0011] In another embodiment, the treatment plan module may remove databases corresponding to temperatures other than 294 K and 0 K from the neutron cross section database corresponding to each of the elements.

[0012] Preferably, the radioactive beam irradiation system may be a neutron capture therapy system, the beam irradiation device may comprise a neutron generation device, a beam forming body and a treatment table. The neutron generation device includes an accelerator and a target, the accelerator accelerates charged particles to generate a beam of charged particles that interacts with the target to generate a neutron beam. The beam forming body is configured to regulate the neutron beam generated by the neutron generation device to ensure a predetermined beam quality, wherein the neutron beam generated by the neutron generation device is radiated onto the irradiated body on the treatment table by means of the beam forming body.

[0013] The second aspect of the invention relates to a method for controlling a radioactive beam irradiation system, wherein the beam irradiation system includes a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a therapeutic beam and irradiates an irradiated body with the therapeutic beam to form an irradiated area. The treatment plan module performs modeling and dose calculation in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area, and generates a treatment plan. The control module controls the radiation of the beam irradiation device in accordance with the treatment plan. The method for controlling the radioactive beam irradiation system includes the following operations. Elements that influence the modeling and calculation results of neutrons and photons in scenes of application of the radioactive beam irradiation system are filtered by taking into account the weight proportions of a plurality of elements in the human body and the intensities of the reaction of the plurality with neutrons and photons, so that only the filtered elements are modeled during the modeling.

[0014] In other embodiments of the invention, the elements filtered by the treatment plan module may be selected from one or more of H, He, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, Ar, K and Ca.

[0015] In particular, the elements filtered by the treatment plan module may be H, He, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, Ar, K and Ca

[0016] In another embodiment, the method of controlling a radioactive beam irradiation system may further include the following operations. The treatment plan module deletes databases corresponding to temperatures other than 294 K and 0 K from the neutron cross-section database corresponding to each of the elements.

[0017] Preferably, the radioactive beam irradiation system may be a neutron capture therapy system, the beam irradiation device may comprise a neutron generation device, a beam forming body and a treatment table. The neutron generation device includes an accelerator and a target, the accelerator accelerates charged particles to generate a neutron beam that interacts with the target to generate a neutron beam. The beam forming body is configured to regulate the neutron beam generated by the neutron generation device to ensure a predetermined beam quality, wherein the neutron beam generated by the neutron generation device is radiated onto the irradiated body on the treatment table by means of the beam forming body.

[0018] According to the radioactive ray irradiation system and the radioactive ray irradiation system control method described in the embodiments of the invention, elements influencing the simulation and calculation results of neutrons and photons in application scenes of the radioactive ray irradiation system are filtered by considering the weight proportions of a plurality of elements in the human body and the reaction intensities of the plurality of elements with neutrons and photons, so that only the filtered elements are simulated during the simulation. In this way, the calculation speed of the radiation exposure plan module can be significantly increased and the calculation time can be shortened.

[0019] In a third aspect of the invention, a radioactive irradiation system is provided, including a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a therapeutic beam and irradiates the therapeutic beam onto an irradiated body to form an irradiated area. The treatment plan module performs modeling and dose calculation according to the parameters of the therapeutic beam and medical imaging data of the irradiated area and generates a treatment plan, wherein the tasks of modeling various source particles are assigned to various processes or flows, and summarization is performed after completing the calculation task of each of the processes or flows to obtain the final calculation result. The control module controls the radiation of the beam irradiation device according to the treatment plan.

[0020] Preferably, the radioactive beam irradiation system may be a neutron capture therapy system, the beam irradiation device may comprise a neutron generation device, a beam forming body and a treatment table. The neutron generation device includes an accelerator and a target, the accelerator accelerates charged particles to generate a beam of charged particles that interacts with the target to generate a neutron beam. The beam forming body is configured to regulate the neutron beam generated by the neutron generation device to ensure a predetermined beam quality, wherein the neutron beam generated by the neutron generation device is radiated onto the irradiated body on the treatment table by means of the beam forming body.

[0021] In an embodiment of the invention, particles can be simulated by process parallelism and thread parallelism of the central processing unit (CPU) and graphics processing unit (GPU) acceleration, where each of the process parallelism and the thread parallelism implements parallel computing by a multi-core processor, wherein the process parallelism and the thread parallelism calculation process of the CPU is as follows. First, the system obtains the number of processes or threads to obtain a numerical value n; then the system equally divides the particles necessary for the simulation into n parts; then each of the threads or processes separately simulates and counts each part of the particles; and finally, the system performs the counting of the amount obtained by each of the processes or threads to obtain the final dose.

[0022] In an embodiment, the GPU acceleration can be realized by parallel computing of a plurality of GPU processors, and the process of simulating and computing the GPU acceleration is as follows. First, the system transfers a random number, cross-sectional data, or the like from the processor memory to the video memory of the GPU, and then each of the plurality of GPU processors simulates, calculates, and counts one particle and counts the counting result into a global count; then the system determines whether there are particles that have not yet been simulated, transfers the counting result from the GPU memory to the CPU memory in response to the absence of particles that have not yet been simulated, or returns to the previous operation in response to the fact that particles have not yet been simulated to continue simulating and counting particles that have not yet been simulated until the simulation of all particles is completed.

[0023] A fourth aspect of the invention relates to a method for controlling a radioactive beam irradiation system, wherein the beam irradiation system includes a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a therapeutic beam and irradiates the therapeutic beam onto an irradiated body to form an irradiated area. The treatment plan module performs modeling and dose calculation in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area and generates a treatment plan. The control module controls the radiation of the beam irradiation device in accordance with the treatment plan. The method for controlling a radioactive beam irradiation system includes the following operations. The tasks of modeling different source particles are assigned to different processes or flows, and summarizing is performed after completing the calculation task of each of the processes or flows to obtain the final calculation result.

[0024] In an embodiment of the invention, particles may be simulated by process parallelism and CPU thread parallelism and GPU acceleration.

[0025] Preferably, each of the process parallelism and the thread parallelism can implement parallel computing by means of a multi-core CPU.

[0026] In an embodiment of the invention, the process of calculating the process parallelism and the CPU thread parallelism may be as follows. First, the system obtains the number of processes or threads to obtain a numerical value n; then the system equally divides the particles needed for modeling into n parts; then each of the threads or processes separately models and counts each part of the particles; and finally, the system performs a count of the amount received by each of the processes or threads to obtain the final dose.

[0027] Preferably, GPU acceleration may be implemented by parallel computing of multiple GPU processors.

[0028] The process of simulating and calculating the acceleration of the GPU may be as follows. First, the system transfers a random number, cross-section data, or the like from the processor memory to the video memory of the GPU, then each of the plurality of processors of the GPU simulates, calculates, and counts one particle and counts the counting result in the global count; then the system determines whether there are particles that have not yet been simulated, transfers the counting result from the GPU memory to the CPU memory in response to the absence of particles that have not yet been simulated, or returns to the previous operation in response to the presence of particles that have not yet been simulated to continue simulating and counting particles that have not yet been simulated until the simulation of all particles is completed.

[0029] According to the radioactive beam irradiation system and the radioactive beam irradiation system control method described in the embodiments of the invention, the tasks of simulating different source particles are assigned to different processes or flows, and a summary is performed after the calculation task of each of the processes or flows is completed to obtain the final calculation result. Therefore, the calculation speed can be significantly increased and the calculation time can be shortened.

[0030] A fifth aspect of the invention relates to a radioactive beam irradiation system comprising a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a therapeutic beam and radiates the therapeutic beam onto an irradiated body to form an irradiated area. The treatment plan module performs modeling and dose calculation in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area and generates a treatment plan, wherein the photon modeling is terminated in response to the fact that the half-absorption thickness of the photon is less than or equal to a first preset value. The control module controls the radiation of the beam irradiation device in accordance with the treatment plan.

[0031] Preferably, the radioactive beam irradiation system may be a neutron capture therapy system, the beam irradiation device may comprise a neutron generation device, a beam forming body and a treatment table. The neutron generation device includes an accelerator and a target, the accelerator accelerates charged particles to generate a beam of charged particles that interacts with the target to generate a neutron beam. The beam forming body is configured to regulate the neutron beam generated by the neutron generation device to ensure a specified beam quality, wherein the neutron beam generated by the neutron generation device is radiated onto the irradiated body on the treatment table by means of the beam forming body.

[0032] In one embodiment of the invention, the treatment plan module may calculate the photon half-absorption thickness t using formula (1-1):

[0033] where - the linear attenuation coefficient of a photon, determined by the material through which the photon passes and the energy of the photon.

[0034] Here, the first preset value may be a cell size, preferably the first preset value may be 0.2 mm, wherein when the photon energy is less than or equal to the second preset value, the half-absorption thickness of the photon corresponding to the photon energy is less than or equal to the first preset value, wherein the second preset value is 10 keV.

[0035] A sixth aspect of the invention relates to a method for controlling a radioactive beam irradiation system, wherein the beam irradiation system includes a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a therapeutic beam and radiates the therapeutic beam onto an irradiated body to form an irradiated area. The treatment plan module performs modeling and dose calculation in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area, and generates a treatment plan. The control module controls the radiation of the beam irradiation device in accordance with the treatment plan. The method for controlling the radioactive beam irradiation system includes the following operations. Photon modeling is terminated in response to the fact that the photon absorption half-width is less than or equal to a first preset value.

[0036] In one embodiment of the invention, the treatment plan module may calculate the photon half-absorption thickness t using formula (1-1):

[0037] where - the linear attenuation coefficient of a photon, determined by the material through which the photon passes and the energy of the photon.

[0038] Here the first preset value can be the cell size.

[0039] Preferably, the first predetermined value may be 0.2 mm.

[0040] When the photon energy is less than or equal to the second preset value, the half-absorption thickness of the photon corresponding to the photon energy is less than or equal to the first preset value.

[0041] Preferably, the second preset value may be 10 keV.

[0042] According to the radioactive beam irradiation system and the radioactive beam irradiation system control method described in the embodiments of the invention, the photon simulation is stopped in response to the fact that the half-absorption thickness of the photon is less than or equal to a first preset value. In this way, the calculation time can be shortened while ensuring the calculation accuracy.

[0043] A seventh aspect of the invention relates to a radioactive ray irradiation system comprising a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a therapeutic beam and irradiates the therapeutic beam onto an irradiated body to form an irradiated area. The treatment plan module performs modeling and dose calculation in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area and generates a treatment plan, wherein the treatment plan module models particles using dispersion reduction. The control module controls the radiation of the beam irradiation device in accordance with the treatment plan.

[0044] In this case, the dispersion reduction may include absorption by weight reduction, weight window play and splitting bet. Preferably, the radioactive beam irradiation system may be a neutron capture therapy system, the beam irradiation device may comprise a neutron generation device, a beam forming body and a treatment table. The neutron generation device includes an accelerator and a target, the accelerator accelerates charged particles to generate a neutron beam that interacts with the target to generate a neutron beam. The neutron generation device includes an accelerator and a target, the accelerator accelerates charged particles to generate a neutron beam that interacts with the target to generate a neutron beam. The beam-forming body is designed with the possibility of regulating the neutron beam generated by the neutron generation device to ensure a preset beam quality, wherein the neutron beam generated by the neutron generation device is radiated onto the irradiated body on the treatment table by means of the beam-forming body.

[0045] An eighth aspect of the invention relates to a method for controlling a radioactive beam irradiation system, wherein the beam irradiation system includes a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a therapeutic beam and irradiates the therapeutic beam onto an irradiated body to form an irradiated area. The treatment plan module performs modeling and dose calculation in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area, and generates a treatment plan. The control module controls the radiation of the beam irradiation device in accordance with the treatment plan. The method for controlling a radioactive beam irradiation system includes the following operations. The treatment plan module models particles using dispersion reduction.

[0046] In this case, the variance reduction may include absorption by weight reduction, a weight window game, and a split bet. The particle simulation operation using the variance reduction may include the following operations S1 to S9. In operation S1, an initial particle is obtained. In operation S2, it is determined whether the particle collides with a lattice element, and operations S3 and S4 are successively performed in the case of a collision, otherwise operations S5 and S6 are successively performed. In operation S3, absorption by weight reduction is performed. In operation S4, it is determined whether the weight is less than the weight window, and operation S7 is performed in response to the fact that the weight is less than the weight window, otherwise the process returns to operation S2. In operation S5, a split bet is performed. In operation S6, it is determined whether the bet is made, and operation S7 is performed in response to the bet being made, otherwise the process returns to operation S2. In operation S7, it is determined whether there is a dead bet, and operation S8 is performed in response to the presence of a dead bet, otherwise operation S9 is performed, after which the process returns to operation S2. In operation S8, it is determined whether the processing of particles is completed, and the process is terminated in response to the processing of particles being completed, otherwise the process returns to operation S1. In operation S9, the weight is divided by the probability of a dead bet.

[0047] In this case, the splitting bet may include the following operations S1 through S9. In operation S1, the importance of the cell size of each cell is calculated and recorded. In operation S2, particles are obtained. In operation S3, a weight window check is performed on the particle and an operation is performed. In operation S4, the importance of the grid space I _n and I _n+1 is calculated before and after the particle overlaps the grid boundary. In operation S5, a comparison is performed to determine whether I _n is greater than I _n+1 , and operation S6 is performed, after which the process returns to operation S3 in response to the fact that I _n is not greater than I _n+1 , otherwise operation S7 is performed. In operation S6, the particle is split, and the weight of the particle is decreased. In operation S7, a determination is made whether there is a dead bet for the particle, and operation S9 is performed in response to the determination that there is a dead bet for the particle, otherwise operation S8 is performed, after which the process returns to operation S3. In operation S8, the weight of the particle is increased. In operation S9, it is determined whether the particle simulation is completed, and the process returns to operation S2 in response to a determination that the particle simulation is not completed, otherwise the process is terminated.

[0048] In this case, the absorption by weight reduction may include the following operations S1-S7. In operation S1, particles are obtained. In operation S2, it is determined whether the particle collides with the lattice element, and operation S3 is performed in the case of a collision, otherwise the process returns to operation S1. In operation S3, the weight is multiplied by the probability of occurrence of scattering. In operation S4, it is determined whether the weight of the particle is less than the smallest weight, and operation S5 is performed in response to the fact that the weight is less than the smallest weight, otherwise the process returns to operation S2. In operation S5, it is determined whether there is a dead bet for the particle, and operation S7 is performed in response to the presence of a dead bet for the particle, otherwise operation S6 is performed, after which the process returns to operation S2. In operation S6, the weight is divided by the probability of a dead bet. In operation S7, it is determined whether the modeling of the particles is complete, and the process returns to operation S1 in response to determining that the modeling of the particles is not complete, otherwise the process is terminated.

[0049] In this case, the weight window game may include the following operations S1-S6. In operation S1, the movement of particles is simulated. In operation S2, it is determined whether the weight of the particle falls within the range of the weight window, and the process returns to operation S1 in response to the weight of the particle falling within the range of the weight window, otherwise, operation S3 is performed. In operation S3, it is determined whether the weight is greater than the weight window, operation S4 is performed, after which the process returns to operation S1 in response to the fact that the weight is greater than the weight window, otherwise, operation S5 is performed. In operation S4, the particle is split and the weight is reduced. In operation S5, it is determined whether there is a dead bet, and the process is terminated in response to the presence of a dead bet, otherwise, operation S6 is performed, after which the process returns to operation S1. In operation S6, the weight of the particle is increased.

[0050] In one embodiment of the invention, the importance of the grid space can be obtained by solving the accompanying transport equation, which is as follows:

[0051] where - accompanying flow, - accompanying source, speed of particle movement, - direction of particle movement, - the cross section of the reaction of a particle colliding with a substance, - scattering cross-section, r - particle position, E - particle energy, t - time.

[0052] In particular, a particle can move out of a grid with grid space importance into the grid with the importance of grid space in this case when given the particle splits into m particles, the weight of each particle decreases to 1/t of the original weight; when the betting technique is performed on the particle, is set, a random number x is selected between 0 and 1, and when x is less than P, the particle is saved and the weight is multiplied by 1/P, otherwise a dead rate is provided for the particle and the particle simulation is terminated.

[0053] In other embodiments of the invention, the weight of a particle may have an upper limit of 10 and a lower limit of 0.25, wherein when the weight w of the particle is greater than 10, the integer part of w is set equal to w1, the decimal part of w is set equal to w2, a random number x is selected between 0 and 1, w = w1 + 1 is set when x is less than w2, and w = w1 is set when x is greater than w2. The particle is then split into w particles for simulation until the weight of each of the particles decreases to 1.

[0054] According to the radioactive beam irradiation system and the method for controlling the radioactive beam irradiation system described in the embodiments of the invention, particles are modeled using dispersion reduction. Thus, the calculation time can be reduced while ensuring the calculation accuracy.

[0055] A ninth aspect of the invention relates to a radioactive beam irradiation system comprising a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a therapeutic beam and radiates the therapeutic beam onto an irradiated body to form an irradiated area. The treatment plan module performs modeling and dose calculation in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area and generates a treatment plan, wherein the treatment plan module performs modeling and calculation using non-uniform grids. The control module controls the radiation of the beam irradiation device in accordance with the treatment plan.

[0056] In this case, the cell size of the important region may be set by the treatment plan module to be smaller than the cell size of the unimportant region.

[0057] In this case, the grid size of the area where the tumor is located can be set by the treatment plan module to be less than or equal to 0.4 mm.

[0058] In this case, the grid size of the region in which each of blood, air and bone is located can be set by the treatment plan module to be greater than or equal to 1.6 mm.

[0059] In this case, the mesh size of the area where the normal muscle is located can be set by the treatment plan module to be larger than 0.8 mm and smaller than 1.6 mm.

[0060] Preferably, the radioactive beam irradiation system may be a neutron capture therapy system, the beam irradiation device may comprise a neutron generation device, a beam forming body and a treatment table. The neutron generation device includes an accelerator and a target, wherein the accelerator accelerates charged particles to generate a neutron beam that interacts with the target to generate a neutron beam. The beam forming body is configured to regulate the neutron beam generated by the neutron generation device to ensure a specified beam quality, wherein the neutron beam generated by the neutron generation device is radiated onto the irradiated body on the treatment table by means of the beam forming body.

[0061] The tenth aspect of the invention relates to a method for controlling a radioactive beam irradiation system, wherein the beam irradiation system includes a beam irradiation device, a treatment plan module and a control module. The beam irradiation device generates a therapeutic beam and irradiates the therapeutic beam onto an irradiated body to form an irradiated area. The treatment plan module performs modeling and dose calculation in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area, and generates a treatment plan. The control module controls the radiation of the beam irradiation device in accordance with the treatment plan. The method for controlling a radioactive beam irradiation system includes the following operations. The treatment plan module performs modeling and calculation using non-uniform grids.

[0062] In another embodiment of the invention, the method for controlling the radiation irradiation system may further include the following operations. The treatment plan module sets the cell size of the important region to be smaller than the cell size of the unimportant region.

[0063] The method for controlling a radioactive irradiation system may further include the following operations. The treatment plan module sets the grid size of the region in which the tumor is located to be less than or equal to 0.4 mm.

[0064] The method for controlling a radioactive ray irradiation system may further include the following operations. The treatment plan module sets the grid size of the region where each of blood, air, and bone is located to be greater than or equal to 1.6 mm, and sets the grid size of the region where normal muscle is located to be greater than 0.8 mm and less than 1.6 mm.

[0065] According to the radioactive beam irradiation system and the radioactive beam irradiation system control method described in the embodiments of the invention, the modeling and calculation are performed using non-uniform grids. Therefore, the calculation accuracy of the important region can be improved without significantly increasing the calculation time, and the calculation time is reduced in the case where the calculation accuracy of the unimportant region is achieved.

BRIEF DESCRIPTION OF DRAWINGS

[0066] Fig. 1 is a schematic modular diagram of a BNCT system according to an embodiment of the invention.

[0067] Fig. 2 is a schematic block diagram of a BNCT system according to an embodiment of the invention.

[0068] Fig. 3 is a block diagram of the computation of process parallelism and CPU thread parallelism in accordance with an embodiment of the invention.

[0069] Fig. 4 is a block diagram of GPU acceleration modeling and computation in accordance with an embodiment of the invention.

[0070] Fig. 5 is a block diagram of particle modeling using a dispersion reduction technique in accordance with an embodiment of the invention.

[0071] Fig. 6 is a block diagram of a split bet in accordance with an embodiment of the invention.

[0072] Fig. 7 is a block diagram of weight loss absorption in accordance with an embodiment of the invention.

[0073] Fig. 8 is a block diagram of a weight window game according to an embodiment of the invention.

IMPLEMENTATION OF THE INVENTION

[0074] Embodiments of the invention will be further described in detail below with reference to the drawings to enable a person skilled in the art to implement these embodiments with reference to the text of the description.

[0075] The application of neutron capture therapy as an effective cancer treatment has been gradually increasing in recent years, with BNCT being the most frequently observed, and the neutrons supplied in BNCT can be supplied by a nuclear reactor or accelerator. In BNCT, two highly charged particles ^{of 4} He and ⁷ Li are produced by using the characteristic of a boron-containing drug ( ¹⁰ B) having a high capture cross section for a thermal neutron, and by capturing a neutron of ¹⁰ B (n, α) ⁷ Li and undergoing a nuclear fission reaction. These two highly charged particles have an average energy of about 2.33 MeV and have the characteristics of high linear energy transfer (LET) and short range. LET and the range of α particles are 150 keV/μm and 8 μm, respectively, LET and the range of the highly charged particle ⁷ Li are 175 keV/μm and 5 μm, respectively, and the two particles have a total range approximately equivalent to the cell size, so that the radiation damage to the body can be limited to the cellular level. When the boron-containing drug is selectively aggregated in the tumor cell, the goal of precisely killing the tumor cell can be achieved under the condition of not causing too much damage to normal tissues in combination with a suitable neutron radiation source.

[0076] Regardless of whether the BNCT neutron source is from a nuclear reactor or a nuclear reaction of charged particles with a target, a mixed radiation field is created, that is, the beam contains neutrons and photons from low energy to high energy. For BNCT of a tumor in a deep position, the greater the content of other radiation beams except for epithermal neutrons, the greater the proportion of non-selective dose deposition in normal tissues, so the radiation induction of excess dose deposition should be minimized. In order to better understand the neutron dose distribution in the human body, in addition to the air beam quality factor, a human head tissue prosthesis is used to calculate the dose distribution in embodiments of the invention, and the beam quality factor of the prosthesis is used as a reference for creating a neutron beam.

[0077] The International Atomic Energy Agency (IAEA) makes five recommendations on the air beam quality factor of the neutron source used in clinical BNCT. The five recommendations allow the advantages and disadvantages of different neutron sources to be compared and serve as a reference for the selection of the neutron generation path and the design of the beam shaping body. The five recommendations are as follows:

[0078] Epithermal neutron flux >1 x 10 ⁹ n/cm ² s

[0079] Fast neutron contamination <2 x 10 ^-13 Gy- ^cm2 /n

[0080] Photon pollution <2 x 10 ^-13 Gy- ^cm2 /n

[0081] Thermal to epithermal neutron flux ratio <0.05

[0082] Ratio of epithermal neutron current to flux >0.7

[0083] Note: An epithermal neutron has an energy region between 0.5 eV and 40 keV, a thermal neutron has an energy region less than 0.5 eV, and a fast neutron has an energy region greater than 40 keV.

[0084] As shown in Fig. 1, the radioactive ray irradiation system in this embodiment of the invention is a BNCT system 100 and includes a neutron beam irradiation device 10, a treatment plan module 20 and a control module 30. The neutron beam irradiation device 10 generates a therapeutic beam of N neutrons and irradiates the therapeutic beam of N neutrons onto an irradiated body 200 to form an irradiated region. The treatment plan module 20 performs modeling and dose calculation according to the parameters of the therapeutic beam of N neutrons generated by the neutron beam irradiation device 10 and the medical image data of the irradiated region, and generates a treatment plan that determines the position of the irradiated region relative to the neutron beam irradiation device 10 during the irradiation treatment and the corresponding irradiation time. After the irradiated body 200 is positioned in accordance with the position determined by the treatment plan, the treatment can be started, the control unit 30 outputs the current treatment plan corresponding to the irradiated body 200 from the treatment plan unit 20, and controls the radiation of the neutron beam irradiation device 10 in accordance with the treatment plan. The control unit 30 can also receive other data information, such as data of the neutron beam irradiation device 10, data of the irradiated body 200, or the like.

[0085] As shown in Fig. 2, in this embodiment of the invention, the neutron beam irradiation device 10 comprises a neutron generation device 11, a beam forming body 12, a collimator 13 and a treatment table 14. The neutron generation device 11 comprises an accelerator 111 and a target T, the accelerator 111 accelerates charged particles (such as protons, deuterium nuclei, etc.) to generate a charged particle beam P, such as a proton beam, wherein the charged particle beam P is radiated onto the target T and interacts with the target T to generate a neutron beam (neutron beam) N. Preferably, the target T is a metal target. The appropriate nuclear reaction is selected according to characteristics such as the desired neutron yield and energy, available energies of accelerated charged particles, current, physical and chemical properties of the metal target, etc. Nuclear reactions commonly discussed include ^7Li (p,n) ^7Be and ^9Be (p,n)9B, both of which are endothermic reactions and have energy thresholds of 1.881 MeV and 2.055 MeV, respectively. The ideal neutron source for BNCT is the epithermal neutron at the keV energy level, whereby, theoretically, when protons with energies just above the threshold are used to bombard a lithium metal target, relatively low energy neutrons for clinical use can be generated without slowing down the treatment too much. However, the proton reaction cross sections of lithium (Li) and beryllium (Be) metal targets with a threshold energy are not high, so protons with higher energies are usually selected to initiate a nuclear reaction in order to generate a sufficiently large neutron flux. An ideal target should have a high neutron yield, the distribution of the generated neutron energy is close to the energy region of epithermal neutrons (this will be described in detail below), there is no too strong penetrating radiation, and there are characteristics such as safety, low cost, ease of operation, resistance to high temperature, or the like. However, nuclear reactions that meet all the requirements cannot actually be found, and a Li metal target is used in the embodiments of the invention. However, it is well known to a person skilled in the art that the target T can also be made of a metal material other than Li and Be, for example, the target T can be formed by tantalum (Ta), tungsten (W), etc. The target T may be a round plate or other solid shape, or may use a liquid (liquid metal). The accelerator 111 may be a linear accelerator, cyclotron, synchrotron, synchrocyclotron, and the neutron generation device 11 may also be a nuclear reactor without using an accelerator and a target. Regardless of whether the neutron source of BNCT is from a nuclear reactor or a nuclear reaction of charged particles of the accelerator with the target, in fact, a mixed radiation field is generated, that is, the beam contains neutrons and photons from low energy to high energy. For BNCT of a tumor in a deep position, the larger the content of other radiation beams except epithermal neutrons, the greater the proportion of non-selective dose deposition in normal tissues, so the radiation induction of excess dose deposition should be reduced as much as possible. At the same time, too many types of radiation rays must be prevented from reaching the normal tissues of the irradiated body, which also causes excessive dose deposition.

[0086] The beam of N neutrons generated by the generating device 11 is radiated onto the irradiated body 200 on the treatment table 14 by successively passing through the beam forming body 12 and the collimator 13. The beam forming body 12 is capable of adjusting the quality of the beam of N neutrons generated by the neutron generating device 11, and the collimator 13 collides the beam of N neutrons so that the beam of N neutrons has high guidance characteristics during treatment. It should be understood that the invention may not have a collimator, and the beam is directly radiated onto the irradiated body 200 on the treatment table 14 after exiting the beam forming body 12.

[0087] The beam-forming body 12 further comprises a reflector 121, a moderator 122, a thermal neutron absorber 123, a radiation screen 124 and a beam outlet 125. The neutrons generated by the neutron generation device 11 have large energy spectra, the content of other types of neutrons and photons must be reduced as much as possible, with the exception of epithermal neutrons that meet the treatment requirements in order to avoid injury to the operator or the irradiated body. Therefore, the neutrons emitted by the neutron generation device 10 must pass through the moderator 22 in order to adjust its fast neutron energy (>40 keV) to the epithermal neutron energy region (from 0.5 eV to 40 keV) and to reduce thermal neutrons (<0.5 eV) as much as possible. The moderator 22 is made of a material having a large cross-section interacting with fast neutrons and a small cross-section interacting with epithermal neutrons. In a preferred embodiment of the invention, the moderator 122 is made of at least one of D ₂ O, AlF ₃ , Fluenal™, CaF ₂ , Li ₂ CO ₂₃ , MgF ₂ or Al ₂ O ₃ . The reflector 121 surrounds the moderator 122 and reflects neutrons scattered around when passing the moderator 122 back into the neutron beam N to improve the neutron utilization factor, and is made of a material having a high neutron reflectivity. In a preferred embodiment of the invention, the reflector 121 is made of at least one of Pb or Ni, the thermal neutron absorber 123 is located in the rear part of the moderator 122 and is made of a material having a large cross-section for interaction with thermal neutrons. In a preferred embodiment, the thermal neutron absorber 123 is made of Li-6, wherein the thermal neutron absorber 123 absorbs thermal neutrons passing through the moderator 122 to reduce the thermal neutron content in the N neutron beam, preventing the induction of an excessive dose on normal tissues in shallow layers during treatment. It should be understood that the thermal neutron absorber can also be integrated with the moderator, and the moderator material contains Li-6. The radiation screen 124 screens neutrons and photons leaking from positions outside the beam exit 125, and the material of the radiation screen 124 comprises at least one of a photon-screening material or a neutron-screening material. In a preferred embodiment, the material of the radiation screen 124 includes lead (PB) used as a photon-screening material and polyethylene (PE) used as a neutron-screening material. The collimator 13 is located at the rear of the beam exit 125, and the epithermal neutron beam exiting the collimator 13 is radiated onto the irradiated body 200, wherein after passing through normal tissues in shallow layers, the epithermal neutron beam is slowed down to thermal neutrons in order to reach the tumor cell M. It should be understood that the beam forming body 20 can also have other configurations, provided that the epithermal neutron beam necessary for the treatment is obtained. For ease of description, when a collimator 13 is provided, the output of the collimator 13 can also be used as a beam output 125, as described below. In this embodiment of the invention, a radiation shielding device 15 is also located between the irradiated body 200 and the beam output 125 to shield the radiation of the beam exiting from the beam output 125 to normal tissues of the irradiated body, and it should be understood that the radiation shielding device 15 may not be provided.

[0088] After the drug containing boron (B-10) is taken or injected into the irradiated body 200, the drug containing boron selectively aggregates in the tumor cell M, after which two highly charged particles are generated using the characteristic of a drug containing boron (B-10) having a high capture cross-section for a thermal neutron, and by means of neutron capture of 10 B (n, α) and nuclear fission reactions. These two charged particles have an average energy of about 2.33 MeV and have the characteristics of high LET and the range of α-particles are 150 keV/μm and 8 μm, respectively, the LET and range of the highly charged particle are 175 keV/μm and 5 μm, respectively, and these two particles have a common range approximately equivalent to the size of a cell, so that radiation damage to the body can be limited to the cellular level, and the task of local destruction of a tumor cell can be solved under the assumption that it does not cause too much damage to normal tissues.

[0089] The BNCT system 100 is an integral part of a building of concrete structure. In particular, the BNCT system 100 further comprises an irradiation chamber 101 and a charged particle beam generation chamber 102. The irradiated body 200 on the treatment table 14 is irradiated with a neutron beam N in the irradiation chamber 101. The charged particle beam generation chamber 102 at least partially comprises an accelerator 111, and the beam forming body 12 is at least partially contained in a partition 103 between the irradiation chamber 101 and the charged particle beam generation chamber 102. It should be understood that the partition 103 can completely separate the irradiation chamber 101 from the charged particle beam generation chamber 102; or may be a partial separation between the irradiation chamber 101 and the charged particle beam generation chamber 102, so that the irradiation chamber 101 is in communication with the charged particle beam generation chamber 102. One or more targets T may be provided, the charged particle beam P may selectively interact with one or more targets T or simultaneously interact with a plurality of targets T to generate one or more beams of therapeutic neutrons N. In accordance with the number of targets T, there may also be one or more beam forming bodies 12, collimators 13 and treatment tables 14; a plurality of treatment tables may be located in the same irradiation chamber, or a separate irradiation chamber may be provided for each treatment table. The irradiation chamber 101 and the charged particle beam generation chamber 102 are spaces formed by a surrounding concrete wall W (including a partition 103), and the concrete structure can shield neutrons and other radiation beams leaking during the operation of the BNCT system 100.

[0090] In order to maximize the destruction of cancer cells while reducing the damage to normal tissues caused by irradiation rays, the accuracy of the dose distribution of epithermal neutrons and photons is especially important when setting up the treatment plan module 20. In the application scenes of the radiation beam irradiation system, it is necessary to read the photon and neutron cross-section databases for dose calculation, and the neutron cross-section database is very large, which leads to the occupation of a large storage space when installing the treatment plan development software of the radiation beam irradiation system, and also requires a large memory space and a long calculation time when the software is used for dose calculation and simulation. In the following embodiments of the invention, a number of methods are provided for optimizing the treatment plan module, which can reduce the dose calculation time while reducing the working memory, to meet the requirements of quickly generating a radiation therapy plan outline. Each optimization method is described below with reference to the drawings.

[0091] Option 1: Optimizing the database in module 20 of the treatment plan.

[0092] There are more than 60 elements in the human body in total, of which more than 20 elements are essential elements and are of great importance for maintaining the normal physiological functions of the body. Elements with a large content in the human body are carbon, hydrogen, oxygen, nitrogen, phosphorus, chlorine, sodium, magnesium, potassium, calcium or the like, wherein carbon, hydrogen, oxygen and nitrogen are the main elements constituting the organic substances of the human body and account for 96% of the total weight of the human body, and the remaining macroelements exceeding 0.01% of the human body are calcium, potassium, phosphorus, sulfur, chlorine, magnesium and sodium, respectively, which account for 1.5%, 0.35%, 1%, 0.25%, 0.15%, 0.05% and 0.15% of the total weight of the human body, respectively. It can be seen that the content of carbon, hydrogen, oxygen, nitrogen, calcium, potassium, phosphorus, sulfur, chlorine, magnesium and sodium reaches 99.45% of the total weight of the human body. At the same time, Monte Carlo stores the simulation and calculation systems of more than 100 elements, and each time Monte Carlo is used for simulation and calculation, each element is simulated and calculated, which takes a long time. At the same time, since the calculation and simulation systems corresponding to many elements are stored, the Monte Carlo database is very large. However, in the application scenes of the radioactive ray irradiation system, when particles such as photons, neutrons, etc. in the human body are simulated and calculated by the Monte Carlo method, only the elements in the human body with large proportions (more than 0.01%) relative to the total weight of the human body need to be simulated and calculated to obtain a relatively accurate calculation result, while the simulation and calculation of some impurity elements may take a long time, and when the obtained calculation result is compared with the result obtained by performing the simulation and calculation of only the elements occupying 0.01% or more of the total weight of the human body, the difference is less than 0.001%.

[0093] In this case, some elements such as Fe have a large weight fraction in the human body, but these elements do not react with neutrons and photons or have a small reaction intensity with neutrons and photons, so the influence of the simulation and calculation results of such elements on the overall calculation result can be ignored. However, some elements such as boron have a small weight fraction in the human body, but a large amount of boron preparations can be introduced into the human body in the scenes of using the radioactive ray irradiation system and have a relatively intense reaction with neutrons, and thus have a great influence on the final calculation result.

[0094] Elements (selected from one or more of H, He, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, Ar, K and Ca) that affect the simulation and calculation results of neutrons and photons in scenes of application of the radioactive ray irradiation system are filtered, taking into account the weight proportions of the elements in the human body and the intensity of the elements reacting with neutrons and photons, so that only the filtered elements are simulated in the simulation process. Thus, the simulation and calculation time is shortened while ensuring the calculation accuracy, which improves the performance of the Monte Carlo system. Meanwhile, the capacity of the database corresponding to the filtered elements is only 2% to 3% of the original capacity of the database, which significantly reduces the need for disk memory and reduces the production cost.

[0095] On the other hand, in the neutron cross section database corresponding to each element, the cross section data of a plurality of temperatures is stored in the Monte Carlo system for selecting, for example, 0 K, 1200 K, 2500 K, 250 K, 294 K, 600 K, 900 K, but in the application scenes of the radioactive ray irradiation system, it is necessary to simulate and calculate only the human body model, and the normal temperature of the human body is from 310 K to 315 K. According to the working principle of Monte Carlo, after inputting the temperature, the system automatically selects the temperature closest to the input temperature from among the plurality of stored temperatures and performs the simulation and calculation by taking this temperature as the simulation parameter. Therefore, after inputting the human body temperature, the system automatically selects the temperature 294 K stored in the database and performs the simulation and calculation by taking 294 K as the parameter. In this case, the temperature of 0 K is used in the calculation of the Doppler effect. In other words, in the scenes of application of the radioactive beam irradiation system, only two temperatures of 294 K and 0 K are used among the temperatures stored in the database, and the other temperatures are not relevant for the application of the radioactive beam irradiation system. Therefore, the databases corresponding to temperatures other than 294 K and 0 K are deleted, and the size of the database can be reduced to about one-third of the original size without affecting the structure of the simulation and calculation, which reduces the operating cost of the database.

[0096] Table 1 shows the result of the Monte Carlo system performing the simulation and calculation of the neutron dose on the same human body model at temperatures of 294 K, 310 K and 330 K, respectively.

[0097] It can be seen from the data in Table 1 that the influence of temperature fluctuation in a small range on the final calculation result of neutron dose can be ignored, and when using the radioactive beam irradiation system, only the database corresponding to 294 K is saved and stored in the system, which can meet the requirements of use, and the error generated by the final calculation result is within the acceptable range.

[0098] Option 2: Interrupt the process during particle simulation to reduce particle simulation time.

[0099] In the application scenes of the radioactive ray irradiation system, the dose calculation includes the neutron dose calculation and the photon dose calculation. When photons move in the irradiated body, the energy and half-absorption thickness of the photon gradually decrease, and the absorption cross section of the photon rapidly increases with the decrease in the energy and half-absorption thickness of the photon. When the half-absorption thickness of the photon is less than the cell size, the photon is absorbed in the cell where the photon is located with a very high probability, and at this time, the photon directly begins to give up all its energy in the cell where the photon is located, and the photon simulation no longer continues. The error between this process and the dose distribution calculated by continuously simulating the photon is within 0.1%.

[00100] The human cell has a size of 2 to 200 microns, and in general, when using a radioactive beam irradiation system, the model grid has a minimum size of 0.8 mm, so the photon modeling can be stopped when the photon half-absorption thickness is less than a quarter of the minimum grid size.

[00101] In particular, the half-absorption thickness t of a photon is calculated using formula (1-1):

[00102] where - the linear attenuation coefficient of a photon, determined by the material through which the photon passes and the energy of the photon.

[00103] The half-absorption thicknesses of various photon energies in the human skeleton are calculated below using the human skeleton as an example.

[00104] It can be seen from Table 2 that in the skeleton, the half-absorption thickness of the photons is less than 0.2 mm when the photon energy is less than 10 keV, that is, when the photon energy is less than 10 keV, the photon can be absorbed in the grid where the photon is currently located with a very high probability, and therefore the difference between the dose obtained when the photon simulation is stopped at this time and the dose obtained when the photon simulation continues is within an acceptable range.

[00105] Since the half-absorption thickness of a photon in a cellular tissue depends on the material passing through the photon and the energy of the photon, a corresponding interruption energy can be provided for each material, and the photon simulation can be terminated when the photon energy is lower than the interruption energy. The photon dose corresponding to different photon interruption energies is obtained by performing the simulation and calculation in the human body model below.

[00106] It can be seen from Table 2 that the photon simulation is stopped when the photon energy is less than or equal to 10 keV, and the error between the final calculated photon dose and the dose distribution calculated by continuous photon simulation is within 0.1%.

[00107] Finally, during the simulation and calculation of a photon, the half-absorption thickness of a photon is smaller than the cell size when the photon energy is less than the preset interruption energy, and at this time, the simulation of the photon is stopped, which can reduce the calculation time of the photon dose without affecting the calculation result of the dose distribution.

[00108] In other embodiments of the invention, the first preset value and the second preset value may be set in accordance with the actual requirements of the accuracy of the simulation, wherein the simulation of the photon is terminated in response to the fact that the half-absorption thickness of the photon is less than or equal to the first preset value, or in response to the fact that the energy of the photon is less than or equal to the second preset value. The first preset value may be greater than the cell size, or may be less than the cell size, and the second preset value may be 12 keV, 16 keV, or the like.

[00109] Option three: Speeding up the dose calculation module using parallel technology.

[00110] The Monte Carlo method is used in the treatment plan module of the radioactive irradiation system, the simulation processes of different source particles are completely independent, that is, the simulation sequences of different particles do not affect the calculation result. According to such calculation characteristics, the simulation tasks of different source particles can be assigned to different processes or threads, and a summary is performed after the completion of the calculation task of each of the processes or threads to obtain the final calculation result. The computer can provide two forms of parallelism: CPU process parallelism and thread parallelism, and GPU acceleration.

[00111] As shown in Fig. 3, the process of calculating the process parallelism and the CPU thread parallelism is as follows. First, the system obtains the number of processes or threads to obtain a numerical value n; then the system equally divides the particles required for the simulation into n parts; then each of the threads or processes separately simulates and counts each part of the particles; and finally, the system performs a count of the amount obtained by each of the processes or threads to obtain a final dose. Since the particles are assigned to different processes or threads so as to perform the simulation and calculation simultaneously, the simulation and calculation time is reduced to 1/n (process) or 1/2n of the time required to use one thread or process to simulate and calculate all the particles.

[00112] Each of the process parallelism and the thread parallelism realizes parallel computing by a multi-core processor, and the GPU acceleration is realized by parallel computing by a plurality of GPU processors. The effect of each of the process parallelism and the thread parallelism is limited by the number of CPU cores, and the number of cores of the general autonomous computer is four, the process parallelism and the thread parallelism can increase the speed by 4 times (process) or 8 times (thread) at most, and one GPU combines multiple processors, and theoretically the calculation speed can be increased many times.

[00113] As shown in Fig. 4, the process of simulating and calculating the acceleration of the GPU is as follows. First, the system transfers a random number, cross-section data or the like from the memory of the CPU to the video memory of the GPU, then each of the plurality of processors of the GPU simulates, calculates and counts one particle and counts the counting result into the global count; then the system determines whether there are particles that have not yet been simulated, transfers the counting result from the memory of the GPU to the memory of the CPU in response to the absence of particles that have not yet been simulated, or returns to the previous operation in response to the presence of particles that have not yet been simulated, to continue simulating and counting the particles that have not yet been simulated until the simulation of all the particles is completed.

[00114] Since the GPU has multiple processors, GPU-accelerated simulation and calculation can greatly improve the calculation speed and shorten the calculation time. In this case, the CPU uses the Intel Xeon CPU with a frequency of 2.27 GHz, priced at about 6,000 yuan, the GPU uses the NVIDIA Tesla C2050, priced at about 9,000 yuan, and each GPU has a total of 448 processors, equivalent to 448 GPUs. The CPU calculation time is 50 to 70 times that of the GPU calculation. It can be seen that compared with the CPU, the GPU is used to perform simulation and calculation, which has great advantages in both price and time.

[00115] Option 4: Use Monte Carlo variance reduction techniques to speed up convergence.

[00116] The Monte Carlo method is rich in variance reduction techniques to increase the speed of program calculation, and the variance reduction techniques that can be used by the treatment plan system include absorption by weight reduction, weight window game, split bet, and the like. In this case, split bet is formed from the bet technique and the split technique. The bet technique, split technique, importance of grid space, absorption by weight reduction, and weight window game involved in the variance reduction technique are described below.

[00117] Betting technique: In general, the particle weight has an upper limit of 10 and a lower limit of 0.25. When the weight w of a particle decreases to less than a certain preset value (e.g. 0.25), a random number x is sampled between 0 and 1, and when x is less than w, the particle survives and the particle weight returns to 1; when x is greater than or equal to w, there is a dead bet for the particle and the particle simulation is terminated.

[00118] Splitting technique: When the weight w of a particle is greater than a certain value (e.g. 10), the integer part of w is set to w1, the decimal part of w is set to w2, a random number x is chosen between 0 and 1, w=w1+1 is set when x is less than w2, and w=w1 is set when x is greater than w2. The particle is then split into w particles for simulation until the weight of each particle decreases to 1.

[00119] Grid space importance: particles in different regions of the model have different dose contributions, the dose contribution of a particle is characterized by the space importance, and the grid space importance of each grid is calculated according to parameters such as tumor position, model characteristics, or the like before modeling. The particle moves from a grid with a grid space importance of 1 _n to a grid with a grid space importance of and when is established, the particle splits into m particles, the weight of each of the particles decreases to 1/t of the original weight; when the betting technique is performed on the particle, it is installed a random number x is chosen between 0 and 1, and when x is less than P, the particle is saved and the weight is multiplied by 1/P, otherwise a dead bet is given to the particle and the particle simulation is terminated.

[00120] Method for calculating the importance of the grid space: The importance of the grid space can be obtained by calculating the associated flow obtained by solving the associated transport equation, which is:

[00121] where - accompanying flow, - accompanying source, - the speed of the particle, - direction of particle movement, - the cross section of the reaction of a particle colliding with a substance, - scattering cross-section, r - particle position, E - particle energy, t - time.

[00122] Weight reduction absorption: When a particle interacts with matter, the particle is only scattered and not absorbed, and the particle's weight is multiplied by P _scattering / P _total in each collision. When the particle's weight is reduced to a certain value (e.g. 0.25), the particle is processed using the betting technique.

[00123] Weight window game: When the particle weight is greater than a certain value such as 10, the splitting technique is performed, and when the particle weight is less than a certain value such as 0.25, the betting technique is performed. Here, the two values are set according to the performance of the software.

[00124] In general, when a particle moves from a place with a small grid space value to a place with a large grid space value, the weight of the particle increases and there is no dead weight for the particle; wherein the case where the weight decreases means that when a particle moves from a place with a large grid space value to a place with a small grid space value, the weight of the particle decreases, the particle is split or absorption by weight reduction occurs. Since a thermal neutron has a large absorption cross section in the human body (mainly elements such as N, B or the like), in order to accelerate the convergence rate of the dose calculation of the grids far from the neutron source, absorption by weight reduction should be used to move the particle to the grids. When the particle is far from the treatment area, continuing the particle dose calculation simulation on the treatment area does not have a great significance, and at this time, the waste of computing resources may be caused by the continuous calculation, so the probability of this situation occurring can be reduced by setting the importance of the grid space. When the particle weight is too small, the particle contribution to the count is very small, and continuing the simulation will waste computational resources, and when the particle weight is too large, the single count is too large, which will lead to the risk of distorting the result, and thus a weight window must be provided to ensure that the particle weight is in a suitable range. Therefore, it is necessary to combine the absorption of weight reduction, reasonable adjustment of the importance of the grid space and the play of the weight window to accelerate the convergence speed of the Monte Carlo program.

[00125] As shown in Fig. 5, the process of modeling particles using the dispersion reduction technique is as follows. In operation S1, an original particle is obtained. In operation S2, it is determined whether the particle collides with a lattice element, and operations S3 and S4 are sequentially performed in the case of a collision, otherwise operations S5 and S6 are sequentially performed. In operation S3, absorption by reducing the weight is performed. In operation S4, it is determined whether the weight is less than the weight window, and operation S7 is performed in response to the fact that the weight is less than the weight window, otherwise the process returns to operation S2. In operation S5, a split bet is performed. In operation S6, it is determined whether the bet is performed, and operation S7 is performed in response to the performance of the bet, otherwise the process returns to operation S2. In operation S7, it is determined whether there is a dead bet, and operation S8 is performed in response to the presence of a dead bet, otherwise operation S9 is performed, after which the process returns to operation S2. In operation S8, it is determined whether the particle processing is complete, and the process is terminated in response to the completion of the particle processing, otherwise the process returns to operation S1. In operation S9, the weight is divided by the probability of a dead bet.

[00126] As shown in Fig. 6, the splitting bet process is as follows. In operation S1, the grid space importance of each grid is calculated and recorded. In operation S2, particles are obtained. In operation S3, a weight window check is performed on the particle and the operation is performed. In operation S4, the grid space importance is calculated before and after the particle overlaps the mesh boundary. In operation S5, a comparison is made to determine whether the particle is greater than how and operation S6 is performed, after which the process returns to operation S3 in response to the fact that no more otherwise, operation S7 is performed. In operation S6, the particle is split and the weight of the particle is decreased. In operation S7, it is determined whether there is a dead rate for the particle, and operation S9 is performed in response to the determination that there is a dead rate for the particle, otherwise operation S8 is performed, after which the process returns to operation S3. In operation S8, the weight of the particle is increased. In operation S9, it is determined whether the particle simulation is complete, and the process returns to operation S2 in response to the determination that the particle simulation is not complete, otherwise the process ends.

[00127] As shown in Fig. 7, the process of absorption by weight reduction is as follows. In operation S1, particles are obtained. In operation S2, it is determined whether the particle collides with the lattice element, and operation S3 is performed in the case of a collision, otherwise the process returns to operation S1. In operation S3, the weight is multiplied by the probability of occurrence of scattering. In operation S4, it is determined whether the weight of the particle is less than the smallest weight, and operation S5 is performed in response to the fact that the weight is less than the smallest weight, otherwise the process returns to operation S2. In operation S5, it is determined whether there is a dead bet for the particle, and operation S7 is performed in response to the presence of a dead bet for the particle, otherwise operation S6 is performed, after which the process returns to operation S2. In operation S6, the weight is divided by the probability of a dead bet. In operation S7, it is determined whether the simulation of the particles is complete, and the process returns to operation S1 in response to determining that the simulation of the particles is not complete, otherwise the process ends.

[00128] As shown in Fig. 8, the process of the weight window game is as follows. In operation S1, the movement of particles is simulated. In operation S2, it is determined whether the weight of the particle falls within the range of the weight window, and the process returns to operation S1 in response to the weight of the particle falling within the range of the weight window, otherwise, operation S3 is performed. In operation S3, it is determined whether the weight is greater than the weight window, operation S4 is performed, after which the process returns to operation S1 in response to the fact that the weight is greater than the weight window, otherwise, operation S5 is performed. In operation S4, the particle is split and the weight is reduced. In operation S5, it is determined whether there is a dead bet, and the process ends in response to the presence of a dead bet, otherwise, operation S6 is performed, after which the process returns to operation S1. In operation S6, the weight of the particle is increased.

[00129] To compare the effects of accelerating the dispersion reduction technique leading to dose calculation, the standard deviations of the results and the calculation time are compared in the case of using the weight reduction absorption technique and without using the weight reduction absorption technique. For the same individual membrane model, ten million particles are calculated, the calculation time is 1584 s, and the standard deviation of the dose calculation is 11% using the weight reduction absorption technique, while without using the weight reduction absorption technique, the calculation time is 1258 s, and the standard deviation of the dose calculation is 14.5%. To reduce the standard deviation to the same level as when using the weight reduction absorption, it is necessary to increase the number of simulated particles to 16 million and increase the simulation and calculation time to 2045 s, which increases the calculation time by 461 s compared to using the weight reduction absorption technique, that is, using the weight reduction absorption technique reduces the calculation time by about 20%.

[00130] Option 5: Performing simulation and calculation using non-uniform rectangular grids.

[00131] The traditional calculation of the body membrane usually uses uniform meshes, and sometimes the number of meshes needs to be increased to accurately calculate some areas, so that the calculation time increases and the memory increases exponentially. Generally, uniform meshes are obtained from the model read from CT or PET, here many related areas have the same material such as air, blood or the like, and some of the areas do not need to accurately calculate the careful dose distribution, so the fine mesh can be replaced by a thicker mesh; and for important positions such as tumors or the like, finer meshes need to be divided to achieve more accurate calculation. The dose is calculated using non-uniform meshes, the calculation accuracy of the important area can be improved without significantly increasing the calculation time, and the calculation time is shortened in the case where the calculation accuracy of the unimportant area is achieved.

[00132] To perform the simulation and calculation below, grids of different sizes are specified and the effects of different grid sizes on the simulation time and calculation accuracy are compared.

[00133] In this experiment, a mixed grid is used, formed by 0.4 mm grids occupying the first 5 cm in the direction of neutron incidence, 0.8 mm grids occupying the middle 5 cm in the direction of neutron incidence, and 1.6 mm grids occupying the rest of the mixed grid. It can be seen from Table 3 that the total number of grids can be reduced to 2/5 of the original number of 0.4 mm grids when using non-uniform grids, the memory required by the grids is also reduced to 1/3 of the original memory, and the calculation time is reduced to 34.7% of the original calculation time. The result obtained by calculation using 0.4 mm grids is taken as a standard, while the result obtained by calculation using mixed grids has a neutron error of less than 0.1% and a photon error of less than 0.2%.

[00134] The composition of the mixed mesh is not limited to the mode as illustrated above, and can be set according to the specific situation of the body to be irradiated. Generally, the mesh size is determined by the degree of importance of the region, for example, the mesh size of the region where a tumor is located is small and is set as a mesh smaller than or equal to 0.4 mm, the mesh size of the region where each of blood, air, and bone is located is relatively large and is set as greater than or equal to 1.6 mm, and the mesh size of the region where other tissue such as normal muscle or the like is located is set as greater than 0.8 mm and smaller than 1.6 mm.

[00135] The algorithm used for calculating the dose is a Monte Carlo algorithm, which has the advantage of high calculation accuracy and the disadvantages of low convergence speed and long calculation time, so that optimization of the calculation efficiency is the most important part of the optimization. According to the optimization method of each of the first to fifth embodiments of the invention, the calculation efficiency can be improved to different degrees, and the operating time can be shortened.

[00136] The above described illustrative specific embodiments of the invention are such that a person skilled in the art understands the invention, however, it should be obvious that the invention is not limited to the scope of these specific embodiments, various changes are obvious to a person skilled in the art and fall within the scope of legal protection of the invention if these changes correspond to the essence and scope of legal protection of the invention, as defined in the attached claims.

Claims (12)

1. A radioactive radiation irradiation system, characterized in that the radioactive radiation irradiation system contains: a beam irradiation device that generates a therapeutic beam and radiates the therapeutic beam onto an irradiated body to form an irradiated area; a treatment plan module that performs dose modeling and calculation according to the parameters of the treatment beam and the medical imaging data of the irradiated area and generates a treatment plan that filters elements that affect the results of modeling and calculation of neutrons and photons in scenes of application of the radiation irradiation system by taking into account the weight proportions of a plurality of elements in the human body and the intensities of the reaction between the plurality of elements and neutrons and photons, so that only the filtered elements are modeled during the modeling; and a control module that controls the radiation of the beam irradiation device in accordance with the treatment plan. 2. The radiation delivery system of claim 1, wherein the elements filtered by the treatment plan module are selected from one or more of H, He, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, Ar, K and Ca. 3. The radiation delivery system of claim 2, wherein the elements filtered by the treatment plan module are H, He, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, Ar, K and Ca. 4. A radiation irradiation system according to any one of claims 1 to 3, wherein the radiation irradiation system is a neutron capture therapy system, and the beam irradiation device comprises a neutron generation device, a beam shaping body and a therapeutic table, the neutron generation device comprises an accelerator and a target, the accelerator accelerates charged particles to generate a charged particle beam that interacts with the target to generate a neutron beam, the beam shaping body is capable of regulating the neutron beam generated by the neutron generation device to have a preset beam quality, wherein the neutron beam generated by the neutron generation device is radiated onto the irradiated body on the therapeutic table by means of the beam shaping body. 5. A radiation irradiation system according to any one of claims 1 to 3, wherein the radiation irradiation system is a neutron capture therapy system, and the beam irradiation device comprises a neutron generation device, a beam shaping body and a treatment table, wherein the neutron generation device comprises an accelerator and a target, the accelerator accelerates charged particles to generate a beam of charged particles interacting with the target to generate a neutron beam, the beam shaping body is capable of adjusting the neutron beam generated by the neutron generation device to have a preset beam quality, wherein the neutron beam generated by the neutron generation device is radiated onto the irradiated body on the treatment table by means of the beam shaping body. 6. A method for controlling a system for delivering radioactive radiation, characterized in that the beam irradiation system comprises: a beam irradiation device that generates a therapeutic beam and emits the therapeutic beam onto an irradiated body to form an irradiated area; a treatment plan module that performs modeling and dose calculation in accordance with the parameters of the therapeutic beam and medical imaging data of the irradiated area and generates a treatment plan; and a control module that controls the radiation of the beam irradiation device in accordance with the treatment plan, a method for controlling a radioactive radiation irradiation system comprises: filtering elements that affect the results of modeling and calculating neutrons and photons in scenes of using the radioactive radiation irradiation system by taking into account the weight proportions of a plurality of elements in the human body and the intensity of the reaction between the plurality of elements and neutrons and photons, so that only the filtered elements are modeled during modeling. 7. The method for controlling the radiation delivery system of claim 6, wherein the elements filtered by the treatment plan module are selected from one or more of H, He, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, Ar, K and Ca. 8. The method for controlling the radiation delivery system of claim 7, wherein the elements filtered by the treatment plan module are H, He, Li, Be, B, C, N, O, F, Na, Mg, Al, Si, P, S, Cl, Ar, K and Ca.