WO2023123420A1 - 放疗剂量确定方法、装置、设备及存储介质 - Google Patents

放疗剂量确定方法、装置、设备及存储介质 Download PDF

Info

Publication number
WO2023123420A1
WO2023123420A1 PCT/CN2021/143835 CN2021143835W WO2023123420A1 WO 2023123420 A1 WO2023123420 A1 WO 2023123420A1 CN 2021143835 W CN2021143835 W CN 2021143835W WO 2023123420 A1 WO2023123420 A1 WO 2023123420A1
Authority
WO
WIPO (PCT)
Prior art keywords
voxel
simulation
deposition
area
voxels
Prior art date
Application number
PCT/CN2021/143835
Other languages
English (en)
French (fr)
Inventor
李波
Original Assignee
西安大医集团股份有限公司
深圳市奥沃医学新技术发展有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 西安大医集团股份有限公司, 深圳市奥沃医学新技术发展有限公司 filed Critical 西安大医集团股份有限公司
Priority to PCT/CN2021/143835 priority Critical patent/WO2023123420A1/zh
Publication of WO2023123420A1 publication Critical patent/WO2023123420A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
    • A61N5/00—Radiation therapy
    • A61N5/10—X-ray therapy; Gamma-ray therapy; Particle-irradiation therapy
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
    • G16H20/40—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to mechanical, radiation or invasive therapies, e.g. surgery, laser therapy, dialysis or acupuncture

Definitions

  • the present application relates to medical information technology, in particular to a radiotherapy dose determination method, device, equipment and storage medium.
  • the dose distribution of the target subject can be obtained through radiotherapy simulation, which can be used as a reference to formulate a matching radiotherapy plan for the target subject.
  • serial Monte Carlo simulation refers to sequentially performing simulation on a plurality of sampled particles, and the simulation of the next particle can only be performed after the simulation of one particle is completed.
  • the time required for the serial Monte Carlo simulation is longer, which prolongs the time for formulating the treatment plan, thus increasing the time of the whole treatment process.
  • the embodiments of the present application provide a radiotherapy dose determination method, device, equipment, and storage medium, so as to shorten the entire duration of the radiotherapy simulation process, shorten the time for formulating a treatment plan, and thereby shorten the time for the entire treatment process.
  • the embodiment of the present application provides a radiotherapy dose determination method comprising:
  • the deposition dose of each voxel is determined.
  • the parallel Monte Carlo simulation is performed on the multiple particles sampled in the simulation area, and the simulation results of the multiple particles are obtained, including:
  • the cache Space is used to store the deposited energy of the particle in the corresponding voxel.
  • a cache space is allocated to some voxels in the simulation area, and Monte Carlo simulation is performed until the particles pass through all the voxels.
  • the above simulation area is used to obtain the simulation results, including:
  • the data in the cache space is transferred to the shared storage space to obtain the simulation result.
  • the deposition area corresponding to each layer voxel of the particles in the simulation area is sequentially determined, for each layer
  • Each voxel in the deposition area corresponding to the voxel allocates a corresponding cache space, including:
  • the incident angle determine the incident voxel and the voxels in the first layer of voxels within a preset range around the incident voxel as the particles corresponding to the first layer of voxels The first layer of deposition area;
  • the area corresponding to the deposition area of the first layer and the voxels in the surrounding preset range in the voxel of the next layer are determined as the particle in the voxel
  • the deposition area when the particles enter each layer of voxels is sequentially determined, and the corresponding buffer space is allocated to store the deposition energy of each voxel in the corresponding deposition area.
  • the transfer of the data in the cache space to the shared storage space to obtain a simulation result includes:
  • the data stored in the released cache space is transferred to the shared storage space.
  • the method further includes:
  • the calculation of the uncertainty of the parallel Monte Carlo simulation according to the deposited energy of the Nth particle in the shared memory space includes:
  • the deposition energy of the Nth particle in a single voxel is the average deposition dose of the Nth particle in each corresponding voxel
  • the energy variance of the Nth particle in a single voxel is: The deposition energy of the Nth particle in each corresponding voxel and the variance sum of the average deposition energy.
  • the embodiment of the present application can also provide a radiotherapy dose determination device, including:
  • the division module is used to perform three-dimensional grid division on the simulation area to obtain multiple voxels
  • a simulation module configured to perform parallel Monte Carlo simulation on a plurality of particles sampled in the simulation area, to obtain simulation results of the plurality of particles
  • the determining module is configured to determine the deposition dose of each voxel according to the simulation result.
  • an embodiment of the present application further provides a computer device, including: a memory and a processor, the memory stores a computer program executable by the processor, and the processor implements the above-mentioned first method when executing the computer program.
  • a radiotherapy dose determination method according to any one of the aspects.
  • the embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored.
  • a computer program is stored.
  • the radiotherapy described in any one of the above-mentioned first aspects can be realized. Dose Determination Method.
  • the radiotherapy dose determination method, device, equipment, and storage medium provided in the embodiments of the present application can divide the simulation area into a three-dimensional grid to obtain multiple voxels, and perform parallel Monte Carlo on the multiple particles sampled in the simulation area. Simulate, obtain the simulation results of multiple particles, and then determine the deposition dose of each voxel according to the simulation results. That is to say, the radiotherapy dose determination method provided in the embodiment of the present application is actually a method for determining the radiotherapy dose through a parallel Monte Carlo simulation method. The Monte Carlo simulation method for multiple particles is executed in parallel, which effectively shortens the entire radiotherapy dose. The time-consuming determination of radiotherapy dose during the simulation process improves the efficiency of Monte Carlo simulation, shortens the time for making treatment plans, and thus shortens the time for the entire treatment process.
  • Fig. 1 is a flowchart of a radiotherapy dose determination method provided in the embodiment of the present application
  • Fig. 2 is a method flowchart of parallel Monte Carlo simulation in a radiotherapy dose determination method provided by the embodiment of the present application;
  • Fig. 3 is a method flowchart 1 of a Monte Carlo simulation process in a radiotherapy dose determination method provided in the embodiment of the present application;
  • Fig. 4 is a method flowchart two of a Monte Carlo simulation process in a radiotherapy dose determination method provided in the embodiment of the present application;
  • FIG. 5 is a flow chart of a method for transferring data in a cache space in a radiotherapy dose determination method provided in an embodiment of the present application
  • Fig. 6 is a schematic diagram of a radiotherapy dose determination device provided in an embodiment of the present application.
  • FIG. 7 is a schematic diagram of a computer device provided by an embodiment of the present application.
  • first and second are used for description purposes only, and cannot be interpreted as indicating or implying relative importance or implicitly indicating the quantity of indicated technical features. Thus, a feature defined as “first” or “second” may explicitly or implicitly include one or more of said features.
  • “plurality” means two or more, unless otherwise specifically defined.
  • Monte Carlo simulation As the most accurate algorithm in the field of radiotherapy, it can use the preset probability distribution model to sample the emitted particles of the radiation source model, and then simulate the movement of particles in the area to be simulated according to the statistical laws of microscopic particle movement Through the process, the radiation dose distribution of the sampled particles in the area to be simulated is obtained.
  • the embodiment of the present application provides a method of performing radiation therapy simulation using a parallel Monte Carlo simulation method to determine multiple possible implementation modes of the radiation therapy plan.
  • the device for performing the radiotherapy dose determination method may be a computer device installed with a radiotherapy dose determination application, and the computer device may execute the corresponding radiotherapy dose determination method by running the radiotherapy dose determination application.
  • the radiotherapy dose determination application can be a sub-function module of a radiotherapy planning system (Treatment Plan System, TPS), which can also be called a Monte Carlo calculation module.
  • TPS Treatment Plan System
  • Fig. 1 is a flowchart of a radiotherapy dose determination method provided in an embodiment of the present application.
  • radiotherapy dose determination methods may include:
  • the simulation area is also called the calculation area, which may be a preset three-dimensional area of a target object, and the target object may be an object to be treated or a preset phantom object. Since the simulation area is actually a three-dimensional area, the division of the simulation area can be in three-dimensional space. According to the preset size of a single voxel, the simulation area is divided into three-dimensional grids to obtain multiple voxels, so that the simulation area There are a, b, and c voxels respectively on the three coordinate axes in the preset three-dimensional coordinate system.
  • the preset three-dimensional coordinate system may be, for example, an xyz coordinate system, and the three coordinate axes may be an x coordinate axis, a y coordinate axis, and a z coordinate axis in turn.
  • the emitted particles of the source module can be sampled according to the preset probability distribution model to obtain a plurality of sampled particles.
  • the preset probability distribution model may be, for example, a random probability distribution model, or another probability distribution model, which is not limited in this embodiment of the present application.
  • a parallel Monte Carlo module can be performed on multiple particles in the simulation area. During the Monte Carlo simulation process for each particle, the movement of the corresponding particle in the simulation area Carry out the simulation, and in the process of motion simulation, calculate the deposition energy of the corresponding particle on each voxel, so as to obtain the simulation result of the corresponding particle.
  • the simulation results of each particle can include the deposition energy of the corresponding particle on each voxel, so in a possible implementation, the deposition dose of each voxel can be determined according to the simulation results of multiple particles, thus obtaining The radiation dose distribution of each voxel in the simulation area.
  • the deposition energy for the same voxel can be combined among the simulation results of multiple particles to obtain the deposition dose of the corresponding voxel .
  • the radiotherapy dose determination method provided in the embodiment of the present application can divide the simulation area into a three-dimensional grid to obtain multiple voxels, and perform parallel Monte Carlo simulation on multiple particles sampled in the simulation area to obtain the The simulation results, and then according to the simulation results, the deposition dose of each voxel is determined. That is to say, the radiotherapy dose determination method provided in the embodiment of the present application is actually a method for determining the radiotherapy dose through a parallel Monte Carlo simulation method.
  • the Monte Carlo simulation method for multiple particles is executed in parallel, which effectively shortens the entire radiotherapy dose.
  • the time-consuming determination of radiotherapy dose during the simulation process improves the efficiency of Monte Carlo simulation, shortens the time for making treatment plans, and thus shortens the time for the entire treatment process.
  • Fig. 2 is a flow chart of a parallel Monte Carlo simulation method in a radiotherapy dose determination method provided in an embodiment of the present application.
  • parallel Monte Carlo simulation is performed on multiple particles sampled in the simulation area, and the simulation results of multiple particles are obtained, which may include:
  • the exit angle of each sampled particle from the source model is determined. Therefore, for each particle, the exit angle of the corresponding particle determined during the sampling process and the relative emission angle of the simulation area can be determined.
  • the placement position of the source model determines the incident angle and incident voxel of the corresponding particles entering the simulation area.
  • the incident voxel is the voxel where the particles meet in the simulation area from outside the simulation area, and it can also be called the voxel that the particle passes through on the movement track in the simulation area.
  • the incident angle of the particle is the negative direction of the z coordinate
  • the first voxel that the particle meets in the simulation area is the cth voxel along the positive direction of the z coordinate. Therefore, the incident The voxels include the first encountered voxel v(i, j, c) and other voxels incident after passing through the first encountered voxel.
  • the cache space is used to store the deposited energy of particles in corresponding voxels.
  • Monte Carlo simulation is performed based on the incident angle and incident voxel of the particles determined above.
  • the corresponding buffer space is allocated to some voxels in the simulation area until the particle moves through the simulation. area, the Monte Carlo simulation process for the particles is completed, and the simulation results are obtained.
  • part of the voxels in the simulation process can be: the voxels associated with the trajectory of the particle in the simulation area, which can include, for example: the voxels that the trajectory of the particle in the simulation area passes through, and the particles in the simulation area Voxels within a preset range around the motion track of the target, etc.
  • Allocating a cache space for some voxels in the simulation area refers to allocating a corresponding cache space for each voxel in some voxels in the simulation area to store the deposited energy of the particles in the corresponding voxels.
  • the size of each buffer space may be, for example, a buffer area with a preset number of bytes, such as a four-byte floating-point number buffer.
  • voxels of the particles in the simulation area include: k voxels, and during this process, k buffer spaces may be allocated to store the deposition energy of the particles in the k voxels respectively.
  • the radiotherapy dose determination method can determine the incident angle and incident voxel of each particle entering the simulation area, and allocate buffer space for some voxels in the simulation area based on the incident angle and incident voxel, The Monte Carlo simulation is performed until the particles pass through the simulation area to obtain the simulation results, and the allocated buffer space is used to record the deposition energy of the particles in the corresponding voxels during the simulation.
  • the Monte Carlo simulation for each particle does not allocate the corresponding buffer space for all voxels in the simulation area, but only Allocate corresponding cache space for some voxels associated with particles in the simulation area to store the deposition energy of particles in the corresponding voxels, effectively reducing the Monte Carlo simulation process for each particle in the parallel Monte Carlo simulation.
  • the required memory resources ensure the parallel efficiency of Monte Carlo simulation for multiple particles, thereby effectively shortening the overall calculation time for radiotherapy dose determination, shortening the time for making treatment plans, and thus shortening the time for the entire treatment process.
  • FIG. 3 is a method flowchart 1 of a Monte Carlo simulation process in a radiotherapy dose determination method provided in an embodiment of the present application.
  • Figure 3 based on the incident angle and incident voxels in S202 as shown above, cache space is allocated for some voxels in the simulation area, and Monte Carlo simulation is performed until the particles pass through the simulation area, and the simulation results can include :
  • the particles can be simulated in the simulation area based on the incident angle and incident voxels, and then according to the results of the motion simulation, the deposition area corresponding to each layer of voxels of the particles in the simulation area can be sequentially determined.
  • the deposition area refers to a three-dimensional area composed of multiple voxels with particle deposition energy.
  • the result of running the simulation can include, for example: the movement trajectory of the particle in the simulation area, and the movement trajectory of the secondary particles generated during the movement of the particle in the simulation area, and the movement trajectory of the particle in the simulation area can be obtained through the particle
  • the movement trajectory of the secondary particle in the simulation area can be represented by the coordinate position of the voxel that the secondary particle passes through.
  • the corresponding buffer space can be assigned to each voxel in the deposition area corresponding to each layer of voxels, that is, for each layer of voxels in the deposition area corresponding to Voxels divide a cache space.
  • the number of voxels in the deposition area corresponding to voxels in different layers may be different, therefore, the amount of cache space allocated to the voxels in the deposition area corresponding to voxels in different layers is different.
  • the number of voxels in the deposition area corresponding to each layer of voxels can be based on the voxel density, medium type, particle type, energy of particles entering each layer of voxel, incident angle, and medium type in the simulation area.
  • At least one parameter such as the free path, combined with the value determined by the laws of physical motion of the particles.
  • k1 buffer spaces can be allocated to the k1 voxels in the deposition area corresponding to the first layer of voxels to store the recording particles in k1 voxels respectively The deposition energy in ; and the deposition area corresponding to the voxel in the second layer includes: k2 voxels, then k2 buffer spaces can be allocated for the k2 voxels in the deposition area corresponding to the voxels in the second layer to store the recording particles respectively Deposited energy in k2 voxels, and so on.
  • the allocation of buffer space can be performed after entering a certain layer of voxels and determining the deposition area corresponding to the corresponding layer of voxels. When not entering a new voxel layer, it is impossible to determine the corresponding The deposition area, therefore, does not allocate cache space.
  • the above example is only for the convenience of understanding, and the relationship between the voxels in the deposition area corresponding to the voxels of each layer and the corresponding allocated buffer space is given as an example.
  • the buffer space allocated for each voxel in the deposition area of each layer can be used to store the deposition energy of the particles and/or the secondary particles of the particles in the corresponding voxels.
  • all the data in the buffer space for the corresponding particles can be transferred to the shared storage space to obtain the simulation results; it can also be used during the simulation process Transfer the data in part of the cache space to the shared storage space, and after the simulation process is completed, transfer the data in other cache spaces to the shared storage space to obtain the simulation results.
  • the shared storage space is a shared memory space of multiple parallel Monte Carlo processes for multiple particles.
  • Fig. 4 is a method flowchart II of a Monte Carlo simulation process in a radiotherapy dose determination method provided in an embodiment of the present application.
  • the deposition area corresponding to each layer of voxel in the simulation area is sequentially determined, and the deposition area corresponding to each layer of voxel in S302
  • Each voxel in is assigned a corresponding cache space, which may include:
  • the incident angle determine the incident voxel and the voxels within the preset range around the incident voxel in the first layer voxel as the first layer deposition area corresponding to the particle in the first layer voxel.
  • the voxels in the first layer of voxels within a preset range around the incident voxel can be determined, and the incident voxels and the voxels within a preset range around the incident voxel can be determined is the first-layer deposition area corresponding to the first-layer voxel of the particle. That is to say, the voxels in the deposition area of the first layer include: the incident voxels and the voxels within the preset range around the incident voxels, wherein the incident voxels can be particles and/or secondary particles in the first layer of voxels.
  • the voxels that the trajectory of the first-level particles pass through, the voxels within the preset range around the incident voxel can be the voxels in the preset range around the motion trajectory of the corresponding particles in the first layer of voxels, and the size of the preset range can be is the range of the three-dimensional area determined according to the incident angle and taking the incident voxel as the reference point.
  • a cache space can be allocated for the incident voxel , to store the deposition energy of the particle in the incident voxel; when the particle moves to the first adjacent voxel v(i+1, j, c) whose z coordinate is still c in the first layer of voxels, then it can be The first adjacent voxel allocates another cache space to store the deposited energy in the first adjacent voxel; when the particle continues to move to the second adjacent voxel v( When i+1, j+1, c), another buffer space may be allocated to the second adjacent voxel to store the deposited energy in the second adjacent voxel.
  • the first layer of voxels may be a layer of voxels whose z coordinate is c, therefore, each voxel in the first layer of deposition area corresponding to the first layer of voxels determined from the first layer of voxels
  • the z coordinates of are also c.
  • the particle enters the voxel of the next layer that is, from the voxel of the first layer whose z coordinate is c to the voxel of the next layer whose z coordinate is c-1, it can determine the particle trajectory in the next layer of voxel
  • the voxel is the area corresponding to the deposition area of the first layer in the voxels of the next layer, combined with the voxels within the preset range around the particle trajectory, it is determined as the deposition area of the second layer corresponding to the voxels of the particle in the next layer .
  • the particle trajectory refers to the trajectory of the particle and/or the secondary particles generated by the particle. That is to say, the deposition area of the second layer includes: the voxels that the particle trajectory passes through, and the voxels within a preset range around the particle trajectory.
  • the specific realization of allocating the corresponding buffer space for each voxel in the deposition area of the second layer may be similar to the implementation process of allocating the buffer space for each voxel in the deposition area of the first layer above, and will not be repeated here.
  • S405 sequentially determine the deposition area when the particles enter each layer of voxels, and allocate corresponding buffer space to store the deposition energy of each voxel in the corresponding deposition area.
  • the method provided in this embodiment by determining the corresponding deposition area of particles in each layer of voxels, makes each layer of voxels correspond to some voxels in the deposition area more accurately, ensuring that each layer of voxels corresponds to the deposition area
  • the accurate determination of ensures the accuracy of the number of cache spaces allocated to voxels in the deposition area for each layer of voxels, and effectively reduces the cache allocated for each particle in the Monte Carlo simulation process in the simulation area
  • the consumption of the number of spaces can effectively ensure the parallel execution efficiency of the parallel Monte Carlo simulation process for multiple particles.
  • FIG. 5 is a flowchart of a method for dumping data in a buffer space in a radiotherapy dose determination method provided in an embodiment of the present application.
  • the data in the cache space is transferred to the shared storage space, and the simulation results obtained may include:
  • the specific value of m can be determined according to the voxel parameters of the simulation area and the attribute parameters of the particles.
  • the voxel parameters may include: at least one parameter in the voxel density, the medium type in the voxel, and the free path corresponding to the medium type, and the attribute parameters include: the type of particle, the initial energy of the particle, the incident angle of the particle, etc. At least one parameter.
  • the method provided in this embodiment can buffer the deposition energy in voxels within an acceptable range to the greatest extent, that is, when particles enter a certain layer, the particles will also be generated in voxels within the preset range of the previous layer Deposit energy, therefore, in this case, the data in the cache space cannot be released directly. Rather, when a particle enters the nth layer of the simulation area, based on the voxel parameters of the simulation area and the property parameters of the particle, it is determined that the particle will not deposit energy in the mth layer voxel again, so the mth layer can be The voxels of the previous layers correspond to the cache space corresponding to each voxel in the deposition area to be released.
  • the scattering particles of the particles entering the nth layer are also called secondary particles, and may return to affect the deposited energy in their voxels Therefore, the buffer space corresponding to each voxel in the deposition area corresponding to each voxel between the nth layer and the mth layer is not released temporarily, so as to ensure that the deposition energy generated by the scattering particles can still be recorded.
  • the particles when the particles pass out of the simulation area, the particles will not generate deposition energy for the voxels in the simulation area. In this case, the buffer space corresponding to each voxel in the deposition area corresponding to voxels in other layers can be released.
  • the data stored in the cache space can be used as the energy contribution of the particle and/or its secondary particles generated by the corresponding voxels in the simulation area, and it can be dumped into the shared storage space.
  • the buffer space corresponding to each voxel in the voxel corresponding to the deposition area of each layer before the mth layer can be released, and the released data can be transferred to
  • the shared storage space not only ensures the accuracy of statistical results, but also ensures the flexible utilization of cache resources.
  • the embodiment of the present application also provides a possible implementation of calculating the uncertainty, so as to evaluate the reliability of the simulation process.
  • the radiotherapy dose determination method may also include:
  • the Nth particle refers to the last particle that releases the cache space
  • the deposition energy of the Nth particle refers to the statistics obtained by sequentially counting the deposition energy of each voxel of the N particles in the corresponding layer deposition area Statistical parameters of deposition energy.
  • the order of the N particles refers to the order in which the cache space is released and transferred to the shared storage space.
  • the data stored in the buffer space corresponding to each voxel in the deposition area of each layer is the deposition energy in each voxel in the deposition area of the particle corresponding to the layer, when the buffer space of the released data is the first particle in For the buffer space of each voxel in the corresponding layer deposition area, the deposition energy of each voxel in the corresponding layer deposition area of the l-th particle can be counted to obtain the deposition energy of the l-th particle, and the shared storage space The statistical energy is updated as the deposited energy of the lth particle. That is to say, the deposited energy of the lth particle in the shared storage space is the statistical parameter of deposited energy after being transferred into the lth particle.
  • the deposition energy of each voxel in the corresponding layer deposition area of the first particle can be counted to obtain the first Particle deposition energy, and the deposition energy of the first particle obtained by statistics is stored in the shared storage space.
  • the cache space of the released data is the cache space of each voxel in the corresponding layer deposition area of the number of particles greater than 1, according to the deposition energy of the particles stored in the shared storage space, and the particles in the currently released cache space are in the corresponding
  • the deposition energy of each voxel in the layer deposition area is calculated again to obtain the deposition energy of the current particle, and the deposition energy statistical parameters in the shared storage space are updated to the deposition energy of the current particle.
  • the deposition energy of the Nth particle stored in the shared storage space is, is the deposition energy of all the particles sampled, at this time, according to the deposition energy of the Nth particle, the uncertainty of the parallel Monte Carlo simulation for multiple particles can be calculated.
  • the deposited energy of the Nth particle when the deposited energy of the Nth particle is dumped in the shared storage space, that is, when there is the deposited energy of the last particle among the N particles, it can be based on the Nth particle , to calculate the uncertainty of the parallel Monte Carlo simulation, that is, only when the Monte Carlo simulation of all sampled particles ends, it is necessary to calculate the parallel Monte Carlo simulation based on the deposition energy of the Nth particle
  • the process of Monte Carlo simulation of particles it is only necessary to update the statistics of the deposition energy of new particles based on the buffer space of voxels in the corresponding layer deposition area of new particles in the released buffer space , and according to the update result, the deposition energy of the particles in the shared storage space can be updated without calculating the uncertainty.
  • the method of performing statistics to calculate the uncertainty, compared with the one-time calculation of the uncertainty after the deposition energy of all the particles is obtained, can effectively improve the stability of the calculation uncertainty and reduce the calculation uncertainty.
  • the rounding error ensures the
  • the deposition energy of particles in the shared storage space refers to the statistical value of the deposition energy of particles in voxels in the corresponding layer deposition area, so in the process of parallel Monte Carlo simulation, it can be Only the deposition energy of particles is counted. In the shared storage space, it is only necessary to allocate the corresponding buffer area to store the statistical value of the deposition energy. Only when the parallel Monte Carlo simulations are all over, based on the recorded The deposition energy of the particles is used to calculate the uncertainty of the parallel Monte Carlo simulation. Therefore, in the process of calculating the uncertainty, as few storage resources as possible can be used, that is, the online real-time statistics of the deposition energy can be guaranteed, and the use of As few storage resources as possible.
  • the embodiment of the present application also provides a possible implementation.
  • the parallel Monte Carlo simulation is calculated according to the deposition energy of the Nth particle in the shared storage space.
  • Uncertainties for Lowe simulations can include:
  • the deposition energy of the Nth particle in a single voxel is the average deposition dose of the Nth particle in each corresponding voxel
  • the energy variance of the Nth particle in a single voxel is: The deposition energy in each voxel and the variance sum of the average deposition energy.
  • the following formula (1) is used to calculate the uncertainty of the parallel Monte Carlo simulation.
  • N is the number of particles sampled, that is, the number of multiple particles, which is a positive integer greater than 1
  • u N is the deposition energy of the Nth particle in a single voxel
  • S N is the kth particle in a single voxel The energy variance of the element.
  • the deposition energy of the Nth particle includes: the deposition energy of the Nth particle in a single voxel, and the energy variance of the Nth particle in a single voxel, two methods for the Nth particle Energy statistical parameters, therefore, in the process of parallel Monte Carlo simulation, there are only two storage areas in the shared storage space to store these two energy statistical parameters, and two storage areas for these two energy statistical parameters
  • the storage areas can be the same size.
  • the process of updating the deposition energy of particles in the shared storage space that is, the process of updating the two energy statistical parameters of particles in the process of calculating the uncertainty, is illustrated by an example as follows.
  • the deposition energy u N of the Nth particle in a single voxel is u N-1
  • the deposition energy u N of the Nth particle in a single voxel can be calculated by using the following formula (2).
  • S N -1 can be expressed as the following formula (3).
  • the deposited energy of the N-1th particle in a single voxel is u N-1
  • the deposition energy u N of the Nth particle in a single voxel the following formula (4) is used to calculate the energy variance of the Nth particle in a single voxel.
  • the uncertainty of the parallel Monte Carlo simulation can be calculated according to the deposited energy of the Nth particle in a single voxel and the energy variance of the Nth particle in a single voxel, and, no matter Whether it is the deposited energy of the Nth particle in a single voxel, or the energy variance of the Nth particle in a single voxel, both are updated on the basis of the corresponding energy statistical parameters of the previous particle’s single voxel, which effectively guarantees The statistical accuracy and stability of the energy statistical parameters are improved, and the rounding error in the process of updating the deposited energy is effectively reduced, thereby effectively ensuring the accuracy of the uncertain calculation.
  • radiotherapy dose determination device, equipment, and storage medium provided by the present application for implementation are described below.
  • the specific implementation process and technical effects refer to the above, and will not be repeated below.
  • Fig. 6 is a schematic diagram of a radiotherapy dose determination device provided in an embodiment of the present application. As shown in Fig. 6, the radiotherapy dose determination device 600 may include:
  • a division module 601 configured to perform three-dimensional grid division on the simulation area to obtain multiple voxels
  • the simulation module 602 is used to perform parallel Monte Carlo simulation on multiple particles sampled in the simulation area to obtain simulation results of multiple particles;
  • the determination module 603 is configured to determine the deposition dose of each voxel according to the simulation result.
  • the simulation module 602 is specifically configured to, based on each particle, determine the incident angle of the corresponding particle entering the simulation area and the incident voxel; based on the incident angle and the incident voxel, allocate buffers for some voxels in the simulation area Space, and conduct Monte Carlo simulation until the particles pass through the simulation area to obtain the simulation results; the cache space is used to store the deposited energy of the particles in the corresponding voxels.
  • the simulation module 602 is specifically configured to sequentially determine the deposition area corresponding to each layer of voxels of particles in the simulation area based on the incident angle and the incident voxel; for each volume in the deposition area corresponding to each layer of voxel Allocate the corresponding cache space for the voxel; conduct Monte Carlo simulation based on the incident angle and the incident voxel until the particle passes through the simulation area, and store the deposited energy of the particle in the corresponding voxel in the cache space; The data is transferred to the shared storage space to obtain the simulation results.
  • the simulation module 602 is specifically configured to: according to the incident angle, determine the incident voxel and the voxels within the preset range around the incident voxel in the first layer of voxels as the particles corresponding to the first layer of voxels The first layer of deposition area; allocate corresponding buffer space for each voxel in the first layer of deposition area to store the deposition energy of each voxel in the first layer of deposition area; when the particles enter the next layer of voxels, the next The area corresponding to the deposition area of the first layer in a layer of voxels and the voxels in the surrounding preset range are determined as the second layer deposition area corresponding to the voxels of the next layer of particles; in the second layer deposition area Each voxel of each voxel is assigned a corresponding buffer space to store the deposition energy of each voxel in the second layer of deposition area; the deposition area when the particles enter each layer of vo
  • the simulation module 602 is specifically used for: when the particle enters the nth layer of the simulation area, release the cache space corresponding to each voxel in the voxel in the layer before the mth layer corresponding to each voxel in the deposition area; wherein, m It is a positive integer less than n; transfer the data stored in the released cache space to the shared storage space.
  • the radiotherapy dose determination device 600 may also include:
  • the calculation module is used to calculate the uncertainty of the parallel Monte Carlo simulation according to the deposition energy of the Nth particle in the shared storage space, where N is equal to the number of multiple particles.
  • the calculation module is specifically configured to calculate the uncertainty of the parallel Monte Carlo simulation according to the deposition energy of the Nth particle in a single voxel and the energy variance of the Nth particle in a single voxel.
  • the deposition energy of the Nth particle in a single voxel is the average deposition dose of the Nth particle in each corresponding voxel
  • the energy variance of the Nth particle in a single voxel is: The deposition energy in each voxel and the variance sum of the average deposition energy.
  • the above-mentioned device is used to implement the radiotherapy dose determination method provided by the foregoing embodiments, and its implementation principle and technical effect are similar, and will not be repeated here.
  • the above modules may be one or more integrated circuits configured to implement the above method, for example: one or more specific integrated circuits (Application Specific Integrated Circuit, referred to as ASIC), or, one or more microprocessors (digital singnal processor, DSP for short), or, one or more Field Programmable Gate Arrays (Field Programmable Gate Array, FPGA for short), etc.
  • ASIC Application Specific Integrated Circuit
  • DSP digital singnal processor
  • FPGA Field Programmable Gate Array
  • the processing element may be a general-purpose processor, such as a central processing unit (Central Processing Unit, referred to as CPU) or other processors that can call program codes.
  • CPU central processing unit
  • these modules can be integrated together and implemented in the form of a system-on-a-chip (SOC for short).
  • FIG. 7 is a schematic diagram of a computer device provided by an embodiment of the present application.
  • a computer device 700 includes: a memory 701 and a processor 702 .
  • the memory 701 and the processor 702 are connected via a bus.
  • the memory 701 stores a computer program executable by the processor 702, and when the processor 702 executes the computer program, the above method embodiments can be implemented. The specific implementation manner and technical effect are similar, and will not be repeated here.
  • an embodiment of the present application may also provide a computer-readable storage medium for implementing the above radiotherapy dose determination method, which may be a non-volatile storage medium on which a computer program is stored , the computer program can execute the above method embodiments when being read and executed.
  • the disclosed devices and methods can be implemented in other ways.
  • the device embodiments described above are only illustrative.
  • the division of the units is only a logical function division. In actual implementation, there may be other division methods.
  • multiple units or components can be combined or May be integrated into another system, or some features may be ignored, or not implemented.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
  • each functional unit in each embodiment of the present invention may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit.
  • the above-mentioned integrated units can be implemented in the form of hardware, or in the form of hardware plus software functional units.
  • the above-mentioned integrated units implemented in the form of software functional units can be stored in a computer-readable storage medium.
  • the above-mentioned software functional units are stored in a storage medium, and include several instructions to enable a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (English: processor) to execute the program described in each embodiment of the present invention. part of the method.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (English: Read-Only Memory, abbreviated: ROM), random access memory (English: Random Access Memory, abbreviated: RAM), magnetic disk or optical disc, etc.

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Public Health (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Animal Behavior & Ethology (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Veterinary Medicine (AREA)
  • Surgery (AREA)
  • Urology & Nephrology (AREA)
  • Epidemiology (AREA)
  • Medical Informatics (AREA)
  • Primary Health Care (AREA)
  • Apparatus For Radiation Diagnosis (AREA)

Abstract

本申请实施例提供一种放疗剂量确定方法、装置及存储介质，涉及医疗信息技术领域，方法包括：对模拟区域进行三维网格划分，得到多个体素；在所述模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到所述多个粒子的模拟结果；根据所述模拟结果，确定各个体素的沉积剂量。本申请可有效缩短整个放疗模拟过程中确定放疗剂量的耗时，提高蒙特卡洛模拟的效率。

Description

放疗剂量确定方法、装置、设备及存储介质 技术领域
本申请涉及医疗信息技术，尤其涉及一种放疗剂量确定方法、装置、设备及存储介质。
背景技术
为保证放疗精度，在为目标对象进行治疗之前，可通过进行放疗模拟，得到目标对象的剂量分布，以此作为参照为目标对象制定匹配的放疗计划。
相关的技术中，在进行放疗模拟的过程中，大多是采用串行蒙特卡洛模拟方法。其中，串行蒙特卡洛模拟，指的是，针对抽样到的多个粒子依次进行模拟，只有在一个粒子模拟结束之后，才可进行下一个粒子的模拟。
串行蒙特卡洛模拟所需时间较长，使得制定治疗计划的时间也相应延长，从而增加了整个治疗过程的时间。
发明内容
本申请实施例提供一种放疗剂量确定方法、装置、设备及存储介质，以缩短放疗模拟过程的整个时长，缩短制定治疗计划的时间，从而缩短整个治疗过程的时间。
第一方面，本申请实施例提供一种放疗剂量确定方法包括：
对模拟区域进行三维网格划分，得到多个体素；
在所述模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到所述多个粒子的模拟结果；
根据所述模拟结果，确定各个体素的沉积剂量。
在一种可能的实现示例中，所述在所述模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到所述多个粒子的模拟结果，包括：
基于每一个粒子，确定相应所述粒子进入所述模拟区域的入射角度以及入射体素；
基于所述入射角度和所述入射体素，为所述模拟区域中的部分体素分配缓存空间，并进行蒙特卡洛模拟直至所述粒子穿出所述模拟区域，得到模拟结果；所述缓存空间用于存储所述粒子在相应体素中的沉积能量。
在一种可能的实现示例中，所述基于所述入射角度和所述入射体素，为所述模拟区域中的部分体素分配缓存空间，并进行蒙特卡洛模拟直至所述粒子穿出所述模拟区域，得到 模拟结果，包括：
基于所述入射角度和所述入射体素，依次确定所述粒子在所述模拟区域中每一层体素对应的沉积区域；
为所述每一层体素对应的沉积区域中的各体素分配对应的缓存空间；
基于所述入射角度和所述入射体素进行蒙特卡洛模拟，直至所述粒子穿出所述模拟区域，并在所述缓存空间中存储所述粒子在相应体素中的沉积能量；
将所述缓存空间内的数据转存至共享存储空间，得到模拟结果。
在一种可能的实现示例中，所述基于所述入射角度和所述入射体素，依次确定所述粒子在所述模拟区域中每一层体素对应的沉积区域，为所述每一层体素对应的沉积区域中的各体素分配对应的缓存空间，包括：
根据所述入射角度，将所述入射体素以及所述第一层体素中所述入射体素周边预设范围内的体素，确定为所述粒子在所述第一层体素对应的第一层沉积区域；
为所述第一层沉积区域中的各体素分配对应的缓存空间，以存储所述第一层沉积区域中各体素的沉积能量；
当所述粒子进入下一层体素时，将所述下一层体素中与所述第一层沉积区域对应的区域以及周边预设范围内的体素，确定为所述粒子在所述下一层体素对应的第二层沉积区域；
为所述第二层沉积区域中的各体素分配对应的缓存空间，以存储所述第二层沉积区域中各体素的沉积能量；
依次确定所述粒子进入每一层体素时的沉积区域，并分配对应缓存空间以存储对应沉积区域内各体素的沉积能量。
在一种可能的实现示例中，所述将所述缓存空间内的数据转存至共享存储空间，得到模拟结果，包括：
当所述粒子进入所述模拟区域的第n层时，将第m层之前的各层体素对应沉积区域中各体素对应的缓存空间进行释放；其中，m为小于n的正整数；
将释放的缓存空间内存储的数据，转存至所述共享存储空间。
在一种可能的实现示例中，所述方法还包括：
根据所述共享存储空间中第N个粒子的沉积能量，计算所述并行蒙特卡洛模拟的不确定度，其中，N等于所述多个粒子的数量。
在一种可能的实现示例中，所述根据所述共享内存空间中第N个粒子的沉积能量，计算所述并行蒙特卡洛模拟的不确定度，包括：
根据所述第N个粒子在单个体素中的沉积能量，以及所述第N个粒子在单个体素的能量方差，计算所述并行蒙特卡洛模拟的不确定度；
其中，所述第N个粒子在单个体素中的沉积能量为所述第N个粒子在对应的各体素中的平均沉积剂量，所述第N个粒子在单个体素的能量方差为：所述第N个粒子在对应的各体素中的沉积能量和所述平均沉积能量的方差和。
第二方面，本申请实施例还可提供一种放疗剂量确定装置，包括：
划分模块，用于对模拟区域进行三维网格划分，得到多个体素；
模拟模块，用于在所述模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到所述多个粒子的模拟结果；
确定模块，用于根据所述模拟结果，确定各个体素的沉积剂量。
第三方面，本申请实施例还提供一种计算机设备，包括：存储器和处理器，所述存储器存储有所述处理器可执行的计算机程序，所述处理器执行所述计算机程序时实现上述第一方面中任一所述的放疗剂量确定方法。
第四方面，本申请实施例还提供一种计算机可读存储介质，所述存储介质上存储有计算机程序，所述计算机程序被读取并执行时，实现上述第一方面任一所述的放疗剂量确定方法。
本申请实施例提供的放疗剂量确定方法、装置、设备及存储介质，可通过对模拟区域进行三维网格划分，得到多个体素，并在模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到多个粒子的模拟结果，继而根据模拟结果，确定各个体素的沉积剂量。也就是说，本申请实施例提供的放疗剂量确定方法，实际为通过并行蒙特卡洛模拟方法确定放疗剂量的方法，针对多个粒子的蒙特卡洛模拟方法是并行执行的，有效缩短了整个放疗模拟过程中确定放疗剂量的耗时，提高了蒙特卡洛模拟的效率，缩短了制定治疗计划的时间，从而缩短整个治疗过程的时间。
附图说明
为了更清楚地说明本发明实施例的技术方案，下面将对实施例中所需要使用的附图作简单地介绍，应当理解，以下附图仅示出了本发明的某些实施例，因此不应被看作是对范围的限定，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他相关的附图。
图1为本申请实施例提供的一种放疗剂量确定方法的流程图；
图2为本申请实施例提供的一种放疗剂量确定方法中并行蒙特卡洛模拟的方法流程图；
图3为本申请实施例提供的一种放疗剂量确定方法中一个蒙特卡洛模拟过程的方法流程图一；
图4为本申请实施例提供的一种放疗剂量确定方法中一个蒙特卡洛模拟过程的方法流程图二；
图5为本申请实施例提供的一种放疗剂量确定方法中缓存空间中数据转存的方法流程图；
图6为本申请实施例提供的一种放疗剂量确定装置的示意图；
图7为本申请实施例提供的一种计算机设备的示意图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。
在本申请的描述中，需要理解的是，术语“第一”、“第二”仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括一个或者更多个所述特征。在本申请的描述中，“多个”的含义是两个或两个以上，除非另有明确具体的限定。
在本申请中，“示例性”一词用来表示“用作例子、例证或说明”。本申请中被描述为“示例性”的任何实施例不一定被解释为比其它实施例更优选或更具优势。为了使本领域任何技术人员能够实现和使用本申请，给出了以下描述。在以下描述中，为了解释的目的而列出了细节。应当明白的是，本领域普通技术人员可以认识到，在不使用这些特定细节的情况下也可以实现本申请。在其它实例中，不会对公知的结构和过程进行详细阐述，以避免不必要的细节使本申请的描述变得晦涩。因此，本申请并非旨在限于所示的实施例，而是与符合本申请所公开的原理和特征的最广范围相一致。
在描述本申请实施例提供的放疗剂量确定方法之前，如下先对本申请实施例所涉及的技术术语进行解释。
蒙特卡洛模拟：作为放射治疗领域最为精确的算法，其可采用预设概率分布模型对射源模型的发射粒子进行抽样，然后按照微观粒子运动的统计规律，模拟粒子在待模拟区域中的运动过程，得到抽样粒子在待模拟区域内的辐射剂量分布。
与传统技术中，采用串行蒙特卡洛模拟方法进行放疗模拟的过程相比，本申请实施例提供一种采用并行蒙特卡洛模拟方法进行放疗模拟，以确定放疗计划的多种可能实现方式。
执行本申请实施例提供的放疗剂量确定方法的设备可以为安装有放疗剂量确定应用的计算机设备，计算机设备可通过运行放疗剂量确定应用，执行对应的放疗剂量确定方法。放疗剂量确定应用可以为放疗计划系统(Treatment PlanSystem,TPS)的一个子功能模块，其还可称为蒙特卡洛计算模块。
如下结合多个示例，首先对本申请实施例所提供的放疗剂量确定方法进行示例的解释说明。图1为本申请实施例提供的一种放疗剂量确定方法的流程图。如图1所示，放疗剂量确定方法可包括：
S101、对模拟区域进行三维网格划分，得到多个体素。
其中，模拟区域又称计算区域，其可以为目标对象的预设三维区域，目标对象可以为待治疗对象或者预设模体对象。由于模拟区域实际是三维区域，因此，对模拟区域的划分可以是在三维空间内，根据预设的单个体素的大小，对模拟区域进行三维网格的划分，得到多个体素，使得模拟区域在预设三维坐标系中三个坐标轴上分别具有a、b、c个体素。预设三维坐标系例如可以为xyz坐标系，三个坐标轴依次可以为x坐标轴、y坐标轴以及z坐标轴。
S102、在模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到多个粒子的模拟结果。
在进行并行蒙特卡洛模拟之前，可先根据预设概率分布模型，对射源模块的发射粒子进行采样，得到采样的多个粒子。其中，预设概率分布模型例如可以为随机概率分布模型，或者其他概率分布模型，本申请实施例不对此进行限制。在获取到采样的多个粒子的情况下，可在模拟区域内对多个粒子进行并行蒙特卡洛模块，在对每个粒子进行蒙特卡洛模拟的过程中，对应粒子在模拟区域内的运动进行模拟，并在运动模拟的过程中，计算对应粒子在各体素上的沉积能量，从而得到对应粒子的模拟结果。
S103、根据模拟结果，确定各个体素的沉积剂量。
每个粒子的模拟结果中可包括有对应粒子在各体素上的沉积能量，那么在可能的实现方式中，可根据多个粒子的模拟结果，确定各个体素的沉积剂量，如此便得到了模拟区域内各个体素的辐射剂量分布。
示例的，在根据多个粒子的模拟结果，确定各个体素的沉积剂量的过程中，可将多个粒子的模拟结果中，针对同一体素的沉积能量进行合并，得到对应体素的沉积剂量。
本申请实施例提供的放疗剂量确定方法，可通过对模拟区域进行三维网格划分，得到多个体素，并在模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到多个粒子的模 拟结果，继而根据模拟结果，确定各个体素的沉积剂量。也就是说，本申请实施例提供的放疗剂量确定方法，实际为通过并行蒙特卡洛模拟方法确定放疗剂量的方法，针对多个粒子的蒙特卡洛模拟方法是并行执行的，有效缩短了整个放疗模拟过程中确定放疗剂量的耗时，提高了蒙特卡洛模拟的效率，缩短了制定治疗计划的时间，从而缩短整个治疗过程的时间。
在本申请上述实施例提供的放疗剂量确定方法的基础上，本申请还通过多个实施例提供了针对多个粒子的并列蒙特卡洛模拟的可能实现方式。图2为本申请实施例提供的一种放疗剂量确定方法中并行蒙特卡洛模拟的方法流程图。如图2所示，如上所示的S102中在模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到多个粒子的模拟结果，可包括：
S201、基于每一个粒子，确定相应粒子进入模拟区域的入射角度以及入射体素。
在采样到多个粒子的情况下，即确定了采样的各粒子从射源模型的出射角度，因此，针对每个粒子，可基于采样过程中确定的对应粒子的出射角度，以及模拟区域相对射源模型的放置位置，确定相应粒子进入模拟区域的入射角度以及入射体素。该入射体素即为粒子从模拟区域外，在模拟区域内相遇的体素，也可以称为粒子在模拟区域内运动轨迹上经过的体素。
为方便描述，假设粒子入射角度为z坐标的负方向，则可确定粒子进入模拟区域内第一个相遇的体素为z坐标上沿着正方向的第c个体素，因此，则可确定入射体素包括第一个相遇的体素v(i，j，c)以及从第一个相遇的体素穿出后入射的其它体素。
S202、基于入射角度和入射体素，为模拟区域中的部分体素分配缓存空间，并进行蒙特卡洛模拟直至粒子穿出模拟区域，得到模拟结果。
缓存空间用于存储粒子在相应体素中的沉积能量。
在具体实现过程中，基于上述确定的粒子的入射角度和入射体素，进行蒙特卡洛模拟，在模拟过程中为模拟区域中的部分体素分别分配对应的缓存空间，直至粒子运动穿过模拟区域，则针对粒子的蒙特卡洛模拟过程完成，得到模拟结果。其中，在模拟过程中的部分体素可以为：粒子在模拟区域中运动轨迹相关联的体素，其例如可包括：模拟区域中粒子的运动轨迹穿过的体素，以及，模拟区域中粒子的运动轨迹周围预设范围内的体素等。
为模拟区域中部分体素分配缓存空间，指的是，为模拟区域中部分体素中各体素分配对应的一个缓存空间，以存储粒子在相应体素中的沉积能量。在可能的实现示例中，每个缓存空间的大小例如可以为预设字节数的缓存区，如四字节浮点数缓冲区。
假设，粒子在模拟区域中的部分体素包括：k个体素，在此过程中，可分配k个缓存空间，以分别存储粒子在k个体素中的沉积能量。
本申请实施例提供的放疗剂量确定方法，可针对每个粒子确定其进入模拟区域的入射角度以及入射体素，并基于入射角度和入射体素，为模拟区域中的部分体素分配缓存空间，并进行蒙特卡洛模拟直至粒子穿出模拟区域，得到模拟结果，分配的缓存空间用于在模拟过程中记录粒子在相应体素中的沉积能量。也就是说，该方法在采用并行蒙特卡洛模拟进行放疗剂量确定的过程中，针对每个粒子的蒙特卡洛模拟，并非是针对模拟区域中的所有体素分配对应的缓存空间，而仅是针对模拟区域中的与粒子关联的部分体素分配对应的缓存空间以存储粒子在相应体素中的沉积能量，有效减小了并行蒙特卡洛模拟中针对每个粒子的蒙特卡洛模拟过程所需的内存资源，从而保证了针对多个粒子的蒙特卡洛模拟的并行效率，从而有效缩短了放疗剂量确定的整体计算时间，缩短了制定治疗计划的时间，从而缩短整个治疗过程的时间。
如下继续结合附图通过对针对一个粒子的蒙特卡洛模拟过程的多种实现示例进行示例解释说明，提供放疗剂量确定方法的多种可能实现方式。图3为本申请实施例提供的一种放疗剂量确定方法中一个蒙特卡洛模拟过程的方法流程图一。如图3所示，如上所示的S202中基于入射角度和入射体素，为模拟区域中的部分体素分配缓存空间，并进行蒙特卡洛模拟直至粒子穿出模拟区域，得到模拟结果可包括：
S301、基于入射角度和入射体素，依次确定粒子在模拟区域中每一层体素对应的沉积区域。
在具体执行过程中，可基于入射角度和入射体素，在模拟区域内对粒子进行运行模拟，继而根据运动模拟的结果，可依次确定粒子在模拟区域中的各层体素对应的沉积区域。沉积区域，指的是具有粒子的沉积能量的多个体素构成的三维区域。其中，运行模拟的结果例如可包括：粒子在模拟区域内的运动轨迹，以及粒子在运动过程中产生的次级粒子在模拟区域内的运动轨迹，而粒子在模拟区域内的运动轨迹可通过粒子穿过的体素的坐标位置进行表示，相应的，次级粒子在模拟区域内的运动轨迹可通过次级粒子穿过的体素的坐标位置进行表示。
S302、为每一层体素对应的沉积区域中的各体素分配对应的缓存空间。
在确定每一层体素对应的沉积区域之后，可分别为每一层体素对应的沉积区域中各体素分配对应的缓存空间，即，为每一层体素对应的沉积区域中的每个体素划分一个缓存空间。不同层体素对应的沉积区域中的体素数量可能不同，因此，针对不同层体素对应的沉 积区域中的体素所分配的缓存空间的数量则不同。
其中，每一层体素对应的沉积区域内的体素数量可以是根据模拟区域内的体素密度、介质类型、粒子类型、粒子进入每一层体素的能量、入射角度、介质类型对应的自由程等至少一项参数，结合粒子的物理运动规律所确定的数值。也就是说，当粒子在每一层体素中运动时，无需为第一层体素中的所有体素分配缓存空间，而是为与粒子在每一层体素中的运动规律相关联的部分体素，也即在第一层体素中确定每一层体素对应的沉积区域中的体素分配相应的缓存空间即可，大大减小了粒子在模拟区域内每一层体素内运动时，所需的缓存空间的数量。
由于模拟区域内不同层体素的密度、介质类型、粒子进入本层体素的能量、入射角度、介质中的自由程等均可能不同，因此，模拟区域内不同体素层对应的沉积区域内的体素可能不同。
假设，第一层体素对应的沉积区域中包括：k1个体素，则可为第一层体素对应的沉积区域中的k1个体素分配k1个缓存空间，以分别存储记录粒子在k1个体素中的沉积能量；而第二层体素对应的沉积区域中包括：k2个体素，则可为第二层体素对应的沉积区域中的k2个体素分配k2个缓存空间，以分别存储记录粒子在k2个体素中的沉积能量，以此类推。
需要说明的是，分配缓存空间是在进入某一层体素，确定出对应层体素对应的沉积区域的情况下，即可执行，在未进入新的体素层时，由于无法确定相应的沉积区域，因此，并不会分配缓存空间。上述举例，仅为方便理解，各层体素对应的沉积区域中的体素，以及对应分配的缓存空间之间的关系，所给出的示例。
S303、基于入射角度和入射体素进行蒙特卡洛模拟，直至粒子穿出模拟区域，并在缓存空间中存储粒子在相应体素中的沉积能量。
当粒子从某一层体素中穿出，需要继续基于入射角度和入射体素进行蒙特卡洛模拟，直至粒子穿出整个模拟区域，则完成粒子在模拟区域中的蒙特卡洛模拟，在模拟的过程中，将模拟得到的粒子在本层体素中相应体素中的沉积能量存储至各体素的缓存空间中。
其中，为每一层沉积区域中各体素分配的缓存空间可用于存储粒子和/或粒子的次级粒子在相应体素中的沉积能量。
S304、将缓存空间内的数据转存至共享存储空间，得到模拟结果。
在可能的实现方式中，可在确定粒子在模拟区域中的模拟过程完成之后，将针对相应粒子的缓存空间中的数据均转存至共享存储空间中，从而得到模拟结果；也可模拟过程中将部分缓存空间内的数据转存至共享存储空间中，在模拟过程完成后，将其他缓存空间中 的数据转存至共享存储空间，从而得到模拟结果。其中，共享存储空间为针对多个粒子的多个并行蒙特卡洛过程的共享内存空间。
本实施例提供的方法中，可通过依次确定粒子在模拟区域中每一层体素对应的沉积区域，并为每一层体素对应的沉积区域中的各体素分配对应的缓存空间，针对粒子进行蒙特卡洛模拟，直至粒子穿出模拟区域，并在缓存空间中存储粒子在相应体素中的沉积能量，继而将缓存空间中的数据转存至共享存储空间中得到粒子的模拟结果，可使得针对每一个粒子在蒙特卡洛模拟过程中的沉积区域的确定以及针对沉积区域中各体素的缓存空间的分配，对模拟区域中的部分体素分配缓存空间提供了清楚的示例，有效保证了针对每个粒子的蒙特卡洛模拟过程中，无需针对所有的体素分配缓存空间，有效减小了针对每个粒子在蒙特卡洛模拟过程中内存资源的消耗量，从而保证针对多个粒子的并行蒙特卡洛模拟过程的并行执行效率。
在上述实施例所提供方案的基础上，本申请实施例还提供了针对一个粒子的一个蒙特卡洛模拟过程的其他可能实现方式。图4为本申请实施例提供的一种放疗剂量确定方法中一个蒙特卡洛模拟过程的方法流程图二。如图4所示，如上所示的S301中基于入射角度和入射体素，依次确定粒子在模拟区域中每一层体素对应的沉积区域，以及S302中为每一层体素对应的沉积区域中的各体素分配对应的缓存空间，可以包括：
S401、根据入射角度，将入射体素以及第一层体素中入射体素周边预设范围内的体素，确定为粒子在第一层体素对应的第一层沉积区域。
在可能的实现示例中，可根据入射角度，确定第一层体素中入射体素周围预设范围内的体素，并将入射体素，以及入射体素周边预设范围内的体素确定为粒子第一层体素对应的第一层沉积区域。也就是说，第一层沉积区域内的体素包括：入射体素、以及入射体素周边预设范围内的体素，其中，入射体素可以为第一层体素中粒子和/或次级粒子的运动轨迹穿过的体素，入射体素周围预设范围内的体素可以是第一层体素中相应粒子的运动轨迹周边预设范围的体素，周边预设范围的大小可以是根据入射角度确定的以入射体素为基准点的三维区域范围。
S402、为第一层沉积区域中的各体素分配对应的缓存空间，以存储第一层沉积区域中各体素的沉积能量。
继续结合上述示例，若粒子的入射角度为z坐标的负方向，当粒子进入第一层体素的第一个体素v(i，j，c)时，可为入射体素分配一个缓存空间，以存储粒子在入射体素内的沉积能量；当粒子运动到第一层体素中z坐标仍为c的第一相邻体素v(i+1，j，c)时， 则可为第一相邻体素分配另一个缓存空间，以存储第一相邻体素内的沉积能量；当粒子继续运动到第一层体素中z坐标仍为c的第二相邻体素v(i+1，j+1，c)时，可为第二相邻体素分配另一个缓存空间，以存储第二相邻体素内的沉积能量。
假设，在第一层体素中粒子的运动轨迹经过了z坐标为c的k11个体素，第一层体素中以粒子的运动轨迹周围的预设范围的z坐标为c体素为k12个体素，则可确定第一层体素对应的沉积区域包括：运动轨迹经过的k11个体素，以及运动轨迹周围的预设范围的k12个体素。那么，可为第一层体素对应的沉积区域分配k1个缓存空间，以分别存储k1个体素中的沉积能量，其中，k1＝k11+k12。在本示例中，第一层体素可以是z坐标为c的一层体素，因此，从第一层体素中确定的第一层体素对应的第一层沉积区域中的各体素的z坐标也均为c。
S403、当粒子进入下一层体素时，将下一层体素中与第一层沉积区域对应的区域以及周边预设范围内的体素，确定为粒子在下一层体素对应的第二层沉积区域。
当粒子进入下一层体素，也就是从z坐标为c的第一层体素进入z坐标为c-1的下一层体素时，可确定下一层体素中的粒子运动轨迹上的体素为下一层体素中与第一层沉积区域对应的区域，并结合确定粒子运动轨迹周围预设范围内的体素，确定为粒子在下一层体素对应的第二层沉积区域。其中，粒子运动轨迹指的是粒子和/或粒子产生的次级粒子的运动轨迹。也就是说，第二层沉积区域包括：粒子运动轨迹穿过的体素，以及粒子运动轨迹周围预设范围内的体素。
S404、为第二层沉积区域中的各体素分配对应的缓存空间，以存储第二层沉积区域中各体素的沉积能量。
为第二层沉积区域中各体素分配对应的缓存空间的具体实现，可以与上述第一层沉积区域中各体素分配缓存空间的实现过程类似，在此不再赘述。
当粒子从z坐标为c的第一层体素进入z坐标为c-1的下一层体素时，且，z坐标为c-1的下一层体素中对应的第二层沉积区域内包括k2个体素，则需为第二层沉积区域分配k2个体素，以存储k2个体素中的沉积能量。
S405、依次确定粒子进入每一层体素时的沉积区域，并分配对应缓存空间以存储对应沉积区域内各体素的沉积能量。
由于粒子在模拟区域的运动是连续进行的，因此，在粒子从第二层体素中穿出后，继续对粒子进行模拟，并依次确定粒子进入每一层体素时的沉积区域，直至粒子穿出模拟区域。需要说明的是，在粒子从上一层体素进入下一层体素时确定下一层体素时的沉积区域 的具体实现过程，与上述粒子从第一层体素进入第二层体素时确定第二层沉积区域的实现过程类似，具体参照上述，在此不再赘述。在模拟过程中，为每一层体素对应的沉积区域中的体素分配对应的缓存空间，以存储对应沉积区域内各体素的沉积能量。
本实施例提供的方法，通过对粒子在每一层体素中对应的沉积区域进行确定，使得各层体素对应沉积区域内的部分体素更准确，保证了每一层体素对应沉积区域的准确确定，保证了针对每一层体素对应沉积区域中体素分配的缓存空间的数量的准确性，有效减小了针对每个粒子在模拟区域内进行蒙特卡洛模拟过程中分配的缓存空间数量的消耗量，从而有效保证针对多个粒子的并行蒙特卡洛模拟过程的并行执行效率。
在上述实施例所提供方案的基础上，本申请实施例还提供了缓存空间中的数据转存至共享存储空间的可能实现方式。图5为本申请实施例提供的一种放疗剂量确定方法中缓存空间中数据转存的方法流程图。如图5所示，如上所示方法中的S304中将缓存空间内的数据转存至共享存储空间，得到模拟结果可以包括：
S501、当粒子进入模拟区域的第n层时，将第m层之前的各层体素对应沉积区域中各体素对应的缓存空间进行释放。
其中，m为小于n的正整数。由于粒子在传输过程中会发生散射或运动方向的偏转，粒子会以极小的概率多次进入同一体素，为保证统计结果的准确性，缓冲区的层数不能太少，更不能随意设定，否则会提高统计结果出错的概率，因此，在本实施例中，可根据模拟区域的体素参数以及粒子的属性参数确定m的具体数值。其中，体素参数可包括：体素密度、体素中的介质类型、介质类型对应的自由程中的至少一项参数，属性参数包括：粒子的类型、粒子的初始能量、粒子的入射角度等至少一项参数。
本实施例提供的方法，可最大程度地在可接受范围内缓冲体素中的沉积能量，也就是当粒子进入某一层时，粒子还会在之前层的预设范围内的体素中产生沉积能量，因此，不能在此情况下，直接将缓存空间中的数据释放。而是，当粒子进入模拟区域的第n层时，基于根据模拟区域的体素参数以及粒子的属性参数，确定粒子不会再次在第m层体素中沉积能量，因此，可将第m层之前的各层体素对应沉积区域中各体素对应的缓存空间进行释放。而对于第n-1层或者n-2层等第n层之前的预设范围内，进入第n层的粒子的散射粒子又称次级粒子，还可能返回从而影响其体素中的沉积能量，因此，针对第n层与第m层之间的各层体素对应沉积区域中各体素对应的缓存空间暂不进行释放，保证了散射粒子产生的沉积能量还能被记录。
但是，当粒子穿出模拟区域，粒子便不会对模拟区域中的体素产生沉积能量，在此情 况下，便可将其他层体素对应沉积区域中各体素对应的缓存空间进行释放。
S502、将释放的缓存空间内存储的数据，转存至共享存储空间。
当缓存空间中的数据释放后，便可将缓存空间中存储的数据作为粒子和/或其产生的次级粒子在模拟区域内相应体素的能量贡献，并其转存至共享存储空间内。
本实施例提供的方法中，由于可在粒子进入第n层时，将第m层之前的各层体素对应沉积区域中各体素对应的缓存空间进行释放，并将释放的数据转存至共享存储空间，即保证了统计结果的准确性，又保证了缓存资源的灵活利用。
由于放疗剂量确定过程中进行蒙特卡洛模拟的多个粒子是采样得到的，那么通过针对多个粒子进行并行蒙特卡洛模拟过程所确定的剂量必然存在不确定性，因此，在上述实施例提供的放疗剂量确定方法的基础上，本申请实施例还提供了一种计算不确定度的可能实现方式，以评定模拟过程的可靠程度。在本实施例中，放疗剂量确定方法还可包括：
根据共享存储空间中第N个粒子的沉积能量，计算并行蒙特卡洛模拟的不确定度，其中，N为多个粒子的数量。
其中，第N个粒子指的是最后一个释放缓存空间的粒子，第N个粒子的沉积能量指的是，依次对N个粒子在相应层沉积区域中的各体素的沉积能量进行统计得到的沉积能量统计参数。N个粒子的先后次序指的是，释放缓存空间并转存至共享存储空间的次序。
在可能的实现示例中，每层沉积区域中各体素对应的缓存空间中存储的数据为粒子对应层沉积区域中各体素中的沉积能量，当释放数据的缓存空间为第l个粒子在相应层沉积区域中各体素的缓存空间，则可对第l个粒子在相应层沉积区域中各体素的沉积能量进行统计，得到第l个粒子的沉积能量，并将共享存储空间中的统计能量更新为第l个粒子的沉积能量。也就是说，共享存储空间中第l个粒子的沉积能量即为转存入第l个粒子后的沉积能量统计参数。
当释放数据的缓存空间为第1个粒子在相应层沉积区域中各体素的缓存空间，则可对第1个粒子在相应层沉积区域中各体素的沉积能量进行统计，得到第1个粒子的沉积能量，并将统计得到的第1个粒子的沉积能量存入至共享存储空间中。当释放数据的缓存空间为大于1的数量个粒子在相应层沉积区域中各体素的缓存空间，则可根据共享存储空间中存储粒子的沉积能量，以及当前释放的缓存空间中的粒子在相应层沉积区域中各体素的沉积能量，重新进行沉积能量的统计，得到当前粒子的沉积能量，并对共享存储空间中的沉积能量统计参数更新为当前粒子的沉积能量。
那么，当l＝N，也就是采样的粒子数量时，则可确定多个粒子的并行蒙特卡洛模拟均 结束，在此情况下，共享存储空间中存储的第N个粒子的沉积能量，即为采样的所有粒子的沉积能量，此时，根据第N个粒子的沉积能量，便可计算得到针对多个粒子的并行蒙特卡洛模拟的不确定度。
本实施例提供的方法中，当共享存储空间中转存了第N个粒子的沉积能量的情况下，也就是就有N个粒子中最后一个粒子的沉积能量时，便可基于第N个粒子的沉积能量，计算并行蒙特卡洛模拟的不确定度，也就是说，只有在所有采样的粒子的蒙特卡洛模拟均结束时，才需基于第N个粒子的沉积能量，计算并行蒙特卡洛模拟的不确定度，而在粒子的蒙特卡洛模拟的过程中，只需基于释放的缓存空间中新的粒子在相应层沉积区域内体素的缓存空间进行新的粒子的沉积能量的统计更新，并根据更新结果，可对共享存储空间中的粒子的沉积能量进行更新，而无需进行不确定度的计算，因此，本实施例提供的方法实际是上一种采用在线方式对粒子的沉积能量进行统计用以计算不确定度的方法，相比较，在获取到的所有的粒子的沉积能量之后，一次性计算不确定度，可有效提高计算不确定度的稳定性，减少计算不确定过程中的舍入误差，保证了不确定度的计算准确度。
其次，由于本实施例的方法，共享存储空间中粒子的沉积能量指的是，针对粒子在相应层沉积区域中体素的沉积能量的统计值，因此在并行蒙特卡洛模拟的过程中，可仅对粒子的沉积能量进行统计，在共享存储空间中仅需分配对应的缓存区存储沉积能量的统计值即可，只有在并行蒙特卡洛模拟均结束的情况下，基于共享存储空间中记录的粒子的沉积能量，计算并行蒙特卡洛模拟的不确定度，因此，在计算不确定度的过程中，可采用尽可能少的存储资源，即保证了沉积能量的在线实时统计，还可保证采用尽可能少的存储资源。
针对并行蒙特卡洛模拟的不确定度的计算，本申请实施例还提供一种可能的实现方式，示例的，如上所示的根据共享存储空间中第N个粒子的沉积能量，计算并行蒙特卡洛模拟的不确定度，可包括：
根据第N个粒子在单个体素中的沉积能量，以及第N个粒子在单个体素的能量方差，计算并行蒙特卡洛模拟的不确定度。
其中，第N个粒子在单个体素中的沉积能量为第N个粒子在对应的各体素中的平均沉积剂量，第N个粒子在单个体素的能量方差为：第N个粒子在对应的各体素中的沉积能量和平均沉积能量的方差和。
示例的，根据第N个粒子在单个体素中的沉积能量，以及第N个粒子在单个体素的能量方差，采用下述公式(1)计算并行蒙特卡洛模拟的不确定度。
Figure PCTCN2021143835-appb-000001
其中，N为采样的粒子数量，即多个粒子的数量，其为大于1的正整数，u N为第N个粒子在单个体素中的沉积能量，S N为第k个粒子在单个体素的能量方差。
在本实施例中，由于第N个粒子的沉积能量包括：第N个粒子在单个体素中的沉积能量，以及第N个粒子在单个体素的能量方差，两种针对第N个粒子的能量统计参数，因此，在并行蒙特卡洛模拟的过程中，在共享存储空间中仅具有两个存储区分别用来存储这两种能量统计参数即可，存储这两种能量统计参数的两个存储区的大小可以相同。
如下通过示例对计算不确定度过程中，对共享存储空间中粒子的沉积能量，也就是针对粒子的这两种能量统计参数进行更新的过程进行示例解释说明。
假设，第N-1个粒子在单个体素中的沉积能量为u N-1，那么，当释放缓存空间为第N个粒子在相应层沉积区域中各体素的缓存空间，可根据第N个粒子在相应层沉积区域中各体素的缓存空间，计算第N个粒子在相应层沉积区域中各体素的沉积能量的平均值作为第N个粒子在单个体素中的贡献能量x N，如下，可采用下述公式(2)计算得到第N个粒子在单个体素中的沉积能量u N。
Figure PCTCN2021143835-appb-000002
假设第N-1个粒子在单个体素的能量方差为S N-1，S N-1可表示为下述公式(3)。
Figure PCTCN2021143835-appb-000003
如此，根据第N-1个粒子在单个体素的能量方差S N-1、第N个粒子在单个体素中的贡献能量x N，第N-1个粒子在单个体素中的沉积能量为u N-1、以及第N个粒子在单个体素中的沉积能量u N，采用下述公式(4)，计算第N个粒子在单个体素的能量方差。
Figure PCTCN2021143835-appb-000004
本实施例提供的方法中，可根据第N个粒子在单个体素中的沉积能量，以及第N个粒子在单个体素的能量方差，计算并行蒙特卡洛模拟的不确定度，并且，无论是第N个粒子在单个体素中的沉积能量，还是第N个粒子在单个体素的能量方差，均是在之前粒子的单个体素的相应能量统计参数的基础上更新的，其有效保证了能量统计参数的统计准确度、稳定性，有效减小了更新沉积能量过程中的舍入误差，从而有效保证了不确定计算的准确度。
下述对用以执行的本申请所提供的放疗剂量确定装置、设备及存储介质等进行说明，其具体的实现过程以及技术效果参见上述，下述不再赘述。
图6为本申请实施例提供的一种放疗剂量确定装置的示意图，如图6所示，放疗剂量确定装置600可包括：
划分模块601，用于对模拟区域进行三维网格划分，得到多个体素；
模拟模块602，用于在模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到多个粒子的模拟结果；
确定模块603，用于根据模拟结果，确定各个体素的沉积剂量。
可选的，模拟模块602，具体用于基于每一个粒子，确定相应粒子进入模拟区域的入射角度以及入射体素；基于入射角度和所述入射体素，为模拟区域中的部分体素分配缓存空间，并进行蒙特卡洛模拟直至粒子穿出模拟区域，得到模拟结果；缓存空间用于存储粒子在相应体素中的沉积能量。
可选的，模拟模块602，具体用于基于入射角度和入射体素，依次确定粒子在模拟区域中每一层体素对应的沉积区域；为每一层体素对应的沉积区域中的各体素分配对应的缓存空间；基于入射角度和入射体素进行蒙特卡洛模拟，直至粒子穿出所述模拟区域，并在缓存空间中存储粒子在相应体素中的沉积能量；将缓存空间内的数据转存至共享存储空间，得到模拟结果。
可选的，模拟模块602，具体用于：根据入射角度，将入射体素以及第一层体素中入射体素周边预设范围内的体素，确定为粒子在第一层体素对应的第一层沉积区域；为第一层沉积区域中的各体素分配对应的缓存空间，以存储第一层沉积区域中各体素的沉积能量；当粒子进入下一层体素时，将下一层体素中与第一层沉积区域对应的区域以及周边预设范围内的体素，确定为粒子在所述下一层体素对应的第二层沉积区域；为第二层沉积区域中的各体素分配对应的缓存空间，以存储第二层沉积区域中各体素的沉积能量；依次确定粒子进入每一层体素时的沉积区域，并分配对应缓存空间以存储对应沉积区域内各体素的沉积能量。
可选的，模拟模块602，具体用于：当粒子进入模拟区域的第n层时，将第m层之前的各层体素对应沉积区域中各体素对应的缓存空间进行释放；其中，m为小于n的正整数；将释放的缓存空间内存储的数据，转存至共享存储空间。
可选的，放疗剂量确定装置600还可包括：
计算模块，用于根据共享存储空间中第N个粒子的沉积能量，计算并行蒙特卡洛模拟的不确定度，其中，N等于多个粒子的数量。
可选的，计算模块，具体用于根据第N个粒子在单个体素中的沉积能量，以及第N个粒子在单个体素的能量方差，计算并行蒙特卡洛模拟的不确定度。
其中，第N个粒子在单个体素中的沉积能量为第N个粒子在对应的各体素中的平均沉积剂量，第N个粒子在单个体素的能量方差为：第N个粒子在对应的各体素中的沉积能量 和平均沉积能量的方差和。
上述装置用于执行前述实施例提供的放疗剂量确定方法，其实现原理和技术效果类似，在此不再赘述。
以上这些模块可以是被配置成实施以上方法的一个或多个集成电路，例如：一个或多个特定集成电路(Application Specific Integrated Circuit，简称ASIC)，或，一个或多个微处理器(digital singnal processor，简称DSP)，或，一个或者多个现场可编程门阵列(Field Programmable Gate Array，简称FPGA)等。再如，当以上某个模块通过处理元件调度程序代码的形式实现时，该处理元件可以是通用处理器，例如中央处理器(Central Processing Unit，简称CPU)或其它可以调用程序代码的处理器。再如，这些模块可以集成在一起，以片上系统(system-on-a-chip，简称SOC)的形式实现。
图7为本申请实施例提供的一种计算机设备的示意图。如图7所示，计算机设备700包括：存储器701、处理器702。存储器701和处理器702通过总线连接。
存储器701存储有处理器702可执行的计算机程序，处理器702在执行计算机程序时，可实现执行上述方法实施例。具体实现方式和技术效果类似，这里不再赘述。
在上述放疗剂量确定方法的基础上，本申请实施例还可提供一种用于执行上述放疗剂量确定方法的计算机可读存储介质，其可以为非易失性存储介质，其上存储有计算机程序，计算机程序可在被读取并执行时执行上述方法实施例。
在本发明所提供的几个实施例中，应该理解到，所揭露的装置和方法，可以通过其它的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本发明各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用硬件加软件功能单元的形式实现。
上述以软件功能单元的形式实现的集成的单元，可以存储在一个计算机可读取存储介 质中。上述软件功能单元存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)或处理器(英文：processor)执行本发明各个实施例所述方法的部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(英文：Read-Only Memory，简称：ROM)、随机存取存储器(英文：Random Access Memory，简称：RAM)、磁碟或者光盘等各种可以存储程序代码的介质。
上仅为本申请的具体实施方式，但本申请的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应以权利要求的保护范围为准。

Claims (10)

  1. 一种放疗剂量确定方法，其特征在于，包括：
    对模拟区域进行三维网格划分，得到多个体素；
    在所述模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到所述多个粒子的模拟结果；
    根据所述模拟结果，确定各个体素的沉积剂量。
  2. 根据权利要求1所述的方法，其特征在于，所述在所述模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到所述多个粒子的模拟结果，包括：
    基于每一个粒子，确定相应所述粒子进入所述模拟区域的入射角度以及入射体素；
    基于所述入射角度和所述入射体素，为所述模拟区域中的部分体素分配缓存空间，并进行蒙特卡洛模拟直至所述粒子穿出所述模拟区域，得到模拟结果；所述缓存空间用于存储所述粒子在相应体素中的沉积能量。
  3. 根据权利要求2所述的方法，其特征在于，所述基于所述入射角度和所述入射体素，为所述模拟区域中的部分体素分配缓存空间，并进行蒙特卡洛模拟直至所述粒子穿出所述模拟区域，得到模拟结果，包括：
    基于所述入射角度和所述入射体素，依次确定所述粒子在所述模拟区域中每一层体素对应的沉积区域；
    为所述每一层体素对应的沉积区域中的各体素分配对应的缓存空间；
    基于所述入射角度和所述入射体素进行蒙特卡洛模拟，直至所述粒子穿出所述模拟区域，并在所述缓存空间中存储所述粒子在相应体素中的沉积能量；
    将所述缓存空间内的数据转存至共享存储空间，得到模拟结果。
  4. 根据权利要求3所述的方法，其特征在于，所述基于所述入射角度和所述入射体素，依次确定所述粒子在所述模拟区域中每一层体素对应的沉积区域，为所述每一层体素对应的沉积区域中的各体素分配对应的缓存空间，包括：
    根据所述入射角度，将所述入射体素以及所述第一层体素中所述入射体素周边预设范围内的体素，确定为所述粒子在所述第一层体素对应的第一层沉积区域；
    为所述第一层沉积区域中的各体素分配对应的缓存空间，以存储所述第一层沉积区域中各体素的沉积能量；
    当所述粒子进入下一层体素时，将所述下一层体素中与所述第一层沉积区域对应的区域以及周边预设范围内的体素，确定为所述粒子在所述下一层体素对应的第二层沉积区域；
    为所述第二层沉积区域中的各体素分配对应的缓存空间，以存储所述第二层沉积区域中各体素的沉积能量；
    依次确定所述粒子进入每一层体素时的沉积区域，并分配对应缓存空间以存储对应沉积区域内各体素的沉积能量。
  5. 根据权利要求3所述的方法，其特征在于，所述将所述缓存空间内的数据转存至共享存储空间，得到模拟结果，包括：
    当所述粒子进入所述模拟区域的第n层时，将第m层之前的各层体素对应沉积区域中各体素对应的缓存空间进行释放；其中，m为小于n的正整数；
    将释放的缓存空间内存储的数据，转存至所述共享存储空间。
  6. 根据权利要求5所述的方法，其特征在于，所述方法还包括：
    根据所述共享存储空间中第N个粒子的沉积能量，计算所述并行蒙特卡洛模拟的不确定度，其中，N等于所述多个粒子的数量。
  7. 根据权利要求6所述的方法，其特征在于，所述根据所述共享内存空间中第N个粒子的沉积能量，计算所述并行蒙特卡洛模拟的不确定度，包括：
    根据所述第N个粒子在单个体素中的沉积能量，以及所述第N个粒子在单个体素的能量方差，计算所述并行蒙特卡洛模拟的不确定度；
    其中，所述第N个粒子在单个体素中的沉积能量为所述第N个粒子在对应的各体素中的平均沉积剂量，所述第N个粒子在单个体素的能量方差为：所述第N个粒子在对应的各体素中的沉积能量和所述平均沉积能量的方差和。
  8. 一种放疗剂量确定装置，其特征在于，包括：
    划分模块，用于对模拟区域进行三维网格划分，得到多个体素；
    模拟模块，用于在所述模拟区域内对采样的多个粒子进行并行蒙特卡洛模拟，得到所述多个粒子的模拟结果；
    确定模块，用于根据所述模拟结果，确定各个体素的沉积剂量。
  9. 一种计算机设备，其特征在于，包括：存储器和处理器，所述存储器存储有所述处理器可执行的计算机程序，所述处理器执行所述计算机程序时实现上述权利要求1-7任一项所述的放疗剂量确定方法。
  10. 一种计算机可读存储介质，其特征在于，所述存储介质上存储有计算机程序，所述计算机程序被读取并执行时，实现上述权利要求1-7中任一项所述的放疗剂量确定方法。
PCT/CN2021/143835 2021-12-31 2021-12-31 放疗剂量确定方法、装置、设备及存储介质 WO2023123420A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/143835 WO2023123420A1 (zh) 2021-12-31 2021-12-31 放疗剂量确定方法、装置、设备及存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/143835 WO2023123420A1 (zh) 2021-12-31 2021-12-31 放疗剂量确定方法、装置、设备及存储介质

Publications (1)

Publication Number Publication Date
WO2023123420A1 true WO2023123420A1 (zh) 2023-07-06

Family

ID=86997124

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/143835 WO2023123420A1 (zh) 2021-12-31 2021-12-31 放疗剂量确定方法、装置、设备及存储介质

Country Status (1)

Country Link
WO (1) WO2023123420A1 (zh)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6148272A (en) * 1998-11-12 2000-11-14 The Regents Of The University Of California System and method for radiation dose calculation within sub-volumes of a monte carlo based particle transport grid
CN105979873A (zh) * 2013-12-04 2016-09-28 医科达有限公司 用于基于连续材料指数的剂量计算的方法和系统
CN110504016A (zh) * 2018-05-18 2019-11-26 北京连心医疗科技有限公司 一种蒙特卡罗网格并行剂量计算方法、设备和存储介质
CN110556176A (zh) * 2018-06-01 2019-12-10 北京连心医疗科技有限公司 一种基于蒙特卡罗的剂量优化方法、设备和存储介质
CN112349383A (zh) * 2020-11-20 2021-02-09 林小惟 一种基于机器学习和蒙特卡罗算法的放射治疗剂量计算与验证方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6148272A (en) * 1998-11-12 2000-11-14 The Regents Of The University Of California System and method for radiation dose calculation within sub-volumes of a monte carlo based particle transport grid
CN105979873A (zh) * 2013-12-04 2016-09-28 医科达有限公司 用于基于连续材料指数的剂量计算的方法和系统
CN110504016A (zh) * 2018-05-18 2019-11-26 北京连心医疗科技有限公司 一种蒙特卡罗网格并行剂量计算方法、设备和存储介质
CN110556176A (zh) * 2018-06-01 2019-12-10 北京连心医疗科技有限公司 一种基于蒙特卡罗的剂量优化方法、设备和存储介质
CN112349383A (zh) * 2020-11-20 2021-02-09 林小惟 一种基于机器学习和蒙特卡罗算法的放射治疗剂量计算与验证方法

Similar Documents

Publication Publication Date Title
US11794038B2 (en) Proton beam system and methods for irradiating a target
JP3730514B2 (ja) 放射線量計算のためのシステムと方法
Wan Chan Tseung et al. A fast GPU‐based Monte Carlo simulation of proton transport with detailed modeling of nonelastic interactions
US8972227B2 (en) Computation of radiating particle and wave distributions using a generalized discrete field constructed from representative ray sets
Wang et al. A GPU‐accelerated Monte Carlo dose calculation platform and its application toward validating an MRI‐guided radiation therapy beam model
Wang et al. Development of a fast Monte Carlo dose calculation system for online adaptive radiation therapy quality assurance
JP2000507848A (ja) 全粒子モンテカルロトランスポートを使用する放射線療法線量の計算
CN110504016B (zh) 一种蒙特卡罗网格并行剂量计算方法、设备和存储介质
US9495513B2 (en) GPU-based fast dose calculator for cancer therapy
WO2007106815A2 (en) Methods and apparatus for hardware based radiation dose calculation
Pantaleo New track seeding techniques for the CMS experiment
CN108771794A (zh) 用于生成辐射处理计划中的剂量计算的系统和方法
Lemaréchal et al. GGEMS-Brachy: GPU GEant4-based Monte Carlo simulation for brachytherapy applications
WO2023123420A1 (zh) 放疗剂量确定方法、装置、设备及存储介质
Neylon et al. A nonvoxel‐based dose convolution/superposition algorithm optimized for scalable GPU architectures
US20220241615A1 (en) Methods and systems for particle based treatment using microdosimetry techniques
Traxler Real-time dilepton selection for the HADES spectrometer
Voss et al. Towards real time radiotherapy simulation
Voss et al. Towards real time radiotherapy simulation
CN103810670B (zh) 基于cuda流和共享内存的dvh图并行统计方法
Della-Giustina et al. Hybrid MPI+ OpenMP parallelization of image reconstruction in proton beam therapy on multi-core and many-core processors
Reid III Search for Soft Unclustered Energy Patterns in Proton-Proton Collisions with√ s= 13 TeV Recorded by the Compact Muon Solenoid Detector
CN115410685B (zh) 粒子剂量的确定方法、装置、设备、介质和程序产品
US11809518B1 (en) Methods, systems, and apparatuses for calculating global fluence for neutron and photon monte carlo transport using expected value estimators
Neph Accelerating Radiation Dose Calculation with High Performance Computing and Machine Learning for Large-scale Radiotherapy Treatment Planning

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21969766

Country of ref document: EP

Kind code of ref document: A1