WO2019140886A1 - 一种标准化云放疗计划方法、存储介质和系统 - Google Patents

一种标准化云放疗计划方法、存储介质和系统 Download PDF

Info

Publication number
WO2019140886A1
WO2019140886A1 PCT/CN2018/099445 CN2018099445W WO2019140886A1 WO 2019140886 A1 WO2019140886 A1 WO 2019140886A1 CN 2018099445 W CN2018099445 W CN 2018099445W WO 2019140886 A1 WO2019140886 A1 WO 2019140886A1
Authority
WO
WIPO (PCT)
Prior art keywords
radiotherapy
plan
standard
specific
radiotherapy plan
Prior art date
Application number
PCT/CN2018/099445
Other languages
English (en)
French (fr)
Inventor
李贵
Original Assignee
北京连心医疗科技有限公司
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 北京连心医疗科技有限公司 filed Critical 北京连心医疗科技有限公司
Priority to EP18900998.8A priority Critical patent/EP3742702B1/en
Priority to US16/645,839 priority patent/US20200203022A1/en
Publication of WO2019140886A1 publication Critical patent/WO2019140886A1/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
    • A61N5/103—Treatment planning systems
    • A61N5/1031—Treatment planning systems using a specific method of dose optimization
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
    • A61N5/00—Radiation therapy
    • 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
    • A61N5/103—Treatment planning systems
    • A61N5/1039—Treatment planning systems using functional images, e.g. PET or MRI
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • 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
    • 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
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
    • 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
    • G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/20—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
    • 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
    • G16H70/00—ICT specially adapted for the handling or processing of medical references
    • G16H70/20—ICT specially adapted for the handling or processing of medical references relating to practices or guidelines
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/05—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves 
    • A61B5/055—Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves  involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A61B6/02—Arrangements for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
    • A61B6/03—Computed tomography [CT]
    • A61B6/032—Transmission computed tomography [CT]
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus or devices for radiation diagnosis; Apparatus or devices for radiation diagnosis combined with radiation therapy equipment
    • A61B6/02—Arrangements for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
    • A61B6/03—Computed tomography [CT]
    • A61B6/037—Emission tomography
    • 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
    • A61N5/103—Treatment planning systems
    • A61N5/1031—Treatment planning systems using a specific method of dose optimization
    • A61N2005/1032—Genetic optimization methods
    • 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
    • A61N5/103—Treatment planning systems
    • A61N5/1031—Treatment planning systems using a specific method of dose optimization
    • A61N2005/1034—Monte Carlo type methods; particle tracking

Definitions

  • the invention belongs to the technical field of radiotherapy and cloud computing, and relates to a standardized cloud radiotherapy planning method, a storage medium and a system.
  • the brand effect of the top three hospitals has attracted a large number of patients to flock to the top three hospitals for treatment, especially the large top three hospitals are overcrowded, and the bed occupancy rate has been saturated for a long time.
  • the use rate of bed-level medical institutions is only about 60%, and some medical resources are wasted, and they do not exert the benefits they deserve.
  • a large number of patients are concentrated in the top three hospitals for treatment, which will inevitably lead to overburdened medical staff and nervous equipment and treatment beds, which can not meet the needs of all patients. Excessive concentration of quality medical resources in the top three hospitals has also led to a decline in medical service coverage. It is difficult for people in the area to enjoy the treatment of experienced experts.
  • the types of radiotherapy equipment in some hospitals are different, and may include several types of radiotherapy equipment of several brands. If the radiotherapy plan is completed according to a radiotherapy equipment in advance, the equipment fails unexpectedly and has the same model. When replacing the device, it will inevitably result in the shelving of the previously established radiotherapy plan, delaying the patient's treatment time; or the radiotherapy equipment of the currently established radiotherapy plan has been occupied and it takes a long time to queue, while other radiotherapy devices are idle or waiting in time. Shorter.
  • the present invention adopts the following technical solutions:
  • a standardized cloud radiotherapy planning method suitable for execution in a standardized cloud radiotherapy planning system includes the following steps:
  • the main control cloud server decomposes the computing task and assigns it to the controlled computer.
  • the controlled computer calculates the patient's radiotherapy plan using the standard radiotherapy device mode to generate a standard radiotherapy plan;
  • the master cloud server or the controlled computer converts and generates a specific radiotherapy plan that matches the particular radiotherapy device according to a standard radiotherapy plan, the match being that the generated specific radiotherapy plan can be performed in the corresponding specific radiotherapy device.
  • the patient image comprises one or a combination of a CT image, a nuclear magnetic image or a PET image.
  • the medical order data includes one or a combination of a target radiation dose, a DVH curve, and a dose of each organ radiation dose.
  • the sketch is automatically outlined, semi-automatically outlined or manually outlined.
  • the step of converting the generated radiotherapy plan that matches the specific radiotherapy device according to the standard radiotherapy plan according to the standard radiotherapy plan or the controlled computer further includes:
  • the recalculation includes one or a combination of dose calculation or reverse optimization; wherein the reverse optimization includes using direct subfield optimization (DAO) or flux map optimization method (FMO).
  • DAO direct subfield optimization
  • FMO flux map optimization method
  • Step (4) also includes a quality assurance (QA) step (5), the patient is verified by QA to verify that the specific radiotherapy plan converted is correct; if correct, the specific radiotherapy plan is executed; if not, return to step (4) ) Regenerate new specific radiotherapy plans based on standard radiotherapy plans.
  • QA quality assurance
  • step (3) and step (4) further comprising the step of selecting a radiotherapy device to be switched according to the use of the radiotherapy device: preferentially converting the standard radiotherapy plan into the currently idle radiotherapy device or the number of tasks to be performed is small The radiotherapy equipment or user-defined selection of the type of radiotherapy equipment to be converted.
  • the process queuing or user-defined settings are prioritized based on available computing resources.
  • the method before converting to generate a specific radiotherapy plan matching the specific radiotherapy device, the method further includes the following steps: comparing the standard radiotherapy device with the parameter of the specific radiotherapy device to be matched, if the parameter matching degree meets the preset threshold requirement
  • the standard radiotherapy plan will be directly used as the final radiotherapy plan, otherwise it will enter step (4).
  • the parameters that need to be compared between a standard radiotherapy device and a particular radiotherapy device to be matched include source parameters, multi-leaf collimator parameters.
  • the source parameters are obtained by comparing dose measurement characteristic data of the source in a uniform or non-uniform medium; the dose measurement characteristic data is obtained by a three-dimensional dose curve.
  • the multi-leaf collimator parameters include blade size and logarithm, maximum open field size, and whether staggering is allowed.
  • step (4) when the standard radiotherapy plan is converted to a specific radiotherapy plan, the set conversion generates one or more specific radiotherapy plans that are respectively matched to one or more specific radiotherapy devices.
  • the invention also provides a standardized cloud radiotherapy planning system, comprising: a main control cloud server, a network communication module, a client and a controlled computer, wherein:
  • the main control cloud server, the controlled computer, and the client are communicatively connected through a network communication module;
  • the main control cloud server is used to define a calculation phantom, target area delineation, and define calculation parameters, decompose calculation tasks, optimize allocation scheduling tasks, and monitor controlled computer execution;
  • the controlled computer is configured to receive a running instruction issued by the main control cloud server, determine a task execution, perform a computing task, and feed back a calculation progress and a calculation result;
  • the client is used to upload the patient image, patient data or clinical dose to the master cloud server, view the converted specific radiotherapy plan or calculate the progress.
  • the user sets the model of the specific radiotherapy device to be converted in a customized manner through the client.
  • the invention also provides a computer readable storage medium storing one or more programs, the one or more programs comprising instructions adapted to be loaded by a memory and to perform the standardized cloud radiation therapy planning method described above.
  • the present invention provides a standardized cloud radiotherapy planning method that converts a standard radiotherapy plan into a specific radiotherapy plan through information such as patient data and patient images. Therefore, it is possible to avoid the treatment time of the patient and the idleness of the treatment resources due to the failure of a certain type of machine in the hospital and the availability of other machines.
  • TPS automatic delineation and automatic radiotherapy
  • the dose volume histogram (DVH) and/or isodose line in the standard radiotherapy plan is close to the real situation, so the DVH curve and/or isodose line in the standard radiotherapy plan is used to optimize the dose in subsequent specific radiotherapy plans. Entering values can greatly reduce the time to generate subsequent specific radiotherapy plans.
  • Figure 1 is a block diagram showing the structure of a standardized cloud radiotherapy planning system in a preferred embodiment of the present invention.
  • FIG. 2 is a flow chart of a method of standardized cloud radiotherapy planning in a preferred embodiment of the present invention.
  • FIG. 3 is a flow chart of a method for standardizing cloud radiation therapy planning in another preferred embodiment of the present invention.
  • FIG. 4 is a flow chart of a method for standardizing cloud radiation therapy planning in yet another preferred embodiment of the present invention.
  • FIG. 5 is a schematic diagram of a Monte Carlo-based grid parallel dose calculation principle according to an exemplary embodiment of the present invention.
  • a standardized cloud radiotherapy planning method and system is now described that converts a standard radiotherapy plan into a specific radiotherapy plan in a cloud radiotherapy planning system through information such as patient data and patient images.
  • FIG. 1 is a block diagram showing the structure of a cloud radiotherapy plan system in an embodiment of the present invention.
  • the cloud radiotherapy planning system comprises: a main control cloud server, a network communication module, a client and a controlled computer, wherein: the main control cloud server, the controlled computer and the client are communicatively connected through the network communication module; preferably, the main control cloud server A data communication channel is established between the controlled computer and the client through the DICOM (Digital Imaging and Communications in Medicine) protocol.
  • DICOM Digital Imaging and Communications in Medicine
  • the main control cloud server is used to define the calculation phantom, the target area delineation and define the calculation parameters, decompose the calculation task into subtasks, optimize the allocation scheduling task, and monitor the controlled computer execution; the controlled computer is used to receive the main control cloud server.
  • the master cloud server defines the calculation parameters, optimizes the allocation scheduling task, and monitors the controlled computer execution; wherein the optimized allocation scheduling task is determined by establishing an optimization model including the optimization target and the constraint condition; optionally, the optimization target Include one or more combinations of minimum completion time, maximum number of completed tasks, minimum cost, or priority according to task, urgency weight, etc. Constraints include determining the current number of tasks, the network distribution that can be used, and the available Controlling the distribution of task completion rates of computer distributed or controlled computers; wherein optimizing the allocation scheduling task includes the following steps:
  • optimization model parameter initialization including defining optimization objectives and determining constraints related parameter initialization
  • Iterative solution using optimization algorithm iterative solution using optimization algorithm; optional optimization algorithms include: gradient algorithm, conjugate gradient method, Newton method, quasi-Newton method, multi-scale algorithm, interior point method, simple method, inheritance Algorithm, ant colony algorithm, particle swarm algorithm
  • the optimization results include the priority of the assigned task, the network resources used, and the computer resources.
  • the main control cloud server defines the calculation phantom, the target area delineation, and defines the calculation parameters; the main control cloud server decomposes, allocates, and schedules the calculation tasks according to the current user's calculation requirements, wherein the allocation and scheduling calculation tasks include sending calculation tasks to the controlled computer. Turn off one or more of computing tasks, transfer computing tasks, power on/off management, task prioritization management, or task security management.
  • the master cloud server monitors the controlled computer, and when it finds that any one of the controlled computers loses contact, the task of the controlled computer is reassigned to another controlled computer.
  • the monitoring method includes: actively sending or passively receiving a heartbeat packet, actively requesting or passively receiving a calculation progress, actively requesting, or passively receiving information related to the calculation result.
  • the cloud radiotherapy planning system of the embodiment includes a plurality of controlled computers. After the controlled computer accepts the task assigned by the main control cloud server, it first determines whether the task can be executed. If the current controlled computer judges, the distribution cannot be completed. The task returns the current task status to the master cloud server and requests the master cloud server to dispatch the assigned task to other controlled computers; if the judgment can be executed, the controlled computer continues to perform the computing task (including executing the subtask Etc., feedback steps to calculate progress and calculate results.
  • the computational computer-implemented radiotherapy plan calculation task includes dose calculation and/or dose optimization; the controlled computer performs the calculation task including decomposing into sub-tasks, executing sub-tasks; optionally, the controlled computer decomposes the tasks into the following sub-tasks One or more: GPU parallel tasks, CPU parallel tasks, or CPU-GPU hybrid parallel tasks.
  • the cloud radiotherapy planning system provided by the invention can be accessed by multiple users at the same time.
  • a user such as a doctor, a physicist or a technician accesses the main control cloud server through a client to define a computing phantom and a target area.
  • the user can also select a model or the like of the specific radiotherapy device to be converted through the client.
  • 2 is a flow chart of a method of normalizing cloud radiation therapy planning in accordance with an exemplary embodiment of the present invention.
  • a standardized cloud radiotherapy planning method suitable for execution in a standardized cloud radiotherapy planning system includes the following steps:
  • Step 210 Upload patient data to the main control cloud server, where the patient data includes patient image and medical order data; the patient image includes one or a combination of a CT image, a nuclear magnetic image, or a PET image; the medical data includes a target radiation dose, a DVH curve, One or a combination of radiotherapy dose constraint values for each organ;
  • Step 220 Perform target area delineation according to the patient image
  • the method for delineating the target area is automatic delineation, semi-automatic delineation or manual delineation; wherein the target area automatically delineates the content belonging to the prior art.
  • the publication number CN103247046B "A Method and Apparatus for Automatic Delineation of a Target Area in a Radiation Therapy Plan” in which a physician manually delineates a fault target area as a priori knowledge, and adopts a cyclic two-dimensional tomographic registration to realize automatic contouring. Dissemination; or refer to the “Automatic Delineation Evaluation of Target Areas for Nasopharyngeal Cancer” (Sichuan Medical, Vol. 36, No.
  • U-net Convolutional networks for biomedical image segmentation. In International Conference on Medical Image Computing and Computer-Assisted Intervention (pp. 234-241). Springer, Cham. U-type network: convolutional network for biomedical image segmentation, international conference on medical image computing and computer-assisted intervention.), the entire contents of which are incorporated herein by reference.
  • Step 230 The master cloud server allocates the computing task to the controlled computer, and the controlled computer calculates the patient's radiotherapy plan by using the standard radiotherapy device mode to generate a standard radiotherapy plan;
  • the standard radiotherapy device is a custom radiotherapy device or an established model is selected as a standard radiotherapy device, and further preferably, the most-used radiotherapy device of the type is used as a standard radiotherapy device; thereby reducing the standard radiotherapy plan into a specific radiotherapy The probability of planning further reduces the amount of calculation.
  • the radiotherapy plan calculated based on the standard radiotherapy apparatus in the present invention is a standard radiotherapy plan.
  • the device parameters that need to be defined in the above custom radiotherapy equipment include: source parameters, multi-leaf collimator parameters, tungsten gate parameters, and the like.
  • the source parameter includes the position of the radiation source, the energy spectrum, the direction of motion, the type of the particle, whether or not a homogenizer is used, and the homogenizer is used to reduce the intensity of the middle of the ray, thereby generating a uniform effect on the ray, if not using a homogenizer
  • 3F mode the intensity around the middle of the beam is large and the intensity is Gaussian; if the homogenizer is used, it is 2F mode, the intensity around the center of the beam is even and even;
  • the parameters of the multi-leaf collimator include the size and logarithm of the blade, and the maximum opening Size, whether interlacing is allowed, etc.
  • tungsten gate parameters include the maximum opening size of the tungsten gate.
  • the calculation task in this step is to generate a standard radiotherapy plan.
  • the generation process of the standard radiotherapy plan includes dose calculation and/or dose optimization.
  • the dose calculation parameters of the standard radiotherapy plan include geometric phantoms (determined by image of the target area, wherein the image may be selected from one or a combination of CT images, MRI images, PET images, etc.), medical order data, field size, Determination of the direction of illumination, the source parameter, the total number of tracking particles, the electron cutoff energy, the photon cutoff energy, the bremsstrahlung segmentation, the range exclusion, and the electron segmentation, wherein the source parameters include the position, energy spectrum, and motion of the source.
  • Direction, particle type whether or not to use a leveler.
  • the existing radiation dose calculation model includes: Monte Carlo calculation model, Acuros XB dose calculation model (used by Varian system), convolution superimposed dose calculation model, optional, convolution superimposed dose
  • the computational model further includes: a Collapse Cone Convolution Algorithm (CCC, using such a computational model including, for example, Pinnacle, CMS, XiO, etc.), an Analytical Anisotropic Algorithm (AAA), Pencil Beam Model (PBM).
  • CCC Collapse Cone Convolution Algorithm
  • AAA Analytical Anisotropic Algorithm
  • PBM Pencil Beam Model
  • a standard radiotherapy plan is generated as a computing task, and an optional way of decomposing a computing task into subtasks includes:
  • Method 1 Split a computing task into several subtasks through a flux map. Specifically, the arbitrary cross-section of the incident direction of the beam is divided into two-dimensional fluxes, and the region of interest in the patient image is divided into a three-dimensional voxel grid, and the i-th two-dimensional flux grid contributes to the j-th voxel.
  • the dose is D ij ; the dose calculation task for each D ij is a subtask.
  • each subtask is assigned as a GPU parallel task, a CPU parallel task, or a CPU-GPU hybrid parallel task.
  • Or mode 2 In the intensity-modulated radiotherapy (IMRT) mode or 3D conformal radiotherapy (3D-CRT) mode, set a beam dose calculation or dose optimization as a sub-task.
  • the mode in which the master cloud server allocates the computing task to the controlled computer includes a single plan mode and a multi-plan mode.
  • the single plan mode is the execution mode of the single radiotherapy plan of the cloud radiotherapy planning system;
  • the multi-planning mode has several (more than or equal to 2) radiotherapy plans to be executed, and multiple plans may be from different patients or the same Multiple radiotherapy plans for patients.
  • the single-planning mode of the radiotherapy plan calculation task can be achieved by the following steps:
  • the master cloud server calculates the currently available computing resources
  • the master cloud server decomposes the computing task into subtasks; calculates the computing resources required for the subtask to complete;
  • the master cloud server assigns the subtask to the controlled computer.
  • the calculation mode of the subtask is determined according to the target preset by the user; optionally, the calculation mode is a GPU parallel task, a CPU parallel task, or a CPU-GPU hybrid parallel task; the user preset target may include a subtask calculation time and/or Or the service cost of the computing resources required for the subtask.
  • the allocation of the radiotherapy plan calculation tasks in the multi-planning mode can be determined by the following steps:
  • the evaluation indicators of the priority level include: the severity of the patient's illness, the time sequence of joining the queue, and other indicators;
  • the main control cloud server decomposes the subtask according to the execution order of the radiotherapy plan calculation task and assigns the subtask to the controlled computer; after allocating the previous calculation task, continues to decompose the next radiotherapy plan calculation task into several subtasks.
  • the subsequent radiotherapy plan calculation subtask is assigned to the less-executed computing task or the idle controlled computer; optionally, the calculation mode of each subtask is GPU parallel task, CPU parallel tasks or CPU-GPU hybrid parallel tasks; the completion time of the computing tasks can be minimized or minimized according to the user's preset goals.
  • the standard radiotherapy plan in this embodiment can be generated in the following two ways:
  • FIG. 5 is a schematic diagram of dose calculation based on Monte Carlo-based grid parallel dose calculation principle in an exemplary embodiment of the present invention.
  • the 3D image of the patient or phantom is 3D meshed by the patient image, wherein each 3D mesh is a voxel, and the region of interest in the 3D mesh is selected; preferably, the Monte Carlo calculation is determined according to the region of interest.
  • the area, that is, the grid in a valid electron range around the region of interest and the grid in which the region of interest is located is set as the calculation region or directly as the calculation region; any cross section of the incident direction is divided into two dimensions.
  • volume meshing D ij is the dose contributed by the i-th flux grid to the j-th voxel; each two-dimensional flux grid corresponds to a weight ⁇ i ; input Monte Carlo dose calculation parameters and/or body Modulus parameters; calculating the radiation dose of each voxel based on Monte Carlo particle transport principle, and normalizing the calculation results; then superimposing the calculated results of each grid dose normalized in the calculation area to obtain the total radiation dose.
  • the dose of a single voxel grid D j is:
  • n is the total number of flux grids
  • j is a three-dimensional voxel label
  • m is the total number of voxels
  • ⁇ i is the weight of each flux grid in the Monte Carlo algorithm.
  • D ij is the dose contributed by the ith flux grid to the jth voxel
  • D j is the total dose of the jth voxel deposition.
  • the weight of each grid in the above two-dimensional flux grid is further optimized by the flux map optimization method by optimizing the target; the final reverse dose optimization result is obtained.
  • the dose distribution in each voxel in the region of interest is calculated; thereby determining the isodose line and the Dose-Volume Histogram (DVH), and the target area lesions and other key organs are received by the dose volume histogram.
  • the relationship between dose and volume is graphically represented, indicating how much volume of tissue is exposed to at least how much dose.
  • the dose volume histogram can be used to calculate the dose volume histogram of the lesion or tissue by counting the doses received by each voxel and then accumulating the same dose of voxels to determine the volume value at the corresponding dose.
  • a sequence of sub-fields containing the shape of the MLC and the tungsten gate is determined to obtain a standard radiotherapy plan.
  • Method 2 Automatically generate a standard radiotherapy plan according to the set model through machine learning; in which the standard radiotherapy plan can be generated by machine learning, refer to Dan N, Long T, Jia X, et al.
  • Dose Prediction with U-net A Feasibility Study for Predicting Dose Distributions from Contours using Deep Learning on Prostate IMRT Patients[J].2017(Chinese translation: U-type network for dose prediction ------Deep learning to predict the feasibility of profiling IMRT patients with contour dose distribution Research), the entire contents of which are incorporated herein by reference.
  • a dose volume histogram and an isodose line can be obtained for subsequent optimization of the constraints for generating a specific radiotherapy planning step.
  • Step 240 the master cloud server or the controlled computer converts and generates a radiotherapy plan that matches the specific radiotherapy device according to the standard radiotherapy plan.
  • the “matching” in the present invention is that the generated specific radiotherapy plan can be executed in the corresponding specific radiotherapy device.
  • This step specifically includes:
  • the standard radiotherapy plan was introduced, and the dose volume histogram, isodose line, and hardware parameters of the specific model in the standard radiotherapy plan were used as constraints to recalculate the final specific radiotherapy plan based on the field parameters of the standard radiotherapy plan.
  • the set conversion when converting a standard radiotherapy plan to a particular radiotherapy plan, the set conversion generates one or more specific radiotherapy plans that match one or more specific radiotherapy devices, respectively. That is, when converting a standard radiotherapy plan into a specific radiotherapy plan, you can set up a standard radiotherapy plan to simultaneously convert multiple specific radiotherapy plans (also known as redundant conversions) that match multiple specific radiotherapy devices or convert standard radiotherapy plans into A specific radiotherapy plan (also known as a specific conversion). For redundant conversion, the user can select and deploy the patient to perform treatment on any of the above specific radiotherapy devices according to the preset use of a plurality of specific radiotherapy devices. Redundancy conversion is suitable for situations where real-time requirements are not high and computer resources are abundant. When the real-time requirements are high, a specific conversion mode can be used to complete the conversion of the standard radiotherapy plan to a specific radiotherapy plan as quickly as possible.
  • the specific radiotherapy plan includes a dose volume histogram (DVH) matched with a specific radiotherapy apparatus, an isodose line, an execution sequence, an opening shape of each subfield, and an execution time of each subfield.
  • the above hardware parameter constraints include whether the tungsten gate, the maximum opening size of the tungsten gate, and the moving direction of the multi-leaf collimator, the thickness of the blade, the maximum opening position, the number of blade pairs, the leakage and the transmission, and the like.
  • the recalculation includes one or a combination of dose calculation and/or reverse optimization; wherein, optionally, the reverse optimization includes one or a combination of direct subfield optimization (DAO) or flux map optimization (FMO).
  • DAO direct subfield optimization
  • FMO flux map optimization
  • the DVH curve and/or isodose line of the standard radiotherapy plan is used as a constraint for the optimization of the specific radiotherapy plan, due to the DVH curve of the standard radiotherapy plan, the final DVH curve of the isodose line and the specific radiotherapy plan, and the final optimization result of the isodose line. Close or identical, therefore, speeds up dose optimization in a particular radiotherapy plan.
  • the standard radiotherapy plan in this embodiment can be converted into a dynamic multi-leaf collimator (DMLC) radiotherapy plan, a static multi-leaf collimator (SMLC) radiotherapy plan, and a volume rotation tune according to the model of the specific radiotherapy device to be converted.
  • DMLC dynamic multi-leaf collimator
  • SMLC static multi-leaf collimator
  • VMAT strong radiotherapy program
  • IMAT fixed dose rate rotational intensity
  • a dynamic multi-leaf collimator (DMLC) radiotherapy program that produces a specific radiotherapy device through a standard radiotherapy program, including the following steps:
  • the sequence optimization algorithm generates the motion sequence of the MLC.
  • the optimization method is to decompose the flux map, set the number of decompositions, minimum execution time, multi-leaf collimator blade thickness, distribution, maximum opening and other related constraints.
  • Algorithms such as the gradient method generate an execution sequence.
  • a static multi-leaf collimator (SMLC) radiotherapy program that produces a specific radiotherapy device through a standard radiotherapy program, including the following steps:
  • the subfield opening shape of the standard radiotherapy plan is used as the initial condition; according to the multi-leaf collimator thickness, distribution, and maximum opening of the specific radiotherapy device, an optimization algorithm is adopted.
  • the shape of the opening of the MLC is directly generated.
  • the above optimization algorithm is a direct subfield optimization method (DAO).
  • VMAT volumetric rotational intensity
  • Constraints with DVH and/or isodose lines of standard radiotherapy plans combined with mechanical constraints of specific radiotherapy equipment (including thickness of multi-leaf collimator, leaf distribution of multi-leaf collimator, and maximum opening position);
  • the irradiation of the radiotherapy equipment is divided according to equal angles; the optimization method is used to optimize the opening shape and residence time of the multi-leaf collimator field at each angle.
  • the above optimization algorithm is DAO.
  • Constraints with DVH and/or isodose lines of standard radiotherapy plans combined with mechanical constraints of specific radiotherapy equipment (including thickness of multi-leaf collimator, leaf distribution of multi-leaf collimator, and maximum opening position);
  • the optimization algorithm is used to optimize the opening shape and residence time of the multi-leaf collimator field at each angle.
  • the above optimization algorithm is DAO.
  • step 240 the step 240 of comparing the parameter matching between the standard radiotherapy device and the device to be matched is further included. If the parameter matching degree meets the preset threshold requirement, the standard radiotherapy plan is directly determined as the final specific Radiotherapy plan step 241', otherwise proceeds to step 240;
  • the equipment parameters to be compared include: source parameters, multi-leaf collimator parameters, D ij values under the same conditions, whether the grating models match, the maximum opening position of the tungsten gate, and the like.
  • the source parameters include source spectrum, position, direction, particle type, whether or not a homogenizer is used
  • the multi-leaf collimator parameters include blade size and logarithm, maximum open field size, whether interleaving is allowed, and the like.
  • the energy spectrum of the standard radiotherapy device and the specific radiotherapy device are compared by a similarity method, and when the similarity between the two is within a preset threshold range, the energy spectrum is considered to be similar.
  • the degree of coincidence of the source parameters is obtained by comparing the characteristic data (three-dimensional dose curve) of the dose measurement in the uniform or non-uniform medium, and specifically includes the following steps: (1) comparing the specific radiotherapy devices one by one The similarity between the dose measurement characteristic data and the standard planned dose measurement characteristic data;
  • Consistency is considered when the comprehensive similarity satisfies the preset threshold.
  • step (4) further includes a quality assurance (QA) step to enable the patient to verify that the converted plan was correct by QA prior to treatment.
  • QA quality assurance
  • FIG. 3 is a flow chart of a method for standardizing a cloud radiotherapy plan according to another exemplary embodiment of the present invention, which is suitable for being executed in a standardized cloud radiotherapy planning system, and includes the following steps:
  • Step 310 Upload patient data to the main control cloud server, where the patient data includes patient image and medical order data; the patient image includes one or a combination of a CT image, a nuclear magnetic image, or a PET image; the medical data includes a target radiation dose, a DVH curve, One or a combination of radiotherapy dose constraint values for each organ;
  • Step 320 performing target area delineation according to the patient image; delineating is automatic delineation or manual delineation;
  • Step 330 The master cloud server allocates the computing task to the controlled computer, and the controlled computer calculates the patient's radiotherapy plan by using the standard radiotherapy device mode to generate a standard radiotherapy plan;
  • the method includes the following steps: 340: comparing whether the parameter matching degree between the standard radiotherapy device and the device to be matched is within a threshold range; the parameter includes a source parameter, a multi-leaf collimator parameter; wherein the source parameter is compared to the source in a uniform or non-uniform medium
  • the characteristic data of the dose measurement is obtained (three-dimensional dose curve); the multi-leaf collimator parameters include the blade size and logarithm, the maximum open field size, whether interlacing is allowed; step 340, if the parameter coincidence degree meets the preset threshold requirement,
  • the standard radiotherapy plan is directly used as the final radiotherapy plan, otherwise it proceeds to step 350;
  • Step 350 The master cloud server or the controlled computer converts and generates a radiotherapy plan that matches the specific radiotherapy device according to the standard radiotherapy plan.
  • the step further includes:
  • recalculation includes dose calculation or reverse optimization One or a combination; wherein the reverse optimization includes one or a combination of direct subfield optimization or flux map optimization methods;
  • Step 360 Quality Assurance (QA) process, ie, whether the particular radiotherapy plan converted by the patient passes the quality assurance verification prior to performing the specific radiotherapy plan.
  • QA Quality Assurance
  • the object of the specific radiotherapy plan is replaced with a phantom made of solid water or other body tissue replacement material, and the radiation dose and dose distribution received in the phantom are tested to meet the doctor's prescription. Claim.
  • the particular radiotherapy plan is executed, and if not, return to step 350 to regenerate a new specific radiotherapy plan.
  • the standardized cloud radiotherapy planning method shown in FIG. 3 is the same as the method shown in FIG. 2 except for the above description.
  • a flow chart of a standardized cloud radiotherapy planning method in the process of formulating a radiotherapy plan, further includes the step of selecting a model of the radiotherapy device to be converted according to the use of the radiotherapy device, preferably using idle or
  • the standardized cloud radiation therapy planning method of the embodiment specifically includes the following steps: performing a less-critical device or according to a user-defined selection:
  • Step 410 Upload patient data to the main control cloud server, where the patient data includes patient image and medical order data; the patient image includes one or a combination of a CT image, a nuclear magnetic image, or a PET image; the medical data includes a target radiation dose, a DVH curve, One or a combination of radiotherapy dose constraint values for each organ;
  • Step 420 Perform target area delineation according to the patient image; sketching is automatic delineation or manual delineation;
  • Step 430 the master cloud server allocates the computing task to the controlled computer, and the controlled computer calculates the patient's radiotherapy plan by using the standard radiotherapy device mode to generate a standard radiotherapy plan;
  • Step 440 Select a model of the radiotherapy device to be converted according to the congestion of the radiotherapy device, and preferentially convert the standard radiotherapy plan into a currently available radiotherapy device or a radiotherapy device with a small number of tasks to be performed or a user-selected radiotherapy device model to be converted. ;
  • Step 450 The master cloud server or the controlled computer converts and generates a radiotherapy plan matching the specific radiotherapy device according to the standard radiotherapy plan.
  • the step further includes: introducing a standard radiotherapy plan, and applying a dose volume histogram in the standard radiotherapy plan, an equal dose
  • the line is used as a constraint to recalculate the final radiotherapy plan based on the field parameters of the standard radiotherapy plan; the recalculation includes one or a combination of dose calculation or reverse optimization; wherein the reverse optimization includes direct subfield optimization or flux One or a combination of graph optimization methods;
  • the method further includes a step 460 quality assurance QA step (not shown in FIG. 4), and the patient is verified by the QA verification whether the converted plan is correct before the treatment, and optionally, the specific radiotherapy plan is performed when the quality assurance step is performed.
  • the execution object is replaced with a phantom made of solid water or other body composition replacement material, and the radiation dose and dose distribution received in the phantom are tested to meet the medical requirements.
  • the particular radiotherapy plan is executed, and if not, return to step 350 to regenerate a new specific radiotherapy plan.
  • the standardized cloud radiotherapy planning method shown in FIG. 4 is the same as the method shown in FIG. 2 or FIG. 3 except for the above description.
  • the present invention also provides a computer readable storage medium storing one or more programs, the one or more programs comprising instructions adapted to be loaded by the memory and to perform the standardized cloud radiation therapy planning method described above, the method comprising the steps of:
  • the main control cloud server decomposes the computing task and assigns it to the controlled computer.
  • the controlled computer calculates the patient's radiotherapy plan using the standard radiotherapy device mode to generate a standard radiotherapy plan;
  • the master cloud server or managed computer converts the standard radiotherapy plan into a specific radiotherapy plan that matches the particular radiotherapy device based on the allocation of the master cloud server.
  • the computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture comprising the instruction device.
  • the apparatus implements the functions specified in one or more blocks of a flow or a flow and/or block diagram of the flowchart.
  • These computer program instructions can also be loaded onto a computer or other programmable data processing device such that a series of operational steps are performed on a computer or other programmable device to produce computer-implemented processing for execution on a computer or other programmable device.
  • the instructions provide steps for implementing the functions specified in one or more of the flow or in a block or blocks of a flow diagram.
  • Computer readable media includes both permanent and non-persistent, removable and non-removable media.
  • Information storage can be implemented by any method or technology.
  • the information can be computer readable instructions, data structures, modules of programs, or other data.
  • Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory. (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technology, compact disk read only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape storage or other magnetic storage devices or any other non-transportable media can be used to store information that can be accessed by a computing device.
  • computer readable media does not include temporary storage computer readable media, such as modulated data signals and carrier waves.
  • the standardized cloud radiotherapy planning method can prevent the treatment time of the patient from being delayed and the idleness of the treatment resources due to the failure of a certain type of machine in the hospital, and the radiotherapy plan is to be treated by the main control cloud service.
  • the computational tasks are assigned to the controlled computer, making it difficult to apply the clinical "gold standard" for the calculation of radiation doses - Monte Carlo particle transport simulation dose calculations are clinically possible, greatly saving the development time of the radiation therapy plan and the patient's Waiting time to improve the accuracy of the dose calculation for the radiotherapy plan.
  • through automatic delineation and automatic TPS development it is also possible to balance the differences in the treatment levels of doctors in different hospitals or regions, and also reduce the workload of oncologists and physicists.

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Business, Economics & Management (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Epidemiology (AREA)
  • Medical Informatics (AREA)
  • Primary Health Care (AREA)
  • Biomedical Technology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Veterinary Medicine (AREA)
  • Animal Behavior & Ethology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Pathology (AREA)
  • Bioethics (AREA)
  • Surgery (AREA)
  • Urology & Nephrology (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Marketing (AREA)
  • Development Economics (AREA)
  • Game Theory and Decision Science (AREA)
  • Radiation-Therapy Devices (AREA)

Abstract

本方案属于放疗和云计算技术领域，涉及一种标准化云放疗计划方法、存储介质和系统。该方法包括如下步骤：(1)向主控云服务器上传患者数据（S210），其中所述的患者数据包括患者影像、医嘱数据；(2)根据患者影像进行靶区勾画（S220）；(3)主控云服务器将计算任务分配给受控计算机，受控计算机用标准放疗设备模式计算患者的放疗计划，生成标准放疗计划（S230）；(4)根据标准放疗计划生成与特定放疗设备匹配的特定放疗计划。通过本方案提供的方法可以避免因为医院内某类型放疗设备发生故障而其他放疗设备空闲的情况下耽误患者的治疗时间以及治疗资源的闲置；还能平衡不同医院或地区的医生的治疗水平的差异，也能降低肿瘤医生与物理师的工作负担。

Description

一种标准化云放疗计划方法、存储介质和系统 技术领域
本发明属于放疗和云计算技术领域，涉及一种标准化云放疗计划方法、存储介质和系统。
背景技术
2014年2月3日，世界卫生组织(WHO)发布的《2014年世界癌症报告》指出，目前每年有大约新发癌症病人1400万人，预期未来20年内上升到2200万；相同时期内，癌症死亡人数将从每年820万上升到1300万。WHO在2017年的实况报道中指出，从全球情况看，近六分之一的死亡由癌症造成。其中，当前60～70％肿瘤患者中在其整个患病治疗过程的某一个阶段都需要接受不同类型的放射治疗(简称放疗)。
根据统计，目前中国癌症发病和死亡人数已居世界第一位。2015年，国内癌症新发病例430万，死亡病例280万。放疗与外科手术治疗、化学药物治疗并列为肿瘤治疗的三大方法。而中国目前的物理师人才缺口为1万名，医疗资源极度紧缺。
一方面，三甲医院的品牌效应吸引了大批患者蜂拥进入三甲医院治疗，尤其大型三甲医院更是人满为患，病床使用率长期处于饱和状态。而据统计基层医疗机构的病床使用率仅有60％左右，部分医疗资源被浪费，没有发挥出应有的效益。大量患者集中在三甲医院进行诊疗，势必造成医护人员负担过重以及放疗设备和治疗床位紧张，无法满足所有患者的需求；优质医疗资源过度集中在三甲医院，也导致了医疗服务覆盖范围下降，边远地区的人群很难就近享受到富有经验的专家的诊疗。如果患者由上级医院有经验的医生或物理师确定治疗方案后，再回到基层医院进行放射治疗，则不但诊断和治疗的质量得到保障，而且还可以不必排队或长时间等候床位和放疗设备。
另一方面，部分医院中的放疗设备的类型不尽相同，可能包括若干品牌多个型号的放疗设备，如果事先根据一台放疗设备制定完成放疗计划后，而该设备意外故障又没有相同型号的替换设备时，势必造成先前已经制定好的放疗计划的搁置，延误病人的治疗时间；或者当前已制定的放疗计划的放疗设备已被占用需要长时间排队，而其他放疗设备正在闲置或排队等候时间较短。因此，如何避免因放疗设备故障造成的放疗计划的搁置或延误执行，提高三甲医院中各放疗设备的有效利用率，减少患者的等待时间，将三甲医院的患者分流到基层医疗机构并维持放射治疗水平是现有技术急需解决的问题。
发明内容
本发明的目的在于为克服上述现有技术的缺陷而提供一种标准化云放疗计划方法、存储介质和系统。
为实现上述目的，本发明采用以下技术方案：
一种标准化云放疗计划方法，适于在标准化云放疗计划系统中执行，包括如下步骤：
(1)向主控云服务器上传患者数据，其中所述的患者数据包括患者影像、医嘱数据；
(2)根据患者影像进行靶区勾画；
(3)主控云服务器将计算任务分解后分配给受控计算机，受控计算机用标准放疗设备模式计算患者的放疗计划，生成标准放疗计划；
(4)主控云服务器或受控计算机根据标准放疗计划，转换生成与特定放疗设备匹配的特定放疗计划，所述的匹配是生成的特定放疗计划能够在相应的特定放疗设备中执行。
本发明进一步优选地，所述的患者影像包括CT影像、核磁影像或PET影像中的一种或者组合。
所述的医嘱数据包括目标放疗剂量、DVH曲线、各器官放疗剂量约束值中的一种或组合。
所述的勾画为自动勾画、半自动勾画或者手动勾画。
所述的主控云服务器或受控计算机根据标准放疗计划，转换生成与特定放疗设备匹配的放疗计划步骤进一步包括：
导入标准放疗计划，并将标准放疗计划中的剂量体积直方图、等剂量线作为特定放疗计划的优化约束条件，在标准放疗计划的射野参数基础上，重新计算最终特定放疗计划。步骤(4)中，所述的重新计算包括剂量计算或者逆向优化中的一种或者组合；其中，所述的逆向优化包括采用直接子野优化(DAO)或者通量图优化方法(FMO)中的一种或者组合。
步骤(4)后还包括质量保证(QA)步骤(5)，病人在治疗前通过QA验证所转换的特定放疗计划是否正确；如果正确，执行该特定放疗计划；如果不正确，返回步骤(4)根据标准放疗计划重新生成新的特定放疗计划。
步骤(3)与步骤(4)之间，还包括根据放疗设备的使用拥堵情况，选择待转换放疗设备的步骤：优先将标准放疗计划转换为当前空闲的放疗设备或待执行任务数量较少的放疗设备或用户自定义选择待转换的放疗设备型号。
生成标准放疗计划时或转换生成与特定放疗设备匹配的放疗计划时根据可用计算资源实行进程排队转换或用户自定义设定计算任务的优先次序。
所述的根据标准放疗计划，转换生成与特定放疗设备匹配的特定放疗计划之前还包括如下步骤：比较标准放疗设备与待匹配特定放疗设备的参数吻合度，如果参数吻合度符合预设的阈值要求，将标准放疗计划直接作为最终放疗计划，否则进入步骤(4)。
标准放疗设备与待匹配特定放疗设备需要比较的参数包括源参数、多叶准直器参数。
所述的源参数通过比较源在均匀或者非均匀介质中的剂量测量特性数据获得；所述的剂量测量特性数据通过三维剂量曲线获得。
所述的多叶准直器参数包括叶片大小和对数、最大开野大小、是否允许交错。
步骤(4)中，将标准放疗计划转换为特定放疗计划时，设置转换生成与一 台或多台特定放疗设备分别匹配的一个或多个特定放疗计划。
本发明还提供一种标准化云放疗计划系统，包括：主控云服务器，网络通信模块，客户端以及受控计算机，其中：
所述主控云服务器、受控计算机、客户端通过网络通讯模块通信连接；
所述主控云服务器用来定义计算模体、靶区勾画以及定义计算参数，分解计算任务，优化分配调度任务，并监控受控计算机执行；
所述受控计算机用来接收主控云服务器发出的运行指令、判断任务执行、执行计算任务、反馈计算进度与计算结果；
所述的客户端用来将病人影像、病人数据或临床剂量上传到主控云服务器、查看转换得到的特定放疗计划或计算进度。
优选地，用户通过所述的客户端通过自定义方式设定待转换的特定放疗设备的型号。
本发明还提供一种存储一个或多个程序的计算机可读存储介质，所述的一个或多个程序包括指令，所述指令适于由存储器加载并执行上述标准化云放疗计划方法。
本发明具有以下有益效果：
本发明提供了一种标准化云放疗计划方法，该方法通过患者数据和患者影像等信息将标准放疗计划转换为特定放疗计划。从而避免因为医院内某类型机器发生故障而其他机器可用的情况下耽误患者的治疗时间以及治疗资源的闲置。通过自动勾画及自动放疗计划(TPS)制定，还能平衡不同医院或地区的医生的治疗水平的差异，也能减轻肿瘤医生与物理师的工作负担。通过主控云服务将放疗计划的计算任务分配给受控计算机，使难以应用于临床的放疗剂量计算的“金标准”—基于蒙特卡罗粒子输运模拟的剂量计算应用于临床成为可能。另外，标准放疗计划中的剂量体积直方图(DVH)和/或等剂量线已经接近真实情况，因此，以标准放疗计划中的DVH曲线和/或等剂量线为后续特定放疗计划中剂量优化的输入值，能大大缩短后续特定放疗计划的生成时间。
附图说明
图1为本发明一个优选的实施例中标准化云放疗计划系统的结构示意图。
图2为本发明一个优选的实施例中标准化云放疗计划方法的流程图。
图3为本发明另一个优选的实施例中标准化云放疗计划方法的流程图。
图4为本发明又一个优选的实施例中标准化云放疗计划方法的流程图。
图5为本发明一个示例性实施例中基于蒙特卡罗的网格并行剂量计算原理示意图。
具体实施方式
现在描述一种标准化云放疗计划方法和系统，该方法通过患者数据和患者影像等信息在云放疗计划系统中将标准放疗计划转换为特定放疗计划。
以下结合附图和实施例进一步说明本发明。显然，所描述的实施例仅仅是本发明的一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动的前提下所获得的所有其他实施例，都应属于本发明保护的范围。
图1为本发明一个示出实施例中的云放疗计划系统结构示意图。该云放疗计划系统包括：主控云服务器，网络通信模块，客户端以及受控计算机，其中：主控云服务器、受控计算机及客户端通过网络通信模块通信连接；优选地，主控云服务器、受控计算机及客户端之间通过DICOM(Digital Imaging and Communications in Medicine)协议建立数据通信通道。主控云服务器用来定义计算模体、靶区勾画以及定义计算参数，将计算任务分解为子任务、优化分配调度任务，并监控受控计算机执行；受控计算机用来接收主控云服务器发出的运行指令、判断任务执行、执行计算任务、反馈计算进度与计算结果，其中计算结果包括特定放疗计划、计算进度等；客户端用来将病人影像、病人数据或临床剂量上传到主控云服务器，并查看放疗计划结果。
主控云服务器定义计算参数，优化分配调度任务，并监控受控计算机执行；其中，优化分配调度任务是通过建立优化模型确定的，上述优化模型包括优化目 标与约束条件；可选的，优化目标包括最小完成时间、最多完成任务数、最低费用、或者根据任务的优先等级、紧急程度权重等中的一个或者多个组合，约束条件包括确定当前任务数量、可使用的网络分布、可使用的受控计算机分布或受控计算机的任务完成率分布；其中，优化分配调度任务包括如下步骤：
a)优化模型参数初始化：包括定义优化目标与确定约束条件相关参数初始化；
b)利用优化算法迭代求解：使用优化算法迭代求解；其中可选的优化算法包括：梯度算法、共轭梯度法、牛顿法、拟牛顿法、多尺度算法、内点法、单纯型法、遗传算法、蚁群算法、粒子群算法；
c)结果输出：优化结果包括分配任务的优先级、所使用的网络资源以及计算机资源。
通过主控云服务器定义计算模体、靶区勾画以及定义计算参数；主控云服务器根据目前用户的计算要求分解、分配与调度计算任务，其中分配与调度计算任务包括对受控计算机发送计算任务、关闭计算任务、转移计算任务、开关机管理、任务优先次序管理或任务安全管理中的一种或多种。
主控云服务器对受控计算机进行监控，当发现任意一台受控计算机失去联系，则将该受控计算机的任务重新分配给另外一台受控计算机。监控方法包括：主动发送或者被动接收心跳包、主动请求或者被动接收计算进度、主动请求或者被动接收计算结果相关信息。
本实施例的云放疗计划系统中包含若干台受控计算机，受控计算机接受主控云服务器分配的任务后，先判断任务能否执行，如果当前的受控计算机经判断后发现不能够完成分配的任务，则将当前任务状态反馈给主控云服务器并请求主控云服务器将分配的任务调度给其他受控计算机；如果判断能够执行则该受控计算机继续进行执行计算任务(包括执行子任务等)、反馈计算进度与计算结果的步骤。受控计算机执行的制定放疗计划计算任务包括剂量计算和/或剂量优化；受控计算机执行计算任务包括分解成子任务、执行子任务；可选的，受控计算机将任务分解为以下子任务中的一种或多种：GPU并行任务、CPU并行任务或CPU-GPU混合并行任务。
本发明提供的云放疗计划系统可同时供多个用户访问，如图1所示，用户(如 医生、物理师或技师等)通过客户端访问主控云服务器，来定义计算模体、靶区勾画以及定义计算参数，并通过客户端查看转换进度以及转换得到的特定放疗计划结果。本实施例中进一步优选地，用户还可以通过客户端选择设定待转换的特定放疗设备的型号等。图2为本发明一个示例性实施例中标准化云放疗计划方法的流程图。一种标准化云放疗计划方法，适于在标准化云放疗计划系统中执行，包括如下步骤：
步骤210，向主控云服务器上传患者数据，其中患者数据包括患者影像、医嘱数据；患者影像包括CT影像、核磁影像或PET影像中的一种或者组合；医嘱数据包括目标放疗剂量、DVH曲线、各器官放疗剂量约束值中的一种或组合；
步骤220，根据患者影像进行靶区勾画；
可选的，靶区勾画的方法为自动勾画、半自动勾画或者手动勾画；其中，靶区自动勾画属于现有技术的内容。例如，可以参考公开号CN103247046B《一种放射治疗计划中靶区自动勾画的方法和装置》，以医师手动勾画某一断层靶区作为先验知识，采用循环二维断层配准，实现轮廓的自动传播；或参考《鼻咽癌靶区的自动勾画评价》(《四川医学》2015年6月第36卷(第6期)，p762-p765)；或参考Ronneberger,O.,Fischer,P.,& Brox,T.(2015,October).U-net:Convolutional networks for biomedical image segmentation.In International Conference on Medical Image Computing and Computer-Assisted Intervention(pp.234-241).Springer,Cham.(中文翻译：U-型网络：用于生物医学图像分割的卷积网络，医学图像计算与计算机辅助干预国际会议。)，上述全部内容通过引用合并于此。
步骤230，主控云服务器将计算任务分解后分配给受控计算机，受控计算机用标准放疗设备模式计算患者的放疗计划，生成标准放疗计划；
本实施例中标准放疗设备为自定义放疗设备或选用既定机型作为标准放疗设备，进一步优选地，将保有数量最多的该种型号放疗设备作为标准放疗设备；从而降低标准放疗计划转换为特定放疗计划的概率，进一步减少计算量。本发明中基于标准放疗设备计算得到的放疗计划为标准放疗计划。
其中上述自定义放疗设备需要定义的设备参数包括：源参数、多叶准直器参数、钨门的参数等。其中源参数包括放射源的位置、能谱、运动方向、粒子种类， 是否使用均整器等；所述的均整器用于消减射线中间的强度，从而对射线产生均整作用，如果不采用均整器称为3F模式，射束中间周围强度较大，强度为高斯分布；如果采用均整器则为2F模式，射束中心周围的强度平整均匀；多叶准直器参数包括叶片大小和对数、最大开野大小、是否允许交错等；钨门参数包括钨门的最大开野尺寸。
本步骤中的计算任务为生成标准放疗计划。其中，标准放疗计划的生成过程包括剂量计算和/或剂量优化。
可选的，标准放疗计划的剂量计算参数包括几何模体(通过靶区图像来确定，其中图像可以选自CT图像、MRI图像、PET图像等一种或组合)、医嘱数据、射野大小、照射方向、源参数、跟踪粒子总数、电子截止能量、光子截止能量、韧致辐射分割、射程排除、电子分割中的一种或组合等确定，其中源参数包括放射源的位置、能谱、运动方向、粒子种类，是否使用均整器等。
可选的，现有辐射剂量计算模型包括：蒙特卡罗(Monte Carlo)计算模型、Acuros XB剂量计算模型(瓦里安系统使用)、卷积叠加剂量计算模型，可选的，卷积叠加剂量计算模型进一步包括：卷积叠加剂量计算模型(Collapse Cone Convolution algorithms，CCC，使用这种计算模型的包括例如Pinnacle,CMS,XiO,等)，分析各向异性算法(the Analytical Anisotropic Algorithm，AAA)，笔形束计算模型(Pencil Beam Model，PBM)。
本实施例中优选地，生成一个标准放疗计划为一个计算任务，将一个计算任务分解为子任务的可选的方式包括：
方式一：通过通量图将一个计算任务拆分为若干子任务。具体是将射束入射方向的任意截面划分二维通量网格化，将患者影像中感兴趣区域划分三维体素网格，则第i个二维通量网格对第j个体素贡献的剂量为D ij；每个D ij的剂量计算任务为一个子任务。可选的，将各子任务分配成GPU并行任务、CPU并行任务或CPU-GPU混合并行任务。
或方式二：在调强放疗(IMRT)模式或3D适形放疗(3D-CRT)模式下，设定一个射束(beam)的剂量计算或剂量优化为一个子任务。本实施例中，主控云服务器将计算任务分配给受控计算机的模式包括单计划模式、多计划模式。其 中单计划模式为云放疗计划系统单个放疗计划的执行模式；多计划模式为同时有若干个(大于等于2个)待执行的放疗计划，多个计划可以来自不同的患者，也可以是同一个患者的多个放疗计划。
模式一：单计划模式
单计划模式的放疗计划计算任务的分配方式可以通过如下步骤实现：
(1)主控云服务器计算当前可用计算资源；
(2)主控云服务器将计算任务分解成子任务；计算子任务完成所需的计算资源；
(3)主控云服务器将子任务分配给受控计算机。根据用户预设的目标，确定子任务的计算模式；可选的，计算模式为GPU并行任务、CPU并行任务或CPU-GPU混合并行任务；用户预设的目标可以包括子任务的计算时间和/或子任务所需计算资源的服务费用等。
模式二：多计划模式
多计划模式的放疗计划计算任务的分配方式可以通过如下步骤确定：
(1)计算当前可用计算资源；
(2)判断待执行各放疗计划计算任务的优先等级；可选的，该优先等级的评价指标包括：患者病症危急程度、加入排队等候的时间次序以及其它指标；
(3)根据各放疗计划计算任务的优先等级和计算资源情况确定多个放疗计划计算任务的执行顺序；
(4)主控云服务器按照放疗计划计算任务的执行顺序分解子任务并将子任务分配给受控计算机；当分配完前一个计算任务后，继续将下一个放疗计划计算任务分解为若干子任务，并分配给受控计算机，优选地，将在后的放疗计划计算子任务分配给待执行计算任务较少的或空闲受控计算机；可选的，各子任务的计算模式为GPU并行任务、CPU并行任务或CPU-GPU混合并行任务；可按照用户的预设的目标，使计算任务的完成时间最短或花费最少等。
可选的，本实施例中的标准放疗计划可以通过以下两种方式生成：
方式一：通过通量图优化法(FMO)获得标准放疗计划：
图5为本发明一个示例性实施例中，基于蒙特卡罗的网格并行剂量计算原 理进行剂量计算的示意图。通过患者影像将患者或体模的三维影像进行3D网格化，其中每个3D网格为一个体素，选取3D网格中的感兴趣区域；优选地，根据感兴趣区域确定蒙特卡罗计算区域，即：将感兴趣区域周围一个有效电子射程内的网格及感兴趣区域所在的网格设置为计算区域或直接将感兴趣区域作为计算区域；将入射方向的任意一个截面划分二维通量网格化，D ij为第i个通量网格对第j个体素贡献的剂量；每个二维通量网格对应的权重为ω i；输入蒙特卡罗剂量计算参数和/或体模参数；基于蒙特卡罗粒子输运原理计算每个体素中粒子辐射剂量，并将计算结果归一化；再将计算区域内归一化的各网格剂量计算结果叠加，得到总辐射剂量。其中单个体素网格D j的剂量为：
Figure PCTCN2018099445-appb-000001
其中，
i为二维通量网格标号，
n为通量网格的总数目，
j为三维体素标号，
m为体素的总数目，
ω i为蒙特卡罗算法中各通量网格权重，
D ij为第i个通量网格对第j个体素贡献的剂量，
D j为第j个体素沉积的总剂量。
根据计算得到的D ij，通过优化目标，进一步通过通量图优化法优化上述二维通量网格中各网格的权重；得到最终逆向剂量优化结果。
通过计算得到感兴趣区域内各体素中的剂量分布；从而确定等剂量线和剂量体积直方图(Dose-Volume Histogram，DVH)，通过剂量体积直方图把靶区病灶和其他关键器官所受到的剂量与体积的关系图表化，表示多少体积的组织至少受到了多少剂量照射。剂量体积直方图可以通过统计各体素所受到的剂量，然后把相同剂量的体素累加求出对应剂量下的体积值，从而统计出该病灶或组织的剂量体积直方图。
根据得到的最终逆向优化的二维通量网格权重，确定包含MLC和钨门位置 的射野开口形状的各子射野的序列(leaf sequence)，得到标准放疗计划。
方式二：通过机器学习方式根据设定的机型自动生成标准放疗计划；其中通过机器学习的方式产生标准放疗计划可以参考Dan N,Long T,Jia X,et al.Dose Prediction with U-net:A Feasibility Study for Predicting Dose Distributions from Contours using Deep Learning on Prostate IMRT Patients[J].2017(中文翻译：利用U型网络进行剂量预测------深度学习预测前列腺IMRT患者轮廓剂量分布的可行性研究)，上述全部内容通过引用合并于此。通过机器学习获得的放疗计划中包含的剂量分布信息，可以得到剂量体积直方图和等剂量线，用于后续优化生成特定放疗计划步骤的约束条件。
步骤240，主控云服务器或受控计算机根据标准放疗计划，转换生成与特定放疗设备匹配的放疗计划，本发明中的“匹配”是生成的特定放疗计划能够在相应的特定放疗设备中执行，本步骤具体包括：
导入标准放疗计划，并将标准放疗计划中的剂量体积直方图，等剂量线、以及特定机型的硬件参数作为约束条件，在标准放疗计划的射野参数基础上，重新计算最终特定放疗计划。
可选的，将标准放疗计划转换为特定放疗计划时，设置转换生成与一台或多台特定放疗设备分别匹配的一个或多个特定放疗计划。即：将标准放疗计划转换为特定放疗计划时，可设置将标准放疗计划同时转换为与多台特定放疗设备分别匹配的多个特定放疗计划(又称冗余转换)或将标准放疗计划转换为一种特定放疗计划(又称特定转换)。对于冗余转换，用户可以根据预设的多台特定放疗设备的使用情况，选择、调配病人在上述任一特定放疗设备上进行治疗。冗余转换适用于对实时性要求不高，计算机资源较为充裕的情况。当实时性要求较高时，可使用特定转换模式，以便尽快完成标准放疗计划到特定放疗计划的转换。
其中，在上述特定放疗计划中包括与特定放疗设备匹配的剂量体积直方图(DVH)、等剂量线、执行序列、每个子野开口形状、各子野的执行时间。上述硬件参数约束条件包括是否含有钨门、钨门最大开口尺寸，以及多叶准直器的运动方向、叶片厚度、最大开口位置、叶片对数、漏射和透射情况等。
重新计算包括剂量计算和/或逆向优化中的一种或者组合；其中，可选地， 逆向优化包括采用直接子野优化(DAO)或者通量图优化法(FMO)等中的一种或者组合。将标准放疗计划的DVH曲线和/或等剂量线作为特定放疗计划优化的约束条件，由于标准放疗计划的DVH曲线、等剂量线与特定放疗计划的最终DVH曲线、等剂量线的最终优化结果非常接近或相同，因此，可以加快特定放疗计划中剂量优化的速度。
可选的，本实施例中标准放疗计划根据待转换的特定放疗设备的型号可以转换成动态多叶准直器(DMLC)放疗计划、静态多叶准直器(SMLC)放疗计划、容积旋转调强(VMAT)放疗计划或固定剂量率旋转调强(IMAT)放疗计划。
1、通过标准放疗计划产生特定放疗设备的动态多叶准直器(DMLC)放疗计划，包括如下步骤：
以标准放疗计划的DVH和/或等剂量线作为约束条件，结合特定放疗设备的机械约束(包括多叶准直器的厚度、多叶准直器的叶片分布情况以及最大开口位置)；通过MLC序列优化算法产生MLC的运动序列，其中优化方法是通过对通量图进行分解，设定分解次数、最小执行时间、多叶准直器叶片厚度、分布情况、最大开口等相关约束条件，利用优化算法比如梯度法生成执行序列。
2、通过标准放疗计划产生特定放疗设备的静态多叶准直器(SMLC)放疗计划，包括如下步骤：
将标准放疗计划的DVH和/或等剂量线作为约束条件，标准放疗计划的子野开口形状作为初始条件；根据特定放疗设备的多叶准直器厚度、分布情况、最大开口，采用优化算法，直接产生MLC的开口形状。可选的，上述优化算法为直接子野优化法(DAO)。
3、通过标准放疗计划产生特定放疗设备的容积旋转调强(VMAT)放疗计划，包括如下步骤：
以标准放疗计划的DVH和/或等剂量线作为约束条件，结合特定放疗设备的机械约束(包括多叶准直器的厚度、多叶准直器的叶片分布情况以及最大开口位置)；将特定放疗设备的照射按照等角度分割；利用优化算法优化出每个角度的多叶准直器子野的开口形状及停留时间。可选的，上述优化算法为DAO。
4、通过标准放疗计划产生特定放疗设备的固定剂量率旋转调强(IMAT)放疗计划，其中，剂量率为单位时间的辐射剂量，包括如下步骤：
以标准放疗计划的DVH和/或等剂量线作为约束条件，结合特定放疗设备的机械约束(包括多叶准直器的厚度、多叶准直器的叶片分布情况以及最大开口位置)；将照射按照等角度分割；利用优化算法优化出每个角度的多叶准直器子野的开口形状及停留时间。可选的，上述优化算法为DAO。
可选的，在步骤240前进一步包含比较标准放疗设备与待匹配设备的参数吻合度是否在阈值范围的步骤240’，如果参数吻合度符合预设的阈值要求，将标准放疗计划直接作为最终特定放疗计划步骤241’，否则进入步骤240；
其中，需要比较的设备参数包括：源参数、多叶准直器参数、相同条件下D ij数值、光栅型号是否匹配、钨门最大开口位置等。进一步，源参数包括源能谱、位置、方向、粒子类型，是否使用均整器等等；多叶准直器参数包括叶片大小和对数、最大开野大小、是否允许交错等。上述参数相同或差值在预设的吻合度阈值范围内时，判定标准放疗设备与待匹配的特定放疗设备的参数吻合度在阈值范围。可选的，标准放疗设备与特定放疗设备的能谱采用相似度法进行比较，当二者的相似度在预设的阈值范围内，认为能谱相似。
本实施例中可选的，源参数的吻合度通过比较源在均匀或者非均匀介质中的剂量测量的特性数据(三维剂量曲线)获得，具体包括如下步骤：(1)一一比较特定放疗设备的剂量测量特性数据与标准计划的剂量测量特性数据的相似度；
(2)计算综合的相似度，将所有的相似度通过预设的权重进行加权求和获得综合的相似度；
(3)当综合相似度满足预设阈值则认为具有一致性。
如图3所示的实施例中，在图2所示实施例的基础上，步骤(4)后还包括质量保证(QA)步骤，使病人在治疗前通过QA验证所转换的计划是否正确。
如图3所示为本发明另一个示例性实施例中标准化云放疗计划方法的流程图，适于在标准化云放疗计划系统中执行，包括如下步骤：
步骤310，向主控云服务器上传患者数据，其中患者数据包括患者影像、医 嘱数据；患者影像包括CT影像、核磁影像或PET影像中的一种或者组合；医嘱数据包括目标放疗剂量、DVH曲线、各器官放疗剂量约束值中的一种或组合；
步骤320，根据患者影像进行靶区勾画；勾画为自动勾画或者手动勾画；
步骤330，主控云服务器将计算任务分配给受控计算机，受控计算机用标准放疗设备模式计算患者的放疗计划，生成标准放疗计划；
可选的，包括步骤340，比较标准放疗设备与待匹配设备的参数吻合度是否在阈值范围；参数包括源参数、多叶准直器参数；其中，源参数通过比较源在均匀或者非均匀介质中的剂量测量的特性数据获得(三维剂量曲线)；多叶准直器参数包括叶片大小和对数、最大开野大小、是否允许交错；步骤340，如果参数吻合度符合预设的阈值要求，将标准放疗计划直接作为最终放疗计划，否则进入步骤350；
步骤350，主控云服务器或受控计算机根据标准放疗计划，转换生成与特定放疗设备匹配的放疗计划，优选地，本步骤进一步包括：
导入标准放疗计划，并将标准放疗计划中的剂量体积直方图，等剂量线作为约束条件，在标准放疗计划的射野参数基础上，重新计算最终放疗计划；重新计算包括剂量计算或者逆向优化中的一种或者组合；其中，逆向优化包括采用直接子野优化或者通量图优化方法中的一种或者组合；
步骤360质量保证(QA)过程，即：病人在执行特定放疗计划前通过质量保证验证所转换的特定放疗计划是否正确。可选的，在进行质量保证步骤时，将特定放疗计划的执行对象更换为以固体水或其它人体组织替换材料制成的体模，测试该体模中接受的辐射剂量及剂量分布是否满足医嘱要求。当满足医嘱要求时，执行该特定放疗计划，如果不满足，返回步骤350，重新生成新的特定放疗计划。
图3所示的标准化云放疗计划方法除了上述说明的内容外，其他与图2所示的方法相同。
如图4所示实施例中，一种标准化云放疗计划方法的流程图，在放疗计划的制定过程中，还包括按照放疗设备的使用情况选择所要转换的放疗设备型号的步骤，优先选用空闲或执行任务较少的设备或根据用户自定义选择，本实施例的标 准化云放疗计划方法具体包括如下步骤：
步骤410，向主控云服务器上传患者数据，其中患者数据包括患者影像、医嘱数据；患者影像包括CT影像、核磁影像或PET影像中的一种或者组合；医嘱数据包括目标放疗剂量、DVH曲线、各器官放疗剂量约束值中的一种或组合；
步骤420，根据患者影像进行靶区勾画；勾画为自动勾画或者手动勾画；
步骤430，主控云服务器将计算任务分配给受控计算机，受控计算机用标准放疗设备模式计算患者的放疗计划，生成标准放疗计划；
步骤440，根据放疗设备的使用拥堵情况选择待转换放疗设备型号，优先将标准放疗计划转换为当前空闲的放疗设备或待执行任务数量较少的放疗设备或用户自定义选择待转换的放疗设备型号；
步骤450主控云服务器或受控计算机根据标准放疗计划，转换生成与特定放疗设备匹配的放疗计划，本步骤进一步包括：导入标准放疗计划，并将标准放疗计划中的剂量体积直方图，等剂量线作为约束条件，在标准放疗计划的射野参数基础上，重新计算最终放疗计划；重新计算包括剂量计算或者逆向优化中的一种或者组合；其中，逆向优化包括采用直接子野优化或者通量图优化方法中的一种或者组合；
可选的，还包括步骤460质量保证QA步骤(图4中未示出)，病人在治疗前通过QA验证所转换的计划是否正确，可选的，在进行质量保证步骤时，将特定放疗计划的执行对象更换为以固体水或其它人体组成替换材料制成的体模，测试该体模中接受的辐射剂量及剂量分布是否满足医嘱要求。当满足医嘱要求时，执行该特定放疗计划，如果不满足，返回步骤350，重新生成新的特定放疗计划。
图4所示的标准化云放疗计划方法除了上述说明的内容外，其他与图2或图3所示的方法相同。
本发明还提供了一种存储一个或多个程序的计算机可读存储介质，一个或多个程序包括指令，指令适于由存储器加载并执行上述标准化云放疗计划方法，该方法包括步骤：
向主控云服务器上传患者数据，其中患者数据包括患者影像、医嘱数据；
根据患者影像进行靶区勾画；
主控云服务器将计算任务分解后分配给受控计算机，受控计算机用标准放疗设备模式计算患者的放疗计划，生成标准放疗计划；
主控云服务器或受控计算机根据主控云服务器的分配将标准放疗计划转换生成与特定放疗设备匹配的特定放疗计划。
以上计算机程序被处理器执行时实现的步骤可以参照上文对方法和系统实施例的描述，并且，在不相冲突的前提下，上述系统实施例的内容和上述方法实施例的内容可以互为补充，对此不再予以赘述。
这些计算机程序指令也可存储在能引导计算机或其他可编程数据处理设备以特定方式工作的计算机可读存储器中，使得存储在该计算机可读存储器中的指令产生包括指令装置的制造品，该指令装置实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能。
这些计算机程序指令也可装载到计算机或其他可编程数据处理设备上，使得在计算机或其他可编程设备上执行一系列操作步骤以产生计算机实现的处理，从而在计算机或其他可编程设备上执行的指令提供用于实现在流程图一个流程或多个流程和/或方框图一个方框或多个方框中指定的功能的步骤。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括暂存电脑可读媒体(transitorymedia)，如调制的数据信号和载波。
通过本发明上述实施例提供的标准化云放疗计划方法，可以避免因为医院内某类型机器发生故障而其他机器可用的情况下耽误患者的治疗时间以及治疗资源的闲置；通过主控云服务将放疗计划的计算任务分配给受控计算机，使难以应 用于临床的放疗剂量计算的“金标准”—蒙特卡罗粒子输运模拟剂量计算应用于临床成为可能，大大节省放射治疗计划的制定时间和患者的等待时间，提高放疗计划剂量计算的准确度。另外，通过自动勾画及自动TPS制定，还能平衡不同医院或地区的医生的治疗水平的差异，也能减轻肿瘤医生与物理师的工作负担。
上述的对实施例的描述是为便于该技术领域的普通技术人员能理解和应用本发明。熟悉本领域技术的人员显然可以容易地对这些实施例做出各种修改，并把在此说明的一般原理应用到其他实施例中而不必经过创造性的劳动。因此，本发明不限于这里的实施例，本领域技术人员根据本发明的揭示，不脱离本发明范畴所做出的改进和修改都应该在本发明的保护范围之内。

Claims (15)

  1. 一种标准化云放疗计划方法，适于在标准化云放疗计划系统中执行，其特征在于：包括如下步骤：
    (1)向主控云服务器上传患者数据，其中所述的患者数据包括患者影像、医嘱数据；
    (2)根据患者影像进行靶区勾画；
    (3)主控云服务器将计算任务分解后分配给受控计算机，受控计算机用标准放疗设备模式计算患者的放疗计划，生成标准放疗计划；
    (4)主控云服务器或受控计算机根据标准放疗计划，转换生成与特定放疗设备匹配的特定放疗计划。
  2. 根据权利要求1所述的标准化云放疗计划方法，其特征在于：所述的患者影像包括CT影像、核磁影像或PET影像中的一种或者组合。
  3. 根据权利要求1所述的标准化云放疗计划方法，其特征在于：所述的医嘱数据包括目标放疗剂量、DVH曲线、各器官放疗剂量约束值中的一种或组合。
  4. 根据权利要求1所述的标准化云放疗计划方法，其特征在于：所述的勾画为自动勾画、半自动勾画或者手动勾画。
  5. 根据权利要求1所述的标准化云放疗计划方法，其特征在于：所述的主控云服务器或受控计算机根据标准放疗计划，转换生成与特定放疗设备匹配的特定放疗计划进一步包括：
    导入标准放疗计划，并将标准放疗计划中的剂量体积直方图、等剂量线作为特定放疗计划的优化约束条件，在标准放疗计划的射野参数基础上，重新计算获得特定放疗计划。
  6. 根据权利要求5所述的标准化云放疗计划方法，其特征在于：步骤(4)中，所述的重新计算包括剂量计算或者逆向优化中的一种或者组合；其中，所述的逆向优化包括采用直接子野优化或者通量图优化方法中的一种或者组合。
  7. 根据权利要求1所述的标准化云放疗计划方法，其特征在于：步骤(4)后还包括质量保证步骤，病人在治疗前通过质量保证步骤验证所转换的计划是否正确，如果正确，执行该特定放疗计划；如果不正确，返回步骤(4)根据标准 放疗计划重新生成新的特定放疗计划。
  8. 根据权利要求1所述的标准化云放疗计划方法，其特征在于：步骤(3)与步骤(4)之间，还包括根据放疗设备的使用拥堵情况，选择待转换放疗设备的步骤，优先将标准放疗计划转换为当前空闲的放疗设备或待执行任务数量较少的放疗设备或用户自定义选择待转换的放疗设备型号。
  9. 根据权利要求1所述的标准化云放疗计划方法，其特征在于：生成标准放疗计划时或转换生成与特定放疗设备匹配的放疗计划前根据可用计算资源实行进程排队转换或用户自定义设定计算任务的优先次序。
  10. 根据权利要求1所述的标准化云放疗计划方法，其特征在于：所述的根据标准放疗计划，转换生成与特定放疗设备匹配的特定放疗计划之前还包括如下步骤：比较标准放疗设备与待匹配特定放疗设备的参数吻合度，如果参数吻合度符合预设的阈值要求，将标准放疗计划直接作为最终放疗计划，否则进入步骤(4)。
  11. 根据权利要求10所述的标准化云放疗计划方法，其特征在于：标准放疗设备与待匹配特定放疗设备需要比较的参数包括源参数、多叶准直器参数。
  12. 根据权利要求11所述的标准化云放疗计划方法，其特征在于：所述的源参数通过比较源在均匀或者非均匀介质中的剂量测量特性数据获得；所述的剂量测量特性数据通过三维剂量曲线获得；所述的多叶准直器参数包括叶片大小和对数、最大开野大小、是否允许交错。
  13. 根据权利要求1所述的标准化云放疗计划方法，其特征在于：步骤(4)中，将标准放疗计划转换为特定放疗计划时，设置转换生成与一台或多台特定放疗设备分别匹配的一个或多个特定放疗计划。
  14. 一种标准化云放疗计划系统，其特征在于：包括主控云服务器，网络通信模块，客户端以及受控计算机，其中：
    所述主控云服务器、受控计算机及客户端通过网络通信模块通信连接；
    所述主控云服务器用来定义计算模体、靶区勾画以及定义计算参数，分解计算任务，优化分配调度任务，并监控受控计算机执行；
    所述受控计算机用来接收主控云服务器发出的运行指令、判断任务执行、执 行计算任务、反馈计算进度与计算结果；
    所述的客户端用来将病人影像、病人数据或临床剂量上传到主控云服务器，并查看放疗计划结果。
  15. 一种存储一个或多个程序的计算机可读存储介质，所述的一个或多个程序包括指令，所述指令适于由存储器加载并执行上述权利要求1-13中任一所述的标准化云放疗计划方法。
PCT/CN2018/099445 2018-01-19 2018-08-08 一种标准化云放疗计划方法、存储介质和系统 WO2019140886A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP18900998.8A EP3742702B1 (en) 2018-01-19 2018-08-08 Standardized cloud radiotherapy planning method and storage medium
US16/645,839 US20200203022A1 (en) 2018-01-19 2018-08-08 Standardized Cloud Radiotherapy Planning Method, Storage Medium, and System

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810054052.5A CN110060765B (zh) 2018-01-19 2018-01-19 一种标准化云放疗计划系统和存储介质
CN201810054052.5 2018-01-19

Publications (1)

Publication Number Publication Date
WO2019140886A1 true WO2019140886A1 (zh) 2019-07-25

Family

ID=67301976

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/099445 WO2019140886A1 (zh) 2018-01-19 2018-08-08 一种标准化云放疗计划方法、存储介质和系统

Country Status (4)

Country Link
US (1) US20200203022A1 (zh)
EP (1) EP3742702B1 (zh)
CN (1) CN110060765B (zh)
WO (1) WO2019140886A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111437521A (zh) * 2020-04-23 2020-07-24 华东医院 一种非均匀容积弧形调强方法

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110465004A (zh) * 2019-08-02 2019-11-19 北京全域医疗技术集团有限公司 一种云放射治疗计划系统及放射治疗计划的生成方法
CN110975172B (zh) * 2019-12-18 2022-05-31 上海联影医疗科技股份有限公司 一种通量图重建方法和系统
CN111388879B (zh) * 2020-03-19 2022-06-14 上海联影医疗科技股份有限公司 一种放射剂量确定系统、装置及存储介质
CN112057753B (zh) * 2020-09-23 2022-08-16 上海联影医疗科技股份有限公司 一种放疗计划调整系统和装置
CN112336996A (zh) * 2020-09-30 2021-02-09 四川大学 基于深度神经网络的放疗靶区自动勾画系统
CN112837796A (zh) * 2021-02-09 2021-05-25 安徽医科大学第二附属医院 一种基于遗传算法的后疫情时期癌症患者放疗排程系统
CN114201459B (zh) * 2021-11-03 2024-06-25 中山大学肿瘤防治中心（中山大学附属肿瘤医院、中山大学肿瘤研究所） 一种改进的放疗结构标准化命名方法、装置及存储介质
CN115966281B (zh) * 2023-01-16 2024-02-09 迈胜医疗设备有限公司 一种弧形放疗计划的生成方法、装置、设备及存储介质

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103247046A (zh) 2013-04-19 2013-08-14 深圳先进技术研究院 一种放射治疗计划中靶区自动勾画的方法和装置
CN105688335A (zh) * 2015-11-23 2016-06-22 北京全域医疗技术有限公司 云放疗远程协作的方法
CN105893772A (zh) * 2016-04-20 2016-08-24 上海联影医疗科技有限公司 用于放射治疗计划的数据获取方法和装置
US20170143994A1 (en) * 2015-11-23 2017-05-25 Board Of Regents, The University Of Texas System Systems and Methods for Cloud-Based Radiation Therapy Treatment Planning
WO2017133654A1 (en) * 2016-02-02 2017-08-10 Suzhou Evidance Medical Technologies Inc. Systems and methods for radiation treatment planning
CN107480416A (zh) * 2016-06-08 2017-12-15 瑞地玛医学科技有限公司 一种放射治疗计划系统

Family Cites Families (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100408127C (zh) * 2002-08-09 2008-08-06 定锦霞 远程立体定向放射治疗系统
CN1438602A (zh) * 2003-03-17 2003-08-27 吴大可 处方剂量自动计算的放射治疗系统
US8467497B2 (en) * 2007-10-25 2013-06-18 Tomotherapy Incorporated System and method for motion adaptive optimization for radiation therapy delivery
RU2571374C2 (ru) * 2009-08-31 2015-12-20 Конинклейке Филипс Электроникс, Н.В. Интерактивный компьютеризованный редактор для компенсаторов, используемых в планировании лучевой терапии
US8477903B2 (en) * 2011-03-31 2013-07-02 Axellis Ventures Ltd Validating a compensator for use in a radiation therapy machine to treat a cancer patient
CN102306239B (zh) * 2011-07-22 2014-11-05 上海赛提菲克医疗器械有限公司 基于锥形束ct图像ct值校正技术的放疗剂量评估和优化方法
US9468776B2 (en) * 2012-06-01 2016-10-18 Raysearch Laboratories Ab Method and a system for optimizing a radiation treatment plan based on a reference dose distribution
CN103279660A (zh) * 2013-05-23 2013-09-04 浙江大学 放化疗规范化云中心系统
WO2015090457A1 (en) * 2013-12-20 2015-06-25 Raysearch Laboratories Ab Selection of radiotherapy treatment plans
CN104548372B (zh) * 2015-01-07 2017-12-22 上海联影医疗科技有限公司 放射治疗的剂量确定装置
CN104519140A (zh) * 2015-01-08 2015-04-15 浪潮(北京)电子信息产业有限公司 一种分布式并行计算的服务器系统及其管理方法
CN104815392B (zh) * 2015-04-10 2018-01-26 石峰 一种交互式放射治疗计划系统优化系统
CN105031819B (zh) * 2015-08-25 2018-11-06 上海联影医疗科技有限公司 一种剂量优化系统
WO2017174189A1 (en) * 2016-04-05 2017-10-12 Varian Medical System Particle Therapy Gmbh Time optimized radiation treatment
CN106139424A (zh) * 2016-08-02 2016-11-23 丽水市人民医院 一种精确放射治疗计划系统
CN106920234B (zh) * 2017-02-27 2021-08-27 北京连心医疗科技有限公司 一种复合式自动放疗计划的方法
CN107391920B (zh) * 2017-07-13 2019-03-05 中科超精（安徽）科技有限公司 一种dicom计划文件兼容转换系统和方法
CN107545137A (zh) * 2017-08-16 2018-01-05 强深智能医疗科技（昆山）有限公司 肿瘤放射治疗计划智能优化方法
CN107491659B (zh) * 2017-09-20 2022-03-15 上海联影医疗科技股份有限公司 一种医疗设备系统升级方法和装置

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103247046A (zh) 2013-04-19 2013-08-14 深圳先进技术研究院 一种放射治疗计划中靶区自动勾画的方法和装置
CN105688335A (zh) * 2015-11-23 2016-06-22 北京全域医疗技术有限公司 云放疗远程协作的方法
US20170143994A1 (en) * 2015-11-23 2017-05-25 Board Of Regents, The University Of Texas System Systems and Methods for Cloud-Based Radiation Therapy Treatment Planning
WO2017133654A1 (en) * 2016-02-02 2017-08-10 Suzhou Evidance Medical Technologies Inc. Systems and methods for radiation treatment planning
CN105893772A (zh) * 2016-04-20 2016-08-24 上海联影医疗科技有限公司 用于放射治疗计划的数据获取方法和装置
CN107480416A (zh) * 2016-06-08 2017-12-15 瑞地玛医学科技有限公司 一种放射治疗计划系统

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
"Automatic delineation evaluation of nasopharyngeal carcinoma target area", SICHUAN MEDICAL, vol. 36, no. 6, June 2015 (2015-06-01), pages 762 - 765
"World Cancer Report 2014", 3 February 2014, WORLD HEALTH ORGANIZATION (WHO
DAN NLONG TJIA X, DOSE PREDICTION WITH U -NET: A FEASIBILITY STUDY FOR PREDICTING DOSE DISTRIBUTIONS FROM CONTOURS USING DEEP LEARNING ON PROSTATE IMRT PATIENTS [J, 2017
RONNEBERGER, O.FISCHER, P.BROX, T.: "U-net: Convolutional networks for biomedical image segmentation", INTERNATIONAL CONFERENCE ON MEDICAL IMAGE COMPUTING AND COMPUTER-ASSISTED INTERVENTION, October 2015 (2015-10-01), pages 234 - 241, XP055561106, DOI: 10.1007/978-3-319-24574-4_28
See also references of EP3742702A4

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111437521A (zh) * 2020-04-23 2020-07-24 华东医院 一种非均匀容积弧形调强方法
CN111437521B (zh) * 2020-04-23 2023-01-31 华东医院 一种非均匀容积弧形调强方法

Also Published As

Publication number Publication date
CN110060765B (zh) 2022-06-17
EP3742702B1 (en) 2024-02-21
CN110060765A (zh) 2019-07-26
EP3742702A4 (en) 2021-11-10
EP3742702A1 (en) 2020-11-25
US20200203022A1 (en) 2020-06-25

Similar Documents

Publication Publication Date Title
WO2019140886A1 (zh) 一种标准化云放疗计划方法、存储介质和系统
US11056243B2 (en) Systems and methods for optimizing treatment planning
RU2684173C2 (ru) Система и способ для автоматического планирования лечения
US11992702B2 (en) Machine learning optimization of fluence maps for radiotherapy treatment
CN102812465B (zh) 包括高效剂量计算的剂量计算任务的安排
Wang et al. Fluence map prediction using deep learning models–direct plan generation for pancreas stereotactic body radiation therapy
US11282192B2 (en) Training deep learning engines for radiotherapy treatment planning
CN113891742B (zh) 用于基于连续深度学习的放射疗法治疗规划的方法和系统
CN110504016A (zh) 一种蒙特卡罗网格并行剂量计算方法、设备和存储介质
US20180161596A1 (en) Methods and systems for generating dose estimation models for radiotherapy treatment planning
US20240123258A1 (en) Cloud-based dose verification
WO2020257216A1 (en) Methods and systems for quality-aware continuous learning for radiotherapy treatment planning
CN108697905B (zh) 放射治疗规划设备、方法和存储介质
WO2022243391A1 (en) Iterative image reconstruction
Roberfroid et al. Comparison of Ethos template-based planning and AI-based dose prediction: General performance, patient optimality, and limitations
US11983869B2 (en) Feature-space clustering for physiological cycle classification
WO2022271197A1 (en) Radiotherapy optimization for arc sequencing and aperture refinement
Montero Robust, accurate and patient-specific treatment planning for proton therapy
Chow Performance optimization in 4D radiation treatment planning using Monte Carlo simulation on the cloud
Della-Giustina et al. Hybrid MPI+ OpenMP parallelization of image reconstruction in proton beam therapy on multi-core and many-core processors
Neph Accelerating Radiation Dose Calculation with High Performance Computing and Machine Learning for Large-scale Radiotherapy Treatment Planning
US20230372739A1 (en) Techniques for detecting movement during radiotherapy treatment
Zhu A Deep-Learning-based Multi-segment VMAT Plan Generation from Patient Anatomy for Prostate Simultaneous Integrated Boost (SIB) Cases: A Feasibility Study of Prostate Radiotherapy Application
Liu et al. A deep learning-based dose prediction method for evaluation of radiotherapy treatment planning
CN112017166A (zh) 一种基于cpu/gpu混合集群的计算放疗笔形束射线剂量分布的方法

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: 18900998

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2018900998

Country of ref document: EP

Effective date: 20200819