Radiation therapy anti-collision method and system based on three-dimensional scanning and real-time simulation

Abstract

The embodiment of the specification provides a radiotherapy anti-collision method and system based on three-dimensional scanning and real-time simulation. The method comprises the following steps: generating an initial three-dimensional model based on body surface data of a patient; performing simulation simulation of anti-collision in a radiotherapy process based on the initial three-dimensional model, and determining a first planning path in the radiotherapy process.

Description

TECHNICAL FIELD

[0001] The present specification relates to the field of radiotherapy collision avoidance, and in particular, to a radiotherapy collision avoidance method and system based on three-dimensional scanning and real-time simulation. BACKGROUND

[0002] In the process of radiotherapy, it is very important to prevent collision between the patient, the treatment bed, the treatment head, and the equipment in the treatment room. However, the conventional collision avoidance method has blind spots, such as the collision avoidance technology based on CT or camera alone fails to fully consider the whole body posture of the patient, especially the limbs and the impact of the supporting structure. Therefore, it is desirable to provide a radiotherapy collision avoidance method and system based on three-dimensional scanning and real-time simulation. SUMMARY

[0003] One or more embodiments of the present specification provide a radiotherapy collision avoidance method based on three-dimensional scanning and real-time simulation, the method comprising: generating an initial three-dimensional model based on body surface data of a patient, the initial three-dimensional model comprising at least three-dimensional information of the patient; performing a simulation of collision avoidance in a radiotherapy process based on the initial three-dimensional model to determine a first planned path in the radiotherapy process.

[0004] One or more embodiments of the present specification provide a radiotherapy collision avoidance system based on three-dimensional scanning and real-time simulation, the system comprising a model generation module and a simulation module; the model generation module is configured to generate an initial three-dimensional model based on body surface data of a patient, the initial three-dimensional model comprising at least three-dimensional information of the patient; the simulation module is configured to perform a simulation based on the initial three-dimensional model to determine a first planned path.

[0005] One or more embodiments of the present specification provide a computer-readable storage medium, the storage medium stores computer instructions, when the computer reads the computer instructions in the storage medium, the computer executes the method as described in the above embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0006] The present specification will be further described in the manner of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, the same numbers represent the same structures, wherein:

[0007] Figure 1 is a system schematic diagram of a radiotherapy collision avoidance system based on three-dimensional scanning and real-time simulation according to some embodiments of the present specification;

[0008] Figure 2 is an exemplary flow schematic diagram of a radiotherapy collision avoidance method based on three-dimensional scanning and real-time simulation according to some embodiments of the present specification;

[0009] Figure 3 is an exemplary schematic diagram of a risk prediction model according to some embodiments of the present specification;

[0010] Figure 4 is an exemplary flowchart diagram of obtaining a second planning path according to some embodiments of the present specification;

[0011] Figure 5 is an exemplary flowchart diagram of determining a first planning path according to some embodiments of the present specification. DETAILED DESCRIPTION

[0012] In order to more clearly illustrate the technical solutions of the embodiments of the present specification, the following will briefly introduce the drawings needed to be used in the embodiments description. The drawings do not represent all the embodiments.

[0013] It should be understood that the "system", "device", "unit" and / or "module" used herein is a method for distinguishing different components, elements, parts, portions or assemblies at different levels. If other words can achieve the same purpose, the words can be replaced by other expressions.

[0014] Unless the context clearly indicates otherwise, the words "one", "an", "a", and / or "the" do not mean a single number, but can also include a plurality. Generally, the terms "comprise" and "include" only indicate the inclusion of the steps and elements explicitly identified, and these steps and elements do not constitute an exclusive list, and the method or device can also include other steps or elements.

[0015] In the embodiments of the present specification, the operations performed in steps are exchangeable in order unless otherwise specified, and the steps can be omitted, and other steps can be included in the operation process.

[0016] Figure 1 is a system schematic diagram of a radiotherapy collision avoidance system based on three-dimensional scanning and real-time simulation according to some embodiments of the present specification.

[0017] In some embodiments, the radiotherapy collision avoidance system 100 based on three-dimensional scanning and real-time simulation includes a model generation module 110 and a simulation simulation module 120.

[0018] The model generation module is configured to generate an initial three-dimensional model based on the body surface data of the patient.

[0019] The simulation simulation module is configured to perform simulation simulation of collision avoidance during radiotherapy based on the initial three-dimensional model, and determine a first planning path during radiotherapy.

[0020] In some embodiments, the real-time simulation based radiotherapy collision avoidance system 100 further comprises a processor and a memory.

[0021] The processor is configured to process data from at least one module of the real-time simulation based radiotherapy collision avoidance system 100 or an external data source. In some embodiments, the processor comprises a central processing unit, an application specific integrated circuit, an image processing unit, a controller, etc. or any combination thereof.

[0022] The memory is configured to store data, instructions and / or any other information. In some embodiments, the memory comprises a mass storage, a removable storage, etc. or any combination thereof.

[0023] For detailed description of the foregoing, please refer to the relevant description of Figures 2 to 5 .

[0024] It should be understood that, Figure 1 the real-time simulation based radiotherapy collision avoidance system and its modules shown can be implemented in various ways. It should be noted that the above description of the system and its modules is for the convenience of description, and cannot limit the scope of the embodiments. It can be understood that, for those skilled in the art, after understanding the principle of the system, any combination of the modules or connection of the modules with other modules can be made without departing from the principle. In some embodiments, Figure 1 the model generation module 110 and the simulation simulation module 120 disclosed in the above can be different modules in a system, or can be a module to realize the functions of two or more modules described above. For example, the modules can share a storage module, and each module can have its own storage module. Variations such as this are within the scope of protection of the present specification.

[0025] Figure 2 is an exemplary flowchart of a real-time simulation based radiotherapy collision avoidance method according to some embodiments of the present specification. In some embodiments, the flow 200 is performed by a real-time simulation based radiotherapy collision avoidance system (hereinafter referred to as a collision avoidance system).

[0026] In some embodiments, the collision avoidance system generates an initial three-dimensional model 220 based on the body surface data 210 of the patient, and performs a simulation simulation of collision avoidance during radiotherapy based on the initial three-dimensional model 220 to determine a first planned path 230 during radiotherapy.

[0027] For related content of the collision avoidance system, please refer to the corresponding description of Figure 1 .

[0028] The body surface data refers to data related to the body surface of the patient. For example, at least one of the body shape, height, size of each part, posture, and the like of the patient. The patient includes a person who needs to receive radiotherapy.

[0029] In some embodiments, the anti-collision system is communicatively connected with the data acquisition device, and acquires the body surface data through the data acquisition device.

[0030] In some embodiments, the data acquisition device can be deployed around the patient or any feasible position, including at least one of a 3D scanning device, an optical camera, and an ultrasonic scanning device, and the like. The 3D scanning device includes a laser or a photosensitive sensor, and the like.

[0031] The body surface data acquired by the 3D scanning device is represented by a three-dimensional space point cloud. The body surface data acquired by the optical camera is represented in the form of an image. The body surface data acquired by the ultrasonic scanning device is represented by reflection data of ultrasonic waves.

[0032] The initial three-dimensional model refers to a three-dimensional model obtained based on the body surface data. The initial three-dimensional model includes three-dimensional information of the patient, and the three-dimensional information refers to information corresponding to the body surface data in the three-dimensional model. For example, the three-dimensional information includes the size of the three-dimensional model corresponding to the body shape of the patient, and the like.

[0033] In some embodiments, the anti-collision system generates the initial three-dimensional model based on the body surface data through any feasible modeling method (such as factor decomposition method, neural network method, and the like) and / or external modeling software.

[0034] In some embodiments, the anti-collision system can also acquire structure data of a support structure supporting the patient, and generate the initial three-dimensional model based on the body surface data and the structure data.

[0035] The support structure refers to a structure used for supporting the body of the patient or in contact with the body of the patient during radiotherapy. For example, a bed on which the patient receives treatment, a device connected to the body surface of the patient, and the like.

[0036] The structure data refers to data related to the physical size of the support structure. For example, at least one of the size of the support structure, the relative position to the patient, and the like.

[0037] In some embodiments, the anti-collision system acquires the structure data through the data acquisition device. The data form of the structure data is similar to that of the body surface data.

[0038] In some embodiments, the anti-collision system generates the initial three-dimensional model based on the structure data and the body surface data by any feasible modeling method and / or external modeling software. If the anti-collision system generates the initial three-dimensional model based on the structure data and the body surface data, the three-dimensional information includes information corresponding to the structure data in the three-dimensional model. For example, the size of the three-dimensional model corresponding to the support structure.

[0039] In some embodiments of the present specification, the support structure related to the patient is also considered when establishing the initial three-dimensional model, so as to construct a more accurate initial three-dimensional model, which is beneficial to improve the authenticity and reliability of subsequent simulation.

[0040] The planned path refers to a path indicating the movement of the treatment device. The first planned path refers to a planned path set before the start of radiotherapy. The treatment device refers to a device (such as a CT machine) that performs radiotherapy. The anti-collision system is in communication connection with the treatment device.

[0041] In some embodiments, the anti-collision system generates the movement control instruction based on the first planned path, and sends the movement control instruction to the treatment device to control the treatment device to move in the first planned path after the start of radiotherapy, and perform radiotherapy on the patient.

[0042] In some embodiments, the anti-collision system queries the reference path corresponding to the body surface data and the structure data in the path preset table as the first planned path based on the body surface data and the structure data. The path preset table is set in advance based on historical data and includes multiple sets of body surface data and structure data, and the reference path corresponding to each set of data.

[0043] The anti-collision system counts the historical body surface data and the historical structure data in the historical radiotherapy process as a set of data in the path preset table, and counts the first planned path that actually does not collide in the historical radiotherapy process as the reference path in the path preset table. The reference path includes the coordinates of a plurality of spatial points arranged in order.

[0044] In some embodiments, the anti-collision system performs simulation of anti-collision in the radiotherapy process based on the initial three-dimensional model to determine the first planned path.

[0045] The simulation refers to the simulation of the movement of the treatment device.

[0046] In some embodiments, to prevent collision between the patient and / or the support structure and the treatment device during the radiotherapy, the anti-collision system selects any one of the candidate first paths in which no collision occurs in the simulation simulation based on a plurality of candidate first paths by simulation simulation through simulation software. The simulation software includes any feasible simulation software such as simulation software built-in the system or external simulation software (such as ADAMS, SolidWorks, etc.).

[0047] The candidate first path refers to a planning path to be determined as the first planning path. In some embodiments, the anti-collision system obtains a plurality of candidate first paths through user input or the like. For example, the candidate first paths include paths connected by a plurality of spatial points arranged in order, and the positions of the spatial points in different candidate first paths are different or partially the same. The user includes a doctor or the like. The user can manually set a plurality of planning paths as the candidate first paths based on the treatment needs of the patient. The treatment needs include at least one of the parts that need to be treated by radiotherapy, the amount of radiation or the length of radiation required for each part, and the like.

[0048] In some embodiments, the process of the anti-collision system performing simulation simulation based on a single candidate first path includes: the anti-collision system sets an initial three-dimensional model in the simulation software, and sets a motion path of a virtual device based on the candidate first path, so that the virtual device performs virtual motion in the simulation software, and records whether the virtual device collides with the initial three-dimensional model during the virtual motion. The virtual device refers to a model for simulating the treatment device in the simulation software.

[0049] In some embodiments, the anti-collision system can also perform at least one simulation simulation based on the initial three-dimensional model and the target path library, and determine the first planning path based on the simulation results of the at least one simulation simulation. For details of this part, see Figure 5 and the related description.

[0050] In some embodiments of the present specification, a three-dimensional model that can accurately reflect the patient's body surface data is constructed based on the complete body surface data of the patient, which can effectively eliminate the error influence of the body surface data, and thus effectively improve the accuracy and efficiency of path planning. Through simulation simulation, the collision that may occur during the movement of the treatment device can be effectively predicted, and thus the motion path of the treatment device can be adjusted to reduce the possibility of real collision with the patient.

[0051] In some embodiments, the anti-collision system can also obtain real-time scanning data of the patient during the treatment, update the initial three-dimensional model based on the real-time scanning data to obtain a real-time three-dimensional model, and perform simulation simulation based on the real-time three-dimensional model.

[0052] The real-time scanning data refers to the body surface data of the patient acquired during the radiotherapy. When the posture of the patient changes during the radiotherapy, the real-time scanning data is inconsistent with the body surface data acquired before the radiotherapy.

[0053] In some embodiments, the real-time scanning data can further include structure data acquired during the radiotherapy.

[0054] In some embodiments, the data acquisition device can continuously acquire the body surface data and the structure data as the real-time scanning data during the radiotherapy, and send the real-time scanning data to the collision avoidance system.

[0055] The real-time three-dimensional model refers to a three-dimensional model obtained by updating the initial three-dimensional model.

[0056] In some embodiments, the collision avoidance system updates the initial three-dimensional model based on the real-time scanning data by using a real-time three-dimensional reconstruction technology (such as BundleFusion, etc.), and obtains a real-time three-dimensional model.

[0057] Exemplarily, for the real-time scanning data acquired by the optical camera, the collision avoidance system can pre-process the real-time scanning data, extract information related to the patient and / or the support structure in the image by using image recognition and processing technology, etc., convert the obtained information into three-dimensional information, and update the initial three-dimensional model by using a real-time three-dimensional reconstruction technology (such as BundleFusion, etc.) to obtain a real-time three-dimensional model. The pre-processing includes at least one of removing noise, adjusting resolution, and cropping a region of interest. The region of interest includes a region of the patient that needs to be treated by radiotherapy, etc.

[0058] In some embodiments, after obtaining the real-time three-dimensional model, the collision avoidance system can further update the real-time three-dimensional model again by using the above method of updating the initial three-dimensional model based on subsequent real-time scanning data.

[0059] In some embodiments, after obtaining the real-time three-dimensional model, the collision avoidance system can further perform simulation simulation again based on the real-time three-dimensional model to determine whether the patient and / or the support structure can collide with the treatment device.

[0060] In some embodiments of the present specification, by monitoring the relevant information of the patient and / or the support structure in real time, and updating the initial three-dimensional model based on the obtained information, the accuracy of the three-dimensional model can be effectively improved, and the accuracy of the simulation simulation can be improved, which is beneficial to the subsequent operation.

[0061] In some embodiments, the collision avoidance system can further determine a first collision risk based on the real-time three-dimensional model and the first planned path, and generate a warning information and issue a warning in response to the first collision risk meeting a warning condition.

[0062] The first collision risk refers to a collision risk corresponding to the first future time. The collision risk is used to represent the possibility of collision between the treatment device and the patient and / or the support structure when the treatment device is moving. In some embodiments, the collision risk is represented by a numerical value or the like, and the larger the numerical value, the higher the collision risk.

[0063] The first future time refers to a future time after the real-time three-dimensional model is obtained.

[0064] In some embodiments, the first future time includes at least one future time point. The number of future time points is related to the motion characteristics of the treatment device.

[0065] The motion characteristics are used to represent the motion of the treatment device. In some embodiments, the motion characteristics include at least one of the motion speed, the motion acceleration, or the like. The number of future time points can be positively correlated with the motion speed or the motion acceleration, and the larger the motion speed or the motion acceleration, the more the number of future time points.

[0066] In some embodiments of the present specification, the number of future time points is related to the motion characteristics of the treatment device, so that more future time points can be determined when the treatment device is running faster, and the collision risk in a longer future time can be predicted, thereby improving the safety of radiotherapy.

[0067] In some embodiments, the anti-collision system determines the first collision risk based on the real-time three-dimensional model and the first planned path in various ways. For example, the anti-collision system performs simulation again based on the real-time three-dimensional model and the first planned path, and counts the number of collisions occurring in the simulation process. The more the number of collisions, the greater the first collision risk.

[0068] For another example, the anti-collision system converts and / or fuses the real-time three-dimensional model and the first planned path to obtain a path vector, clusters the path vector in a plurality of historical path vectors with the path vector as a clustering center, and determines the first collision risk based on a vector set corresponding to the clustering center. The anti-collision system can extract a preset number of historical path vectors in the vector set, obtain historical collision results corresponding to the historical path vectors from historical data, calculate a ratio of the number of collisions in the historical collision results to the preset number, and take the ratio as the first collision risk. The preset data is set in advance based on historical experience.

[0069] The historical path vector refers to a feature vector obtained by converting and / or fusing a historical real-time three-dimensional model and a historical first planned path. The historical collision result corresponding to the historical path vector refers to the actual collision result occurring in the historical radiotherapy process corresponding to the historical path vector. The collision result includes collision and no collision.

[0070] In some embodiments, the anti-collision system can further determine, based on the real-time three-dimensional model and the first planning path, a first collision risk at each future time point in the first future time through a risk prediction model. Figure 3 and the related description.

[0071] The warning condition refers to a condition for determining whether to generate a warning information and issue a warning. In some embodiments, the warning condition is pre-set based on historical experience. For example, the first collision risk is greater than a risk threshold. The risk threshold is pre-set based on historical experience.

[0072] In some embodiments, if the anti-collision system determines the first collision risk at each future time point in the first future time through the risk prediction model, the warning condition includes that the average value of the first collision risks at the plurality of future time points is greater than the risk threshold, and the like.

[0073] The warning information refers to information related to the content of the warning. For example, reminding the user that the first collision risk is high or reminding the patient not to move around. In some embodiments, the way of issuing a warning includes a prompt sound or any feasible way.

[0074] In some embodiments of the present specification, the collision risk at the future time is dynamically calculated through the real-time three-dimensional model and the first planning path, which improves the accuracy and timeliness of the collision risk assessment. The warning helps to take measures to avoid collision in time, and ensures the safety of the patient and the treatment device.

[0075] In some embodiments, the anti-collision system can further correct the first planning path based on the real-time three-dimensional model to obtain a second planning path, and determine a second collision risk based on the real-time three-dimensional model and the second planning path.

[0076] In some embodiments, in response to the second collision risk satisfying the warning condition, the anti-collision system generates a warning information and issues a warning, and / or updates the second planning path.

[0077] The second planning path refers to a path for instructing the movement of the treatment device after obtaining the real-time three-dimensional model.

[0078] In some embodiments, the anti-collision system generates a movement control instruction based on the second planning path, and sends the movement control instruction to the treatment device to control the treatment device to move in the second planning path and perform radiotherapy on the patient.

[0079] In some embodiments, the anti-collision system corrects the first planning path based on the real-time three-dimensional model to obtain a plurality of candidate second paths, and simulates each candidate second path to obtain a second planning path. The candidate second path refers to a planning path to be determined as the second planning path. The anti-collision system can select any one of the candidate second paths in which no collision occurs in the simulation as the second planning path.

[0080] In some embodiments, the anti-collision system corrects the first planning path in multiple ways. For example, the anti-collision system obtains the corrected first planning path (i.e., the candidate second path) through user input. The user can manually correct the first planning path to generate a plurality of candidate second paths and input the anti-collision system.

[0081] For another example, the anti-collision system can divide the first planning path into a plurality of spatial points based on a preset distance, and randomly adjust the coordinates of one or more spatial points to complete the correction of the first planning path and obtain a plurality of candidate second paths. The preset distance is pre-set based on historical experience.

[0082] The second collision risk refers to the collision risk corresponding to a second future time. The second future time refers to a future time after the second planning path is obtained. In some embodiments, the second future time is later than or equal to the first future time. The second future time includes at least one future time point.

[0083] In some embodiments, the anti-collision system determines the second collision risk in a similar manner to the method of determining the first collision risk, and the implementation method can refer to the method of determining the first collision risk.

[0084] In some embodiments, the warning condition further includes that the second collision risk is greater than a risk threshold, etc. In response to the second collision risk meeting the warning condition, the anti-collision system generates warning information and issues a warning and / or updates the second planning path. Wherein, the anti-collision system obtains the updated second planning path, and can continue to determine the second collision risk corresponding to the updated second planning path, and in response to the new second collision risk meeting the warning condition, update the second planning path again.

[0085] In some embodiments, the anti-collision system updates the second planning path in multiple ways. For example, the anti-collision system updates the second planning path by the above-mentioned method of determining the second planning path. For another example, the anti-collision system directionally adjusts the second planning path based on the second planning path, and simulates the adjusted second planning path. If no collision occurs in the simulation, the adjusted second planning path is taken as the new second planning path, and if a collision occurs in the simulation, the second planning path continues to be directionally adjusted.

[0086] In some embodiments, the adjusting the second planning path includes: the anti-collision system divides the second planning path into a plurality of spatial points based on a preset distance, and adjusts the coordinates of the spatial points that are prone to collision towards a direction away from the patient, thereby obtaining the adjusted second planning path. The adjustment range is preset based on historical experience. The anti-collision system can regard the spatial points that are too close to the real-time three-dimensional model or have overlaps as the spatial points that are prone to collision.

[0087] In some embodiments of the present specification, the first planning path is corrected based on the real-time three-dimensional model to obtain the second planning path, so as to dynamically adjust the movement path of the treatment equipment during the radiotherapy process, and improve the accuracy, safety and intelligent level of the radiotherapy process.

[0088] Figure 3 is an exemplary schematic diagram of the risk prediction model according to some embodiments of the present specification.

[0089] In some embodiments, the anti-collision system determines the first collision risk 350 based on the real-time three-dimensional model 310 and the first planning path 230 through the risk prediction model 340. For the description of the real-time three-dimensional model, the first planning path and the first collision risk, please refer to Figure 2 and the related description.

[0090] The risk prediction model refers to a model for determining the first collision risk. In some embodiments, the risk prediction model can be a machine learning model. For example, the risk prediction model can include any one or combination of a convolutional neural network (CNN) model, a neural network (NN) model or other custom model structure.

[0091] In some embodiments, the input of the risk prediction model includes the real-time three-dimensional model and the first planning path, and the output includes the first collision risk at each future time point in the first future time.

[0092] In some embodiments, the anti-collision system trains the risk prediction model through gradient descent method or the like based on a large number of first training samples with first labels. The first training sample includes a sample real-time three-dimensional model and a sample first planning path. The first label of the first training sample can be the actual collision risk at each sample future time point in the sample first future time corresponding to the first training sample.

[0093] In some embodiments, the first training samples and the first labels are determined based on historical data. For example, the anti-collision system statistically records historical real-time three-dimensional models and historical first planning paths in a historical radiotherapy process as the first training samples. For the first training samples in which no collision actually occurs, the labels of all time points corresponding to the first training samples are set to 0. For the first training samples in which a collision actually occurs, the label of the time point at which the collision occurs is set to 1, and the labels of the time points at which no collision occurs are set to a value between 0 and 1. The closer the time point at which no collision occurs to the time point at which the collision occurs, the closer the label of the time point at which no collision occurs to 1.

[0094] In some embodiments, the risk prediction model can be trained in the following manner: the plurality of first training samples with the first labels are input into an initial risk prediction model, a loss function is constructed based on the first labels and the prediction results of the initial risk prediction model, the initial risk prediction model is updated iteratively based on the loss function, and the risk prediction model is trained when the loss function of the initial risk prediction model meets a preset condition. The preset condition can be that the loss function converges, the number of iterations reaches a set value, or the like.

[0095] In some embodiments, the input of the risk prediction model further includes the number of data blind areas 320 and the part activity of each data blind area 330.

[0096] A data blind area refers to a part of the patient's body for which the data acquisition device cannot collect surface data. For example, a part of the patient's body that is blocked by the treatment device during the movement of the treatment device.

[0097] In some embodiments, the anti-collision system determines the data blind area based on the real-time scanning data. For example, if the data acquisition device is an optical camera, the anti-collision system determines the data blind area by identifying and processing the part of the patient's body that is blocked by the treatment device based on the real-time scanning data. For another example, if the data acquisition device is a 3D scanning device, the anti-collision system can directly determine the part of the patient's body that is blocked by the treatment device based on the real-time scanning data, because there is a difference between the three-dimensional point cloud obtained by scanning the human body and the three-dimensional point cloud obtained by scanning the treatment device.

[0098] Part activity refers to the activity level of the part of the patient's body corresponding to the data blind area. The part activity corresponding to each data blind area is determined separately. For how to determine the part activity, see Figure 4 and the related description.

[0099] In some embodiments, if the input of the risk prediction model further includes the number of data blind areas and the part activity corresponding to each data blind area, the first training samples further include the number of sample data blind areas and the sample part activity corresponding to each sample data blind area.

[0100] In some embodiments of the present specification, the part activity of the data blind area helps to reflect the motion tendency of the part corresponding to the data blind area of the patient without data support. Adding the part activity of the data blind area in the input of the risk prediction model helps to improve the accuracy of the output first collision risk.

[0101] In some embodiments of the present specification, by processing the real-time three-dimensional model and the first planning path and other data through the risk prediction model, the self-learning ability of the machine learning model can be utilized to find the rules from a large amount of data, and the correlation between the real-time three-dimensional model and the first planning path and other data and the first collision risk can be obtained, thereby improving the accuracy and efficiency of determining the first collision risk of the user.

[0102] Figure 4 is an exemplary flowchart of obtaining a second planning path according to some embodiments of the present specification. In some embodiments, the flow 400 is performed by a radiotherapy anti-collision system based on three-dimensional scanning and real-time simulation (hereinafter referred to as anti-collision system).

[0103] Step 410, determining at least one safety buffer.

[0104] The safety buffer refers to the space reserved to prevent the patient from colliding with the treatment equipment. For example, the safety buffer is the space obtained by expanding the local contour of the three-dimensional model by a certain distance, so the size of the safety buffer can be represented by the expansion distance of the safety buffer.

[0105] In some embodiments, the anti-collision system determines at least one safety buffer in multiple ways. For example, the anti-collision system sets the safety buffer based on the model boundary and the preset space size. The model boundary refers to the boundary points of the initial three-dimensional model or the real-time three-dimensional model and the edges formed by connecting the boundary points. The preset space size is set in advance based on historical experience, such as a space of 5 cm expanded outward from the model boundary.

[0106] In some embodiments, the anti-collision system can also determine at least one safety buffer based on the regional characteristics of the data blind area. For the description of the data blind area, see Figure 3 and related descriptions.

[0107] The regional characteristics refer to information related to the data blind area. In some embodiments, the regional characteristics include the patient body part corresponding to the data blind area, etc. The regional characteristics are determined by the anti-collision system at the same time when the data blind area is determined.

[0108] Exemplarily, the anti-collision system queries the preset feature table for a reference region feature identical to the region feature based on the region feature of the data blind area, and determines a reference buffer corresponding to the reference region feature as the safety buffer corresponding to the data blind area. Each data blind area corresponds to one safety buffer.

[0109] The preset feature table is preset based on historical data, and includes a plurality of reference region features and a reference buffer corresponding to each reference region feature. In some embodiments, the anti-collision system adds the historical region feature of the historical data blind area in the historical radiotherapy process as the reference region feature to the preset feature table, and adds the historical safety buffer (e.g., a 30cm space expanded outward from the data blind area) to the preset feature table.

[0110] In some embodiments, the region feature further includes a region area of the data blind area. The region area refers to an area occupied by the part corresponding to the data blind area. The anti-collision system can further determine the at least one safety buffer based on the region area. Exemplarily, the larger the region area, the greater the expansion distance of the safety buffer corresponding to the data blind area.

[0111] In some embodiments of the present specification, the larger the data blind area, the higher the potential collision risk, and increasing the size of the safety buffer corresponding to the data blind area with a larger region helps to further reduce the potential collision risk.

[0112] In some embodiments of the present specification, different safety buffers are set according to different data blind areas, which improves the flexibility of ensuring the safety of radiotherapy.

[0113] In some embodiments, the anti-collision system can further determine a part activity of the part corresponding to the data blind area based on the region feature, and determine the at least one safety buffer based on the region feature and the part activity. The safety buffer corresponding to each data blind area is determined based on the region feature and the part activity of the data blind area.

[0114] In some embodiments, the anti-collision system determines the activity frequency and the activity amplitude of the part of the patient before the part appears in the data blind area based on the real-time scanning data of the part of the patient corresponding to the region feature, calculates a weighted sum of the activity frequency and the activity amplitude, and determines the weighted sum as the part activity of the data blind area. The activity frequency and the activity amplitude can be determined by any feasible method such as image comparison and three-dimensional space point cloud change.

[0115] In some embodiments, the anti-collision system determines the at least one safety buffer based on the region feature and the region activity, and queries a region preset table. The region preset table is pre-set based on historical experience and includes a plurality of sets of region features and region activities, and an expansion distance of a safety buffer corresponding to each set of region features and region activities. The expansion distance of the safety buffer is positively correlated with the area of the data blind area and the region activity corresponding to the data blind area.

[0116] In some embodiments, the anti-collision system determines the expansion distance of the safety buffer corresponding to the data blind area by a buffer prediction model based on the region feature and the region activity.

[0117] The buffer prediction model refers to a model for determining the expansion distance of the safety buffer. In some embodiments, the buffer prediction model can be a machine learning model. For example, the buffer prediction model can include any one or combination of a convolutional neural network (CNN) model, a neural network (NN) model, or other custom model structures.

[0118] In some embodiments, the anti-collision system trains the buffer prediction model by gradient descent method based on a large number of second training samples with second labels. The second training samples include sample region features and sample region activities, and the second labels of the second training samples can be the actual expansion distance of the safety buffer corresponding to the second training samples.

[0119] In some embodiments, the second training samples and the second labels are determined based on historical data. For example, the anti-collision system counts historical radiotherapy processes in the historical data that have not occurred collision and have high radiotherapy efficiency, takes the region features and region activities of the historical data blind areas in the historical radiotherapy processes as the second training samples, and takes the expansion distance of the safety buffer actually used in the historical radiotherapy processes as the second labels. The radiotherapy efficiency can be represented by radiotherapy time, and the shorter the radiotherapy time, the higher the radiotherapy efficiency.

[0120] In some embodiments, the training process of the buffer prediction model is similar to the training process of the risk prediction model, and the implementation method can refer to the training process of the risk prediction model.

[0121] In some embodiments of the present specification, the higher the region activity, the higher the potential collision risk. Based on the region activity, the size of the safety buffer corresponding to the data blind area is determined in multiple ways, which helps to improve the efficiency of determining the safety buffer and further reduce the potential collision risk.

[0122] In some embodiments, the anti-collision system performs step 420 to obtain a second planning path 430 based on the real-time three-dimensional model 310 and the at least one safety buffer obtained in step 410.

[0123] Step 420, the first planning path is modified.

[0124] In some embodiments, the anti-collision system fuses the safety buffer and the real-time three-dimensional model by any feasible method such as point cloud union to obtain a fused three-dimensional model, and modifies the first planning path based on the fused three-dimensional model to obtain a second planning path. The method of modifying the first planning path based on the fused three-dimensional model by the anti-collision system is similar to the method of modifying the first planning path based on the real-time three-dimensional model, and the implementation method can be referred to Figure 2 and the related description.

[0125] In some embodiments of the present specification, by reserving a safety buffer, it is ensured that even if the patient has a slight accidental movement during radiotherapy, collision will not occur, and the safety of the radiotherapy process is improved.

[0126] Figure 5 is an exemplary flowchart for determining a first planning path according to some embodiments of the present specification. In some embodiments, the flowchart 500 is performed by a radiotherapy anti-collision system based on three-dimensional scanning and real-time simulation (hereinafter referred to as anti-collision system).

[0127] Step 510, determine a target path library.

[0128] The target path library refers to a database containing a plurality of planning paths. In some embodiments, the anti-collision system stores a plurality of target path libraries, each of which corresponds to a type of treatment need. The anti-collision system determines the corresponding type of target path library based on the treatment needs of the current patient. For a description of treatment needs, see Figure 2 and the related description.

[0129] In some embodiments, the construction process of the target path library includes: for each type of treatment need, the anti-collision system counts a large amount of historical data with the same or similar historical treatment needs as the treatment need, determines the historical radiotherapy process corresponding to such historical treatment needs as target historical data, determines the target historical data with better radiotherapy effect as preferred historical data, and adds the planning path used in the preferred historical data as a candidate planning path to the target path library. Wherein the radiotherapy effect is manually annotated by the user based on historical experience.

[0130] The collision avoidance system determines whether the treatment need and the historical treatment need are similar by converting into eigenvectors and calculating vector similarity, etc. The vector similarity is negatively related to the vector distance. The vector distance includes Euclidean distance, etc.

[0131] Step 520, at least one simulation is simulated based on the initial three-dimensional model and the target path library.

[0132] In some embodiments, the process of single simulation includes:

[0133] S1: The collision avoidance system selects a candidate planning path from the target path library as the target planning path of the current simulation based on a preset selection order. The preset selection order includes selecting the candidate planning path with the most historical use times from the candidate planning paths that have not been selected. The candidate planning path that has not been selected refers to the candidate planning path that has not been selected in the current at least one simulation. The historical use times are determined by the collision avoidance system based on historical data.

[0134] Illustratively, the current at least one simulation has been simulated once, which is recorded as the first simulation, and the current simulation is the second simulation. If the target path library includes 5 paths (which are recorded as path 1-5, and the historical use times of path 1-5 are sorted from high to low), the target planning path selected by the first simulation is path 1, and the target planning path selected by the second simulation is path 2. The candidate planning paths that have not been selected include paths 3-5.

[0135] S2: Based on the target planning path and the initial three-dimensional model, the test collision risk corresponding to the target planning path is determined. The test collision risk is used to represent the possibility of collision between the treatment device and the patient and / or support structure when the treatment device moves along the target planning path. The method of determining the test collision risk is similar to the method of determining the first collision risk, and the implementation method is described in Figure 2 and the related description.

[0136] S3: In response to the test collision risk not meeting the end condition, the next simulation is performed. The end condition includes that the test collision risk is not less than the risk threshold. For the description of the risk threshold, see Figure 2 and the related description.

[0137] S4: In response to the collision risk corresponding to the target planning path meeting the end condition, the at least one simulation is ended.

[0138] Step 530, the first planning path is determined based on the simulation result of the at least one simulation.

[0139] For more description of the first planning path, see Figure 2 and the related description.

[0140] In some embodiments, the anti-collision system determines the candidate planning path with the least test collision risk in the simulation results as the first planning path based on the simulation results of at least one simulation simulation. The simulation results include the test collision risk determined in each simulation simulation.

[0141] In some embodiments of the present specification, the optimal first planning path is comprehensively evaluated and determined in combination with the target path library of the treatment needs of the patient, improving the efficiency of determining the first planning path.

[0142] It should be noted that the above description of the processes 400 and 500 is only for example and illustration, and does not limit the scope of the present specification. Various modifications and changes can be made to the processes under the guidance of the present specification. However, these modifications and changes are still within the scope of the present specification.

[0143] Some embodiments of the present specification also provide a computer-readable storage medium, which stores computer instructions. When the computer reads the computer instructions in the storage medium, the computer executes the method described in any one of the above embodiments.

[0144] In addition, some features, structures or characteristics in one or more embodiments of the present specification can be appropriately combined.

[0145] Some embodiments use numbers to describe components, attributes, and the like. It should be understood that such numbers used in the description of the embodiments are, in some examples, modified by the adjectives "about", "approximately", or "generally". Unless otherwise stated, "about", "approximately", or "generally" indicates that the stated number allows for a ±20% variation. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximations that can vary depending on the desired characteristics of the individual embodiments. In some embodiments, numerical parameters should be considered in the context of the number of significant digits and by applying ordinary rounding techniques. Although the numerical ranges and parameters in some embodiments of the present specification are approximations, in specific embodiments, such numerical values are set forth in a manner to provide understanding of the breadth of the scope of the present specification.

[0146] If the description, definitions, and / or use of terms in the materials cited in the present specification are inconsistent or conflict with the description, definitions, and / or use of terms in the present specification, the description, definitions, and / or use of terms in the present specification shall prevail.

Claims

1. A collision avoidance method for radiotherapy based on three-dimensional scanning and real-time simulation, characterized in that, The method includes: Based on the patient's body surface data, an initial three-dimensional model is generated. The initial three-dimensional model includes at least the patient's three-dimensional information. The body surface data includes the patient's body shape, height, dimensions of various body parts, and posture. The body surface data is represented by a three-dimensional spatial point cloud. Based on the initial three-dimensional model, a collision avoidance simulation is performed during radiotherapy to determine the first planned path during the radiotherapy process; Based on the initial 3D model, the simulation of collision avoidance during radiotherapy is performed to determine the first planned path during the radiotherapy process, which includes: Determine the target path library; At least one simulation is performed based on the initial 3D model and the target path library. Each simulation process includes: A candidate planning path is selected from the target path library based on a preset selection order, and used as the target planning path for this simulation. Based on the target planning path and the initial 3D model, determine the test collision risk corresponding to the target planning path; If the test collision risk does not meet the termination condition, proceed to the next simulation. The at least one simulation ends when the collision risk corresponding to the target planned path meets the termination condition. The first planned path is determined based on the simulation results of the at least one simulation.

2. The method as described in claim 1, characterized in that, The method further includes: Acquire real-time scan data of the patient during the treatment process; Based on the real-time scanning data, the initial 3D model is updated to obtain a real-time 3D model; The simulation is performed based on the real-time 3D model.

3. The method as described in claim 2, characterized in that, The method further includes: Based on the real-time 3D model and the first planned path, a first collision risk is determined, where the first collision risk is the collision risk corresponding to the first future time. In response to the first collision risk meeting the warning conditions, a warning message is generated and a warning is issued.

4. The method as described in claim 3, characterized in that, The determination of the first collision risk based on the real-time 3D model and the first planned path includes: Based on the real-time 3D model and the first planned path, the first collision risk is determined by a risk prediction model, which is a machine learning model.

5. The method as described in claim 4, characterized in that, The inputs to the risk prediction model also include the number of data blind spots and the site activity corresponding to each data blind spot, wherein the site activity is the degree of activity of the patient's site in the data blind spot.

6. The method as described in claim 3, characterized in that, The first future time includes at least one future time point, the number of which is related to the motion characteristics of the treatment device.

7. The method as described in claim 2, characterized in that, The method further includes: Based on the real-time 3D model, the first planned path is corrected to obtain the second planned path; Based on the real-time 3D model and the second planned path, a second collision risk is determined. The second collision risk is the collision risk corresponding to a second future time, and the second future time is later than or equal to the first future time. In response to the second collision risk meeting the warning conditions, a warning message is generated and a warning is issued, and / or the second planned path is updated.

8. The method as described in claim 1, characterized in that, The method further includes: Determine at least one safe buffer; Based on the real-time 3D model and the at least one safety buffer, the first planned path is modified to obtain the second planned path.

9. The method as described in claim 8, characterized in that, Determining at least one security buffer includes: Identify data blind spots; Based on the regional characteristics of the data blind zone, the at least one security buffer is determined.

10. The method as described in claim 9, characterized in that, The regional features include the area of ​​the data blind zone, and determining the at least one security buffer based on the regional features of the data blind zone includes: The at least one security buffer is determined based on the area of ​​the region.

11. The method as described in claim 10, characterized in that, The determination of the at least one security buffer based on the regional characteristics of the data blind spot further includes: Based on the aforementioned regional characteristics, the activity level of the corresponding areas in the data blind spots is determined. The at least one safety buffer is determined based on the regional characteristics and the activity level of the location.

12. The method as described in claim 1, characterized in that, The method further includes: Obtain structural data of the support structure that supports the patient; The initial three-dimensional model is generated based on the body surface data and the structural data, and the initial three-dimensional model also includes the structural data.

13. A radiotherapy collision avoidance system based on three-dimensional scanning and real-time simulation, characterized in that, The system includes a model generation module and a simulation module; The model generation module is configured to generate an initial three-dimensional model based on the patient's body surface data. The initial three-dimensional model includes at least the patient's three-dimensional information. The body surface data includes the patient's body shape, height, dimensions of various body parts, and posture. The body surface data is represented by a three-dimensional spatial point cloud. The simulation module is configured to perform collision avoidance simulation during radiotherapy based on the initial three-dimensional model, and determine the first planned path during the radiotherapy process; Based on the initial 3D model, the simulation of collision avoidance during radiotherapy is performed to determine the first planned path during the radiotherapy process, which includes: Determine the target path library; At least one simulation is performed based on the initial 3D model and the target path library. Each simulation process includes: A candidate planning path is selected from the target path library based on a preset selection order, and used as the target planning path for this simulation. Based on the target planning path and the initial 3D model, determine the test collision risk corresponding to the target planning path; If the test collision risk does not meet the termination condition, proceed to the next simulation. The at least one simulation ends when the collision risk corresponding to the target planned path meets the termination condition. The first planned path is determined based on the simulation results of the at least one simulation.

14. A computer-readable storage medium storing computer instructions, wherein when a computer reads the computer instructions in the storage medium, the computer performs the method as described in any one of claims 1 to 12.

Images