Auxiliary positioning method for tumor intensity modulated radiation and related device

Abstract

The embodiment of the invention provides an auxiliary positioning method for tumor intensity modulated radiation and a related device, and belongs to the technical field of image processing. The method comprises the steps of capturing a target position image of a target mark point in real time according to a target sensor; performing position change identification on a target mark point in the target position image by adopting a deep learning algorithm to obtain a motion direction and a motion amplitude of the target mark point; performing risk early warning according to the motion direction and the motion amplitude to obtain a target early warning result; and determining a beam radiation area of radiotherapy equipment corresponding to the target patient according to the target early warning result, the movement direction and the movement amplitude, so that the radiotherapy equipment performs radiation beam positioning according to the beam radiation area. The problem that poor treatment effect exists around the sensitive part due to the influence of movement of a patient in the prior art is solved. And treatment interference caused by movement of a patient can be effectively reduced, so that the accuracy and effectiveness of treatment are improved. And the risk that normal tissues are damaged by unnecessary radiation is also obviously reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to an auxiliary positioning method for tumor intensity modulated radiation and a related device. Background Art

[0002] In traditional conformal radiotherapy, the beam intensity in each radiation field is uniform, resulting in the radiation dose shape being basically limited to simple geometric shapes (such as squares or rectangles), which cannot effectively adapt to the complex shape of the tumor. This method cannot cope with the tiny movements of patients caused by physiological factors such as breathing and heartbeat during radiotherapy, which may cause unnecessary radiation damage to normal tissues. The introduction of intensity-modulated radiotherapy (IMRT), rotational modulated radiotherapy (VMAT) and spiral tomotherapy has significantly improved the formability of the radiation dose, which can accurately fit the shape of the tumor and reduce the radiation dose to normal tissues. However, these treatment methods are still affected by patient movement, especially around sensitive areas (such as lymph nodes, joint epiphyses, etc.), which may lead to poor treatment effects.

[0003] Therefore, it is urgent to design a technical solution to solve at least one of the above technical problems. Summary of the invention

[0004] The main purpose of the embodiments of the present invention is to provide an auxiliary positioning method and related devices for tumor intensity modulated radiation, aiming to solve the problem in related technologies that adverse treatment effects occur around sensitive parts due to the influence of patient movement.

[0005] In a first aspect, an embodiment of the present invention provides an auxiliary positioning method for tumor intensity modulated radiation therapy, comprising:

[0006] According to the target sensor, a target position image corresponding to a target marking point of a target patient is captured in real time;

[0007] Using a deep learning algorithm to identify the position change of the target marking point in the target position image, and obtain the movement direction and movement amplitude corresponding to the target marking point;

[0008] Perform risk warning according to the movement direction and the movement amplitude to obtain a target warning result;

[0009] The beam irradiation area of ​​the radiotherapy equipment corresponding to the target patient is determined according to the target warning result, the movement direction and the movement amplitude, so that the radiotherapy equipment performs radiation beam positioning on the target patient according to the beam irradiation area.

[0010] In a second aspect, an embodiment of the present invention provides an auxiliary positioning system for tumor intensity modulated radiation therapy, comprising:

[0011] A data acquisition module, used for capturing a target position image corresponding to a target marking point of a target patient in real time according to a target sensor;

[0012] A change recognition module is used to use a deep learning algorithm to identify the position change of the target marking point in the target position image, and obtain the movement direction and movement amplitude corresponding to the target marking point;

[0013] An information warning module is used to perform risk warning according to the movement direction and the movement amplitude to obtain a target warning result;

[0014] The area determination module is used to determine the beam irradiation area of ​​the radiotherapy equipment corresponding to the target patient according to the target warning result, the movement direction and the movement amplitude, so that the radiotherapy equipment can perform radiation beam positioning on the target patient according to the beam irradiation area.

[0015] In a third aspect, an embodiment of the present invention further provides a terminal device, comprising a processor, a memory, a computer program stored in the memory and executable by the processor, and a data bus for realizing connection and communication between the processor and the memory, wherein when the computer program is executed by the processor, the steps of any one of the auxiliary positioning methods for tumor intensity-modulated radiation provided in the specification of the present invention are realized.

[0016] In a fourth aspect, an embodiment of the present invention further provides a storage medium for computer-readable storage, characterized in that the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any one of the auxiliary positioning methods for tumor intensity-modulated radiation provided in the specification of the present invention.

[0017] The embodiment of the present invention provides an auxiliary positioning method and related device for tumor intensity modulated radiation. The method can accurately identify the movement direction and movement amplitude of the target patient by capturing the target position image corresponding to the target marker point of the target patient in real time, and using a deep learning algorithm to identify the position change of the target marker point in the target position image, so as to perform risk warning according to the movement direction and movement amplitude, and obtain the target warning result; determine the beam radiation area of ​​the radiotherapy equipment corresponding to the target patient according to the target warning result, movement direction and movement amplitude, so as to ensure that the radiotherapy equipment can be adjusted in time and accurately locate the beam radiation area. This will significantly improve the targeting of radiotherapy and minimize damage to normal tissues. It solves the problem of adverse treatment effects around sensitive parts caused by patient movement in related technologies. It can effectively reduce the treatment interference caused by patient movement, thereby improving the accuracy and effectiveness of treatment. It also significantly reduces the risk of unnecessary radiation damage to normal tissues. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 A schematic flow chart of an auxiliary positioning method for tumor intensity modulated radiation provided by an embodiment of the present invention;

[0020] Figure 2 A schematic diagram of the module structure of an auxiliary positioning system for tumor intensity modulated radiation provided by an embodiment of the present invention;

[0021] Figure 3 A schematic block diagram of the structure of a terminal device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0024] It should be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0025] The embodiment of the present invention provides an auxiliary positioning method for tumor intensity modulated radiation and a related device. The auxiliary positioning method for tumor intensity modulated radiation can be applied to a terminal device, which can be an electronic device such as a tablet computer, a laptop computer, a desktop computer, a personal digital assistant, and a wearable device. The terminal device can be a server or a server cluster.

[0026] In traditional conformal radiotherapy, the beam intensity in each radiation field is uniform, resulting in the radiation dose shape being basically limited to simple geometric shapes (such as squares or rectangles), which cannot effectively adapt to the complex shape of the tumor. This method cannot cope with the tiny movements of patients caused by physiological factors such as breathing and heartbeat during radiotherapy, which may cause unnecessary radiation damage to normal tissues. The introduction of intensity-modulated radiotherapy (IMRT), rotational modulated radiotherapy (VMAT) and spiral tomotherapy has significantly improved the formability of the radiation dose, which can accurately fit the shape of the tumor and reduce the radiation dose to normal tissues. However, these treatment methods are still affected by patient movement, especially around sensitive areas (such as lymph nodes, joint epiphyses, etc.), which may lead to poor treatment effects.

[0027] To this end, an embodiment of the present invention provides an auxiliary positioning method and related devices for tumor intensity modulated radiation. The method can accurately identify the movement direction and movement amplitude of the target patient by capturing the target position image corresponding to the target marker point of the target patient in real time, and using a deep learning algorithm to identify the position change of the target marker point in the target position image, so as to perform risk warning according to the movement direction and movement amplitude, and obtain a target warning result; determine the beam radiation area of ​​the radiotherapy equipment corresponding to the target patient according to the target warning result, movement direction and movement amplitude, so as to ensure that the radiotherapy equipment can be adjusted in time and accurately locate the beam radiation area. This will significantly improve the targeting of radiotherapy and minimize damage to normal tissues. It solves the problem of adverse treatment effects around sensitive parts caused by patient movement in related technologies. It can effectively reduce the treatment interference caused by patient movement, thereby improving the accuracy and effectiveness of treatment. It also significantly reduces the risk of unnecessary radiation damage to normal tissues.

[0028] Some embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0029] Please refer to Figure 1 , Figure 1 A schematic flow chart of a method for auxiliary positioning of tumor intensity modulated radiation provided by an embodiment of the present invention.

[0030] like Figure 1 As shown, the auxiliary positioning method for tumor intensity modulated radiation therapy includes steps S101 to S104.

[0031] Step S101 : capturing a target position image corresponding to a target marking point of a target patient in real time according to a target sensor.

[0032] For example, a target marker point is affixed to the skin surface of a target patient who is undergoing tumor treatment using intensity modulated radiation therapy (IMRT), rotational intensity modulated radiation therapy (VMAT) or spiral tomotherapy. The target marker point can accurately identify the area of ​​the target patient that requires intensity modulated radiation therapy, thereby monitoring the target marker point to achieve the purpose of identifying tiny movements of the target patient.

[0033] For example, before treatment, a specially made high-contrast target marker is pasted on the skin surface of the target patient according to the tumor location and anatomical structure characteristics of the target patient. The design of the target marker can ensure clear identification in the machine vision system. Thus, the target position image corresponding to the target marker is captured in real time according to the target sensor. The target sensor can be a high-definition camera or an image processing unit, etc.

[0034] Step S102: Use a deep learning algorithm to identify position changes of the target marking point in the target position image to obtain the movement direction and movement amplitude corresponding to the target marking point.

[0035] Exemplarily, the initial position corresponding to the target marker point is determined, and then the target marker point in the target position image is recognized using a deep learning algorithm to obtain the position information of the target marker point in the target position image, and then the target position corresponding to the target marker point is determined based on the position information, thereby obtaining the movement direction corresponding to the target marker point based on the coordinate difference between the initial position and the target position, and obtaining the movement amplitude by calculating the Euclidean distance between the initial position and the target position.

[0036] In some embodiments, the use of a deep learning algorithm to identify position changes of the target marker points in the target position image to obtain a motion direction and a motion amplitude corresponding to the target marker points includes: using a target recognition model of the deep learning algorithm to perform target recognition on the target marker points in the target position image to obtain target position information corresponding to the target marker points in the target position image; using a target fitting model of the deep learning algorithm to perform information fitting on the target position information to obtain target edge information corresponding to the target marker points; determining a marking area corresponding to the target marker points based on the target edge information, and determining the motion direction and the motion amplitude corresponding to the target marker points based on the marking area.

[0037] Exemplarily, the target recognition model is YOLO (You Only Look Once) or FasterR-CNN, and the target recognition model is trained using labeled image data to ensure that the target recognition model can accurately identify the target marker point and extract its position. Cross-validation and different loss functions are used to evaluate and optimize model performance. Thus, the trained target recognition model is used to infer the target position image to obtain the target position information of the target marker point in the target position image, and the target position information may be bounding box information.

[0038] Exemplarily, based on the target position information extracted by the target recognition model, a target fitting model (such as curve fitting, contour recognition, etc.) is used to further analyze the shape and edge information of the target marker point. The target position information is input into the target fitting model, and the target edge information of the target marker point is obtained through the fitting process. For example, an edge detection algorithm (such as Canny edge detection) or a shape analysis technique is used to determine the contour of the target marker point, thereby obtaining the target edge information of the target marker point.

[0039] Exemplarily, the target center of the target marking point is determined according to the extracted target edge information, and the target centers of the target marking points are sequentially connected to obtain the marking area.

[0040] Exemplarily, based on the comparison between the marked area at the current moment and the marked area at the previous moment, the movement direction and the movement amplitude are determined by calculating the coordinate change of the center point corresponding to the marked area.

[0041] For example, the first center coordinate corresponding to the marked area at the current moment and the second center coordinate corresponding to the marked area at the previous moment are obtained, and then the coordinate difference between the first center coordinate and the second center coordinate is calculated using a vector to obtain the movement direction, and the movement amplitude is calculated using the length of the coordinate difference.

[0042] Specifically, through the target recognition model of the deep learning algorithm, the target marker points in the target position image can be identified with high precision, and their target position information can be accurately obtained. This accuracy is crucial for subsequent motion analysis and treatment positioning. By fitting the target position information using the target fitting model, the target edge information of the target marker points can be further refined. This helps to more accurately define the contour and shape of the target marker points, thereby improving the accuracy of motion analysis. Based on accurate target edge information, the marked area corresponding to the target marker point can be more reliably determined. This avoids the problem of inaccurate marked areas caused by edge blur or recognition errors, and provides a reliable basis for subsequent calculations of motion direction and amplitude. By comparing the marked areas at different time points, the motion direction and amplitude of the target marker points can be accurately calculated. This provides more powerful information support for the subsequent medical treatment process.

[0043] In some embodiments, the target recognition model of the deep learning algorithm is used to perform target recognition on the target marker points in the target position image to obtain the target position information corresponding to the target marker points in the target position image, including: obtaining the first objective function and the second objective function required for image segmentation of the target position image according to the target determination layer of the target recognition model; calculating the first segmentation threshold corresponding to the target position image under the first objective function using a particle swarm optimization algorithm according to the threshold calculation layer of the target recognition model; calculating the second segmentation threshold corresponding to the target position image under the second objective function using the particle swarm optimization algorithm and the first segmentation threshold according to the threshold calculation layer; performing optimal threshold analysis using the first segmentation threshold and the second segmentation threshold according to the threshold determination layer of the target recognition model to obtain the target segmentation threshold corresponding to the target position image; and performing target recognition on the target marker points in the target position image according to the target segmentation threshold to obtain the target position information corresponding to the target marker points in the target position image.

[0044] Exemplarily, the first objective function and the second objective function required for image segmentation of the target position image are obtained according to the target determination layer of the target recognition model. The first objective function and the second objective function should respectively reflect different optimization goals for image segmentation of the target position image, such as segmentation accuracy, edge smoothness, etc. For example, the first objective function is used to measure the maximum cross entropy information of image segmentation. The second objective function is used to consider the accuracy of edge detection or the smoothness of the segmented area.

[0045] Exemplarily, according to the threshold calculation layer of the target recognition model, a particle swarm optimization algorithm is used to iteratively update the position and velocity of particles, and an optimal solution is found under the first objective function to calculate the first segmentation threshold corresponding to the target position image.

[0046] Exemplarily, based on the first segmentation threshold, the particle swarm optimization algorithm is used again to perform optimization under the second objective function to calculate the second segmentation threshold corresponding to the target position image.

[0047] Exemplarily, according to the threshold determination layer of the target recognition model, a first segmented image of the target position image at a first segmentation threshold and a second segmented image of the target position image at a second segmentation threshold are obtained, a first segmented area corresponding to the target marker point in the first segmented image is similarly calculated to obtain a first similarity value, and a second segmented area corresponding to the target marker point in the second segmented image is similarly calculated to obtain a second similarity value, thereby determining the segmentation threshold corresponding to the maximum of the first similarity value and the second similarity value as the target segmentation threshold. The preset segmentation area is the area obtained by manually segmenting the target image collected by the target marker point under the target sensor.

[0048] Exemplarily, target recognition is performed on target marking points in the target position image according to the target segmentation threshold value to obtain target position information corresponding to the target marking points in the target position image.

[0049] Specifically, by defining the first objective function and the second objective function, multi-objective optimization is performed during the image segmentation process, thereby improving the accuracy and reliability of the segmentation results. By segmenting the target position image with the optimal segmentation threshold, the target marker point can be accurately identified and located, and more reliable target position information of the target marker point can be obtained, thereby providing accurate data support for subsequent motion analysis.

[0050] In some embodiments, the target fitting model of the deep learning algorithm is used to perform information fitting on the target position information to obtain the target edge information corresponding to the target marker point, including: obtaining the image radius corresponding to the target marker point according to the conversion relationship between the shooting position corresponding to the target position image and the real radius corresponding to the target marker point according to the radius determination layer of the target fitting model; performing position clustering according to the image radius and the target position information according to the data clustering layer of the target fitting model to obtain a position clustering result; performing information fitting on the position clustering result according to the target fitting layer of the target fitting model to obtain an initial fitting center and an initial fitting radius corresponding to the target marker point; performing information screening on the position clustering result according to the target determination layer of the target fitting model using the initial fitting radius and the image radius to obtain a target fitting radius; obtaining a target fitting center corresponding to the target fitting radius from the initial fitting center; and obtaining the target edge information corresponding to the target marker point according to the target fitting center and the target fitting radius.

[0051] Exemplarily, the target fitting model includes a radius determination layer, a data clustering layer, a target fitting layer, and a target determination layer.

[0052] Exemplarily, the target marking point is a circular mark, and the real radius of the target marking point is determined by manual measurement, so as to obtain the shooting parameters of the target position image, including the shooting position (i.e., the position and angle of the target sensor) and the properties of the target sensor (such as focal length, sensor size, aperture, etc.). In this way, the conversion relationship between the image coordinate system (in pixels) and the real world coordinate system (in physical units, such as meters or centimeters) is clarified. Based on the shooting position and the real radius of the target marking point, a simple geometric relationship (such as the principle of similar triangles) is used to calculate the image radius corresponding to the target marking point in the target position image.

[0053] Exemplarily, the data clustering layer of the target fitting model performs edge recognition on the target position information in the target position image to obtain the target edge points of the target mark points in the target position image. Thus, the data clustering layer performs position clustering using the image radius and the target edge points to obtain the position clustering result. The purpose of the data clustering layer performing position clustering is to group targets with target edge points of similar image radius.

[0054] For example, the coordinates of the target edge points and the image radius are used as the key features of clustering. According to the characteristics of the target edge points (such as data scale, shape complexity, noise level, etc.), select a suitable clustering algorithm, such as K-means, DBSCAN, spectral clustering, etc. Set appropriate parameters for the selected algorithm, such as the number of clusters K (for K-means), neighborhood radius, and minimum number of points (for DBSCAN). Input the coordinates of the target edge points and the image radius into the selected clustering algorithm, and then group similar edge points together. Thus, the position clustering result is obtained.

[0055] Exemplarily, if the target marker point is a circle, each cluster in the position clustering result is analyzed according to the target fitting layer of the target fitting model, and a suitable circle fitting algorithm is selected. Commonly used methods include least squares circle fitting, Hough transform circle detection, etc. After determining the circle fitting algorithm, each cluster is traversed, and the point coordinates in the cluster are used as input data to apply the selected circle fitting algorithm, and each cluster is fitted, and then the initial fitting center and initial fitting radius of the best fitting circle are calculated.

[0056] Exemplarily, according to the target determination layer of the target fitting model, the gap between the initial fitting radius and the image radius is used to screen information on the position clustering results, and the initial fitting radius at which the gap between the initial fitting radius and the image radius is the smallest is determined as the target fitting radius; and the target fitting center corresponding to the target fitting radius is obtained from the initial fitting center, and then the target fitting circle corresponding to the target marking point is determined according to the target fitting center and the target fitting radius, and then the target fitting circle is determined as the target edge information corresponding to the target marking point in the target position image.

[0057] In some embodiments, the target fitting layer according to the target fitting model performs information fitting on the position clustering result to obtain the initial fitting center and the initial fitting radius corresponding to the target marker point, including: determining the target optimization function and the target parameter corresponding to each cluster in the position clustering result according to the target fitting layer; performing partial derivative calculation on the target parameter in the target optimization function according to the target fitting layer to obtain the target partial derivative function corresponding to the target parameter; determining the solution formula corresponding to the target parameter by fusing the target partial derivative function according to the target fitting layer; using the solution formula and the initial fitting center and the initial fitting radius corresponding to each cluster in the position clustering result according to the target fitting layer; wherein the target optimization function and the target parameter are obtained according to the following formula:

[0058]

[0059] a=-2*x0;

[0060] b = -2*y0;

[0061] c=x0 2 +y0 2 -r 2 ;

[0062] Wherein, LOSS represents the target optimization function, a, b, c all represent the target parameters, M represents the number of targets corresponding to each cluster in the location clustering result, and x i Indicates the horizontal coordinate corresponding to each location information of each cluster in the location clustering result, y i The vertical coordinate corresponding to each location information of each cluster in the location clustering result, x0 represents the horizontal coordinate corresponding to the initial fitting center of each cluster in the location clustering result, y0 represents the vertical coordinate corresponding to the initial fitting center of each cluster in the location clustering result, and r represents the initial fitting radius corresponding to each cluster in the location clustering result.

[0063] Exemplarily, according to the target fitting layer, the sum of squares of distances from each cluster to the fitting center corresponding to the fitting circle in the fitting position clustering result using the least squares method is minimized, thereby avoiding the square root problem while ensuring the minimum sum of squares of distances, thereby resetting the target optimization function and target parameters, as shown in the following formula:

[0064]

[0065] a=-2*x0;

[0066] b = -2*y0;

[0067] c=x0 2 +y0 2 -r 2 ;

[0068] Among them, LOSS represents the target optimization function, a, b, and c represent the target parameters, M represents the number of targets corresponding to each cluster in the location clustering result, and x i Describes the horizontal coordinates corresponding to each location information of each cluster in the location clustering result, y i The vertical coordinate corresponding to each location information of each cluster in the location clustering result, x0 represents the horizontal coordinate corresponding to the initial fitting center of each cluster in the location clustering result, y0 represents the vertical coordinate corresponding to the initial fitting center of each cluster in the location clustering result, and r represents the initial fitting radius corresponding to each cluster in the location clustering result.

[0069] Exemplarily, the present application transforms the target optimization function for solving the initial fitting center and the initial fitting radius into a solution method for target parameter a, target parameter b, and target parameter c, so that after obtaining the target parameters, the initial fitting center and the initial fitting radius are obtained according to the following formula:

[0070] a=-2*x0;

[0071] b = -2*y0;

[0072] c=x0 2 +y0 2 -r 2 ;

[0073] Exemplarily, a partial derivative of a target parameter a in a target optimization function is calculated according to the target fitting layer to obtain a first partial derivative function corresponding to the target parameter a. A partial derivative of a target parameter b in a target optimization function is calculated according to the target fitting layer to obtain a second partial derivative function corresponding to the target parameter b. A partial derivative of a target parameter c in a target optimization function is calculated according to the target fitting layer to obtain a third partial derivative function corresponding to the target parameter c. Thus, a target partial derivative function is obtained according to the first partial derivative function, the second partial derivative function, and the third partial derivative function.

[0074] Exemplarily, according to the target fitting layer, the first partial derivative function, the second partial derivative function and the third partial derivative function are respectively equal to 0. Then, after setting the first partial derivative function, the second partial derivative function and the third partial derivative function to 0, two equations can be obtained: the sum of the horizontal coordinates corresponding to each position information of each cluster under the first partial derivative function minus the target number M multiplied by the third partial derivative function multiplied by the target number M is equal to 0, and the sum of the vertical coordinates corresponding to each position information of each cluster under the second partial derivative function minus the target number M multiplied by the third partial derivative function multiplied by the target number M is equal to 0, thereby fusing the above two equations to obtain the solution formula corresponding to the target parameters.

[0075] Exemplarily, according to the target fitting layer, the position information corresponding to each cluster in the position clustering result is substituted into the solution formula to obtain the initial fitting center and initial fitting radius corresponding to the cluster.

[0076] In some embodiments, the marking area includes a first area and a second area, and determining the movement direction and the movement amplitude corresponding to the target marking point based on the marking area includes: determining a target coordinate system, and determining a first coordinate system corresponding to the first area and a second coordinate system corresponding to the second area; calculating a first transformation relationship corresponding to the first coordinate system and the target coordinate system, and calculating a second transformation relationship corresponding to the second coordinate system and the target coordinate system; converting the first area to the target coordinate system according to the first transformation relationship to obtain a first target area; converting the second area to the target coordinate system according to the second transformation relationship to obtain a second target area; edge-aligning the first target area and the second target area to obtain a target alignment result; and determining the movement direction and the movement amplitude corresponding to the target marking point according to the target alignment result.

[0077] Exemplarily, the marking area includes a first area corresponding to the target marking point at the current moment and a second area corresponding to the previous moment before the current moment, thereby determining the target coordinate system, that is, the world coordinate system, determining the first coordinate system corresponding to the first area, that is, the first image coordinate system corresponding to the target position image at the current moment, and determining the second coordinate system corresponding to the second area, that is, the second image coordinate system corresponding to the target position image at the previous moment. Thus, the first conversion relationship between the first coordinate system and the target coordinate system and the second conversion relationship between the second coordinate system and the target coordinate system are calculated.

[0078] For example, at the current moment, a first image coordinate system is established according to the viewing angle and parameters of the target position image corresponding to the current moment. This includes determining the origin, coordinate axis direction and unit of the target position image corresponding to the current moment, and associating it with the world coordinate system. And obtaining the origin, coordinate axis direction and unit of the target position image corresponding to the target position at the previous moment and associating it with the world coordinate system. Thus, by analyzing the relationship between the first image coordinate system and the target coordinate system, the parameters required for the conversion are determined. These parameters include translation vectors, rotation matrices, etc. According to the above parameters, a first conversion relationship from the first coordinate system to the target coordinate system is established, including: translation, rotation, so that according to the translation and rotation relationship, the first conversion relationship from the first coordinate system to the target coordinate system is expressed. Similarly, the relationship between the second image coordinate system and the target coordinate system is analyzed to identify the required conversion parameters. According to the translation and rotation parameters of the second coordinate system, a second conversion relationship from the second coordinate system to the target coordinate system is established.

[0079] Exemplarily, the first area is converted into the target coordinate system according to the first conversion relationship to obtain the first target area, and the second area is converted into the target coordinate system according to the second conversion relationship to obtain the second target area, so that the first target area and the second target area are edge-aligned according to the requirement when the overlapping area is the largest, to obtain a target alignment result.

[0080] For example, select initial alignment parameters, such as initial values ​​of translation, rotation, and scaling. Set the alignment goal, that is, maximize the overlap area of ​​the first target area and the second target area, and then use an optimization algorithm (such as gradient descent, genetic algorithm, etc.) to adjust the alignment parameters to find the target alignment parameters with the largest overlap area, so as to obtain the target alignment result according to the target alignment parameters.

[0081] Exemplarily, the movement direction corresponding to the target marker point is determined according to the rotation angle in the target alignment result, and the movement amplitude is determined according to the translation vector in the target alignment result.

[0082] Step S103: Perform risk warning according to the movement direction and the movement amplitude to obtain a target warning result.

[0083] Exemplarily, the direction of movement is compared with a pre-set dangerous direction. For example, certain directions may lead to higher risks (such as approaching a border). Then evaluate whether the amplitude of movement exceeds the normal range. A larger amplitude may mean rapid movement, increasing potential risks. Thus, an early warning mechanism is set according to the set rules, that is, an early warning is triggered when certain conditions are met, for example: the amplitude of movement exceeds the set threshold, or the direction of movement approaches the border. The risks are thus divided into different levels, such as low risk: within the normal range, no alarm is required. Medium risk: monitoring is required, and measures may need to be taken. High risk: immediate alarm is required and emergency response is implemented. Finally, the risk level determined according to the direction of movement and the amplitude of movement is determined as the target early warning result.

[0084] In some embodiments, performing risk warning based on the movement direction and the movement amplitude to obtain a target warning result includes: determining the amplitude range corresponding to the radiotherapy equipment in the movement direction; comparing the amplitude range and the movement amplitude to obtain a comparison result; performing risk warning based on the comparison result to obtain the target warning result.

[0085] For example, during radiotherapy, it is crucial to ensure that the target patient's movement amplitude and direction are within a safe range. First, a series of expert consultations are conducted, and the opinions of experts in related fields (such as radiologists, engineers, and physical therapists) are used to set the corresponding amplitude range under different movement directions, and then compare the amplitude range and the movement amplitude to obtain a comparison result. If the monitored movement amplitude is within the preset amplitude range, the target warning result is determined to be low risk and no alarm is required. If the movement amplitude is close but still within the amplitude range, the target warning result is determined to be medium risk. In this case, it is necessary to prepare to take preventive measures. If the movement amplitude exceeds the preset amplitude range, the target warning result is determined to be high risk and corresponding measures need to be taken.

[0086] Step S104, determining the beam irradiation area of ​​the radiotherapy equipment corresponding to the target patient according to the target warning result, the movement direction and the movement amplitude, so that the radiotherapy equipment performs radiation beam positioning for the target patient according to the beam irradiation area.

[0087] Exemplarily, when the target warning result is low risk, the operation of the radiotherapy equipment when performing beam radiation on the target patient does not need to be changed; when the target warning result is medium risk or high risk, it is necessary to adjust the current radiation area according to the movement direction and movement amplitude corresponding to the target patient to obtain the beam radiation area of ​​the radiotherapy equipment corresponding to the target patient, so that the radiotherapy equipment can position the radiation beam for the target patient according to the beam radiation area.

[0088] For example, the initial radiation area of ​​the radiotherapy device is set according to the location and size of the tumor. The initial beam radiation area is adjusted according to the direction and amplitude of movement of the target patient. For example, if the target patient moves in a certain direction, the position of the initial radiation area is adjusted accordingly. If the target patient rotates, the shape of the initial radiation area is adjusted to cover the new position of the tumor, thereby obtaining the beam radiation area. The beam radiation area is redefined as the initial radiation area, so that dynamic monitoring is implemented, the latest movement data of the target patient is continuously obtained, and the beam radiation area is continuously updated. Ensure that the beam radiation area always accurately covers the tumor position, while also avoiding excessive irradiation of surrounding healthy tissues.

[0089] For example, during radiotherapy, the radiation area of ​​the radiotherapy equipment is adjusted in real time according to the movement direction and movement amplitude of the target patient to ensure the accurate positioning of the radiation beam and the effective implementation of the treatment. This method can not only improve the accuracy of radiotherapy, but also maximize the protection of surrounding healthy tissues and improve the treatment effect and safety of the target patient.

[0090] The embodiment of the present invention provides an auxiliary positioning method and related device for tumor intensity modulated radiation. The method can accurately identify the movement direction and movement amplitude of the target patient by capturing the target position image corresponding to the target marker point of the target patient in real time, and using a deep learning algorithm to identify the position change of the target marker point in the target position image, so as to perform risk warning according to the movement direction and movement amplitude, and obtain the target warning result; determine the beam radiation area of ​​the radiotherapy equipment corresponding to the target patient according to the target warning result, movement direction and movement amplitude, so as to ensure that the radiotherapy equipment can be adjusted in time and accurately locate the beam radiation area. This will significantly improve the targeting of radiotherapy and minimize damage to normal tissues. It solves the problem of adverse treatment effects around sensitive parts caused by patient movement in related technologies. It can effectively reduce the treatment interference caused by patient movement, thereby improving the accuracy and effectiveness of treatment. It also significantly reduces the risk of unnecessary radiation damage to normal tissues.

[0091] See also Figure 2 , Figure 2 An auxiliary positioning system 200 for tumor intensity modulated radiation provided in an embodiment of the present application includes a data acquisition module 201, a change recognition module 202, an information warning module 203, and an area determination module 204, wherein the data acquisition module 201 is used to capture a target position image corresponding to a target marker point of a target patient in real time according to a target sensor; the change recognition module 202 is used to use a deep learning algorithm to identify position changes of the target marker point in the target position image to obtain a movement direction and a movement amplitude corresponding to the target marker point; the information warning module 203 is used to perform a risk warning according to the movement direction and the movement amplitude to obtain a target warning result; the area determination module 204 is used to determine the beam radiation area of ​​the radiotherapy equipment corresponding to the target patient according to the target warning result, the movement direction and the movement amplitude, so that the radiotherapy equipment performs radiation beam positioning on the target patient according to the beam radiation area.

[0092] In some implementations, the change recognition module 202 performs, during the process of using the deep learning algorithm to recognize the position change of the target marker point in the target position image and obtaining the movement direction and movement amplitude corresponding to the target marker point:

[0093] Using the target recognition model of the deep learning algorithm to perform target recognition on the target marking point in the target position image, and obtain target position information corresponding to the target marking point in the target position image;

[0094] Using the target fitting model of the deep learning algorithm to perform information fitting on the target position information, the target edge information corresponding to the target marking point is obtained;

[0095] The marking area corresponding to the target marking point is determined according to the target edge information, and the movement direction and the movement amplitude corresponding to the target marking point are determined according to the marking area.

[0096] In some implementations, the change recognition module 202, in the process of using the target recognition model of the deep learning algorithm to perform target recognition on the target marking point in the target position image and obtaining the target position information corresponding to the target marking point in the target position image, executes:

[0097] Obtaining a first objective function and a second objective function required for image segmentation of the target position image according to the target determination layer of the target recognition model;

[0098] Calculating a first segmentation threshold corresponding to the target position image under the first objective function using a particle swarm optimization algorithm according to the threshold calculation layer of the target recognition model;

[0099] Calculate the second segmentation threshold corresponding to the target position image under the second objective function by using the particle swarm optimization algorithm and the first segmentation threshold according to the threshold calculation layer;

[0100] Performing an optimal threshold analysis using the first segmentation threshold and the second segmentation threshold according to the threshold determination layer of the target recognition model to obtain a target segmentation threshold corresponding to the target position image;

[0101] The target marking point in the target position image is subjected to target recognition according to the target segmentation threshold value to obtain the target position information corresponding to the target marking point in the target position image.

[0102] In some implementations, the change recognition module 202 performs, during the process of performing information fitting on the target position information using the target fitting model of the deep learning algorithm to obtain the target edge information corresponding to the target marking point:

[0103] According to the radius determination layer of the target fitting model, the image radius corresponding to the target marking point is obtained by utilizing the conversion relationship between the shooting position corresponding to the target position image and the real radius corresponding to the target marking point;

[0104] Performing position clustering using the image radius and the target position information according to the data clustering layer of the target fitting model to obtain a position clustering result;

[0105] Performing information fitting on the position clustering result according to the target fitting layer of the target fitting model to obtain an initial fitting center and an initial fitting radius corresponding to the target mark point;

[0106] According to the target determination layer of the target fitting model, the initial fitting radius and the image radius are used to perform information screening on the position clustering result to obtain a target fitting radius;

[0107] Obtaining a target fitting center corresponding to the target fitting radius from the initial fitting center;

[0108] The target edge information corresponding to the target marking point is obtained according to the target fitting center and the target fitting radius.

[0109] In some implementations, the change identification module 202 performs, during the process of performing information fitting on the position clustering result according to the target fitting layer of the target fitting model to obtain the initial fitting center and the initial fitting radius corresponding to the target marker point:

[0110] Determining the target optimization function and target parameters corresponding to each cluster in the position clustering result according to the target fitting layer;

[0111] Performing partial derivative calculation on the target parameter in the target optimization function according to the target fitting layer to obtain a target partial derivative function corresponding to the target parameter;

[0112] Determine a solution formula corresponding to the target parameter by fusing the target partial derivative function with the target fitting layer;

[0113] According to the target fitting layer, the initial fitting center and the initial fitting radius corresponding to each cluster in the position clustering result are used by the solution formula;

[0114] Wherein, the objective optimization function and the objective parameters are obtained according to the following formula:

[0115]

[0116] a=-2*x0;

[0117] b = -2*y0;

[0118] c=x0 2 +y0 2 -r 2 ;

[0119] Wherein, LOSS represents the target optimization function, a, b, c all represent the target parameters, M represents the number of targets corresponding to each cluster in the location clustering result, and x iIndicates the horizontal coordinate corresponding to each location information of each cluster in the location clustering result, y i The vertical coordinate corresponding to each location information of each cluster in the location clustering result, x0 represents the horizontal coordinate corresponding to the initial fitting center of each cluster in the location clustering result, y0 represents the vertical coordinate corresponding to the initial fitting center of each cluster in the location clustering result, and r represents the initial fitting radius corresponding to each cluster in the location clustering result.

[0120] In some implementations, the marked area includes a first area and a second area, and the change recognition module 202, in the process of determining the movement direction and the movement amplitude corresponding to the target marked point according to the marked area, performs:

[0121] Determine a target coordinate system, and determine a first coordinate system corresponding to the first area and a second coordinate system corresponding to the second area;

[0122] Calculating a first transformation relationship between the first coordinate system and the target coordinate system and calculating a second transformation relationship between the second coordinate system and the target coordinate system;

[0123] According to the first conversion relationship, the first area is converted into the target coordinate system to obtain a first target area;

[0124] According to the second conversion relationship, the second area is converted into the target coordinate system to obtain a second target area;

[0125] Performing edge alignment on the first target area and the second target area to obtain a target alignment result;

[0126] The movement direction and the movement amplitude corresponding to the target marking point are determined according to the target alignment result.

[0127] In some implementations, the information warning module 203 performs, during the process of performing risk warning according to the movement direction and the movement amplitude and obtaining a target warning result:

[0128] Determining an amplitude range corresponding to the radiotherapy device in the movement direction;

[0129] Compare the amplitude range and the motion amplitude to obtain a comparison result;

[0130] A risk warning is performed according to the comparison result to obtain the target warning result.

[0131] In some embodiments, the auxiliary positioning system 200 for tumor intensity modulated radiation therapy can be applied to a terminal device.

[0132] It should be noted that, those skilled in the art can clearly understand that, for the sake of convenience and brevity of description, the specific working process of the auxiliary positioning system 200 for tumor intensity-modulated radiation therapy described above can refer to the corresponding process in the aforementioned embodiment of the auxiliary positioning method for tumor intensity-modulated radiation therapy, and will not be repeated here.

[0133] See also Figure 3 , Figure 3 A schematic block diagram of the structure of a terminal device provided in an embodiment of the present invention.

[0134] like Figure 3 As shown, the terminal device 300 includes a processor 301 and a memory 302, and the processor 301 and the memory 302 are connected via a bus 303, such as an I2C (Inter-integrated Circuit) bus.

[0135] Specifically, the processor 301 is used to provide computing and control capabilities to support the operation of the entire terminal device. The processor 301 can be a central processing unit (CPU), and the processor 301 can also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0136] Specifically, the memory 302 may be a Flash chip, a read-only memory (ROM) disk, an optical disk, a USB flash drive, or a mobile hard disk.

[0137] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a partial structure related to the embodiment of the present invention, and does not constitute a limitation on the terminal device to which the embodiment of the present invention is applied. The specific server may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0138] The processor is used to run a computer program stored in the memory, and implement any one of the auxiliary positioning methods for tumor intensity-modulated radiation provided by the embodiments of the present invention when executing the computer program.

[0139] In one embodiment, the processor is used to run a computer program stored in the memory, and implements the following steps when executing the computer program:

[0140] According to the target sensor, a target position image corresponding to a target marking point of a target patient is captured in real time;

[0141] Using a deep learning algorithm to identify the position change of the target marking point in the target position image, and obtain the movement direction and movement amplitude corresponding to the target marking point;

[0142] Perform risk warning according to the movement direction and the movement amplitude to obtain a target warning result;

[0143] The beam irradiation area of ​​the radiotherapy equipment corresponding to the target patient is determined according to the target warning result, the movement direction and the movement amplitude, so that the radiotherapy equipment performs radiation beam positioning on the target patient according to the beam irradiation area.

[0144] In some implementations, the processor 301, in the process of using the deep learning algorithm to identify the position change of the target marker point in the target position image and obtaining the movement direction and movement amplitude corresponding to the target marker point, executes:

[0145] Using the target recognition model of the deep learning algorithm to perform target recognition on the target marking point in the target position image, and obtain target position information corresponding to the target marking point in the target position image;

[0146] Using the target fitting model of the deep learning algorithm to perform information fitting on the target position information, the target edge information corresponding to the target marking point is obtained;

[0147] The marking area corresponding to the target marking point is determined according to the target edge information, and the movement direction and the movement amplitude corresponding to the target marking point are determined according to the marking area.

[0148] In some implementations, the processor 301, in the process of using the target recognition model of the deep learning algorithm to perform target recognition on the target marking point in the target position image and obtaining the target position information corresponding to the target marking point in the target position image, executes:

[0149] Obtaining a first objective function and a second objective function required for image segmentation of the target position image according to the target determination layer of the target recognition model;

[0150] Calculating a first segmentation threshold corresponding to the target position image under the first objective function using a particle swarm optimization algorithm according to the threshold calculation layer of the target recognition model;

[0151] Calculate the second segmentation threshold corresponding to the target position image under the second objective function by using the particle swarm optimization algorithm and the first segmentation threshold according to the threshold calculation layer;

[0152] Performing an optimal threshold analysis using the first segmentation threshold and the second segmentation threshold according to the threshold determination layer of the target recognition model to obtain a target segmentation threshold corresponding to the target position image;

[0153] The target marking point in the target position image is subjected to target recognition according to the target segmentation threshold value to obtain the target position information corresponding to the target marking point in the target position image.

[0154] In some implementations, the processor 301, in the process of using the target fitting model of the deep learning algorithm to fit the target position information to obtain the target edge information corresponding to the target marking point, executes:

[0155] According to the radius determination layer of the target fitting model, the image radius corresponding to the target marking point is obtained by utilizing the conversion relationship between the shooting position corresponding to the target position image and the real radius corresponding to the target marking point;

[0156] Performing position clustering using the image radius and the target position information according to the data clustering layer of the target fitting model to obtain a position clustering result;

[0157] Performing information fitting on the position clustering result according to the target fitting layer of the target fitting model to obtain an initial fitting center and an initial fitting radius corresponding to the target mark point;

[0158] According to the target determination layer of the target fitting model, the initial fitting radius and the image radius are used to perform information screening on the position clustering result to obtain a target fitting radius;

[0159] Obtaining a target fitting center corresponding to the target fitting radius from the initial fitting center;

[0160] The target edge information corresponding to the target marking point is obtained according to the target fitting center and the target fitting radius.

[0161] In some implementations, the processor 301 performs, during the process of performing information fitting on the position clustering result according to the target fitting layer of the target fitting model to obtain the initial fitting center and the initial fitting radius corresponding to the target marker point:

[0162] Determining the target optimization function and target parameters corresponding to each cluster in the position clustering result according to the target fitting layer;

[0163] Performing partial derivative calculation on the target parameter in the target optimization function according to the target fitting layer to obtain a target partial derivative function corresponding to the target parameter;

[0164] Determine a solution formula corresponding to the target parameter by fusing the target partial derivative function with the target fitting layer;

[0165] According to the target fitting layer, the initial fitting center and the initial fitting radius corresponding to each cluster in the position clustering result are used by the solution formula;

[0166] Wherein, the objective optimization function and the objective parameters are obtained according to the following formula:

[0167]

[0168] a=-2*x0;

[0169] b = -2*y0;

[0170] c=x0 2 +y0 2 -r 2 ;

[0171] Wherein, LOSS represents the target optimization function, a, b, c all represent the target parameters, M represents the number of targets corresponding to each cluster in the location clustering result, and x i Indicates the horizontal coordinate corresponding to each location information of each cluster in the location clustering result, y i The vertical coordinate corresponding to each location information of each cluster in the location clustering result, x0 represents the horizontal coordinate corresponding to the initial fitting center of each cluster in the location clustering result, y0 represents the vertical coordinate corresponding to the initial fitting center of each cluster in the location clustering result, and r represents the initial fitting radius corresponding to each cluster in the location clustering result.

[0172] In some implementations, the marking area includes a first area and a second area, and the processor 301, in the process of determining the movement direction and the movement amplitude corresponding to the target marking point according to the marking area, performs:

[0173] Determine a target coordinate system, and determine a first coordinate system corresponding to the first area and a second coordinate system corresponding to the second area;

[0174] Calculating a first transformation relationship between the first coordinate system and the target coordinate system and calculating a second transformation relationship between the second coordinate system and the target coordinate system;

[0175] According to the first conversion relationship, the first area is converted into the target coordinate system to obtain a first target area;

[0176] According to the second conversion relationship, the second area is converted into the target coordinate system to obtain a second target area;

[0177] Performing edge alignment on the first target area and the second target area to obtain a target alignment result;

[0178] The movement direction and the movement amplitude corresponding to the target marking point are determined according to the target alignment result.

[0179] In some implementations, the processor 301, in the process of performing risk warning according to the movement direction and the movement amplitude and obtaining a target warning result, executes:

[0180] Determining an amplitude range corresponding to the radiotherapy device in the movement direction;

[0181] Compare the amplitude range and the motion amplitude to obtain a comparison result;

[0182] A risk warning is performed according to the comparison result to obtain the target warning result.

[0183] It should be noted that technicians in the relevant field can clearly understand that for the convenience and simplicity of description, the specific working process of the terminal device described above can refer to the corresponding process in the aforementioned embodiment of the auxiliary positioning method for tumor intensity-modulated radiation therapy, and will not be repeated here.

[0184] An embodiment of the present invention also provides a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of any one of the auxiliary positioning methods for tumor intensity-modulated radiation provided in the description of the embodiment of the present invention.

[0185] The storage medium may be an internal storage unit of the terminal device described in the foregoing embodiment, such as a hard disk or memory of the terminal device. The storage medium may also be an external storage device of the terminal device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc., equipped on the terminal device.

[0186] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware embodiment, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those of ordinary skill in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0187] It should be understood that the term "and / or" used in the present specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations. It should be noted that, in this article, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "including a..." does not exclude the presence of other identical elements in the process, method, article or system including the element.

[0188] The serial numbers of the embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. The above description is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for auxiliary positioning of tumor intensity modulated radiation therapy, characterized in that: The method comprises: According to the target sensor, a target position image corresponding to a target marking point of a target patient is captured in real time; Using a deep learning algorithm to identify the position change of the target marking point in the target position image, and obtain the movement direction and movement amplitude corresponding to the target marking point; Perform risk warning according to the movement direction and the movement amplitude to obtain a target warning result; The beam irradiation area of ​​the radiotherapy equipment corresponding to the target patient is determined according to the target warning result, the movement direction and the movement amplitude, so that the radiotherapy equipment performs radiation beam positioning for the target patient according to the beam irradiation area.

2. The method according to claim 1, characterized in that The step of using a deep learning algorithm to identify the position change of the target marking point in the target position image to obtain the movement direction and movement amplitude corresponding to the target marking point includes: Using the target recognition model of the deep learning algorithm to perform target recognition on the target marking point in the target position image, and obtain target position information corresponding to the target marking point in the target position image; Using the target fitting model of the deep learning algorithm to perform information fitting on the target position information, the target edge information corresponding to the target marking point is obtained; The marking area corresponding to the target marking point is determined according to the target edge information, and the movement direction and the movement amplitude corresponding to the target marking point are determined according to the marking area.

3. The method according to claim 2, characterized in that The target recognition model of the deep learning algorithm is used to perform target recognition on the target marking point in the target position image to obtain target position information corresponding to the target marking point in the target position image, including: Obtaining a first objective function and a second objective function required for image segmentation of the target position image according to the target determination layer of the target recognition model; Calculating a first segmentation threshold corresponding to the target position image under the first objective function using a particle swarm optimization algorithm according to the threshold calculation layer of the target recognition model; Calculate the second segmentation threshold corresponding to the target position image under the second objective function by using the particle swarm optimization algorithm and the first segmentation threshold according to the threshold calculation layer; Performing an optimal threshold analysis using the first segmentation threshold and the second segmentation threshold according to the threshold determination layer of the target recognition model to obtain a target segmentation threshold corresponding to the target position image; The target marking point in the target position image is subjected to target recognition according to the target segmentation threshold value to obtain the target position information corresponding to the target marking point in the target position image.

4. The method according to claim 2, characterized in that: The using the target fitting model of the deep learning algorithm to perform information fitting on the target position information to obtain the target edge information corresponding to the target marking point includes: According to the radius determination layer of the target fitting model, the image radius corresponding to the target marking point is obtained by utilizing the conversion relationship between the shooting position corresponding to the target position image and the real radius corresponding to the target marking point; Performing position clustering using the image radius and the target position information according to the data clustering layer of the target fitting model to obtain a position clustering result; Performing information fitting on the position clustering result according to the target fitting layer of the target fitting model to obtain an initial fitting center and an initial fitting radius corresponding to the target mark point; According to the target determination layer of the target fitting model, the initial fitting radius and the image radius are used to perform information screening on the position clustering result to obtain a target fitting radius; Obtaining a target fitting center corresponding to the target fitting radius from the initial fitting center; The target edge information corresponding to the target marking point is obtained according to the target fitting center and the target fitting radius.

5. The method according to claim 4, characterized in that The step of performing information fitting on the position clustering result according to the target fitting layer of the target fitting model to obtain an initial fitting center and an initial fitting radius corresponding to the target mark point includes: Determining a target optimization function and target parameters corresponding to each cluster in the position clustering result according to the target fitting layer; Performing partial derivative calculation on the target parameter in the target optimization function according to the target fitting layer to obtain a target partial derivative function corresponding to the target parameter; Determine a solution formula corresponding to the target parameter by fusing the target partial derivative function with the target fitting layer; According to the target fitting layer, the initial fitting center and the initial fitting radius corresponding to each cluster in the position clustering result are used by the solution formula; Wherein, the objective optimization function and the objective parameters are obtained according to the following formula: a=-2*x0; b = -2*y0; c=x0 2 +y0 2 -r 2 ; Wherein, LOSS represents the target optimization function, a, b, c all represent the target parameters, M represents the number of targets corresponding to each cluster in the location clustering result, and x i Indicates the horizontal coordinate corresponding to each location information of each cluster in the location clustering result, y i The vertical coordinate corresponding to each location information of each cluster in the location clustering result, x0 represents the horizontal coordinate corresponding to the initial fitting center of each cluster in the location clustering result, y0 represents the vertical coordinate corresponding to the initial fitting center of each cluster in the location clustering result, and r represents the initial fitting radius corresponding to each cluster in the location clustering result.

6. The method according to claim 2, characterized in that The marking area includes a first area and a second area, and determining the movement direction and the movement amplitude corresponding to the target marking point according to the marking area includes: Determine a target coordinate system, and determine a first coordinate system corresponding to the first area and a second coordinate system corresponding to the second area; Calculating a first transformation relationship between the first coordinate system and the target coordinate system and calculating a second transformation relationship between the second coordinate system and the target coordinate system; According to the first conversion relationship, the first area is converted into the target coordinate system to obtain a first target area; According to the second conversion relationship, the second area is converted into the target coordinate system to obtain a second target area; Performing edge alignment on the first target area and the second target area to obtain a target alignment result; The movement direction and the movement amplitude corresponding to the target marking point are determined according to the target alignment result.

7. The method according to claim 1, characterized in that The performing risk warning according to the movement direction and the movement amplitude to obtain a target warning result includes: Determining an amplitude range corresponding to the radiotherapy device in the movement direction; Compare the amplitude range and the motion amplitude to obtain a comparison result; A risk warning is performed according to the comparison result to obtain the target warning result.

8. An auxiliary positioning system for tumor intensity modulated radiation therapy, characterized in that: include: A data acquisition module, used for capturing a target position image corresponding to a target marking point of a target patient in real time according to a target sensor; A change recognition module is used to use a deep learning algorithm to identify the position change of the target marking point in the target position image, and obtain the movement direction and movement amplitude corresponding to the target marking point; An information warning module is used to perform risk warning according to the movement direction and the movement amplitude to obtain a target warning result; The area determination module is used to determine the beam irradiation area of ​​the radiotherapy equipment corresponding to the target patient according to the target warning result, the movement direction and the movement amplitude, so that the radiotherapy equipment can perform radiation beam positioning on the target patient according to the beam irradiation area.

9. A terminal device, characterized in that: The terminal device includes a processor and a memory; The memory is used to store computer programs; The processor is used to execute the computer program and implement an auxiliary positioning method for tumor intensity modulated radiation therapy according to any one of claims 1 to 7 when executing the computer program.

10. A computer storage medium for computer storage, characterized in that: The computer storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps of an auxiliary positioning method for tumor intensity-modulated radiation therapy as described in any one of claims 1 to 7.

Images