A real-time image-guided method and system for organ motion compensation in tumor radiotherapy

By constructing a respiratory motion model and updating the gold standard position in real time, the problem of model prediction bias in traditional tumor radiotherapy is solved, achieving precise localization of the tumor target area and improving treatment efficacy.

CN121921341BActive Publication Date: 2026-06-23SHAANXI CANCER HOSPITAL (SHAANXI INST OF CANCER PREVENTION & TREATMENT) (SHAANXI THIRD PEOPLES HOSPITAL)

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SHAANXI CANCER HOSPITAL (SHAANXI INST OF CANCER PREVENTION & TREATMENT) (SHAANXI THIRD PEOPLES HOSPITAL)
Filing Date
2026-03-26
Publication Date
2026-06-23

AI Technical Summary

Technical Problem

In existing technologies, traditional organ motion compensation methods for tumor radiotherapy do not take into account the displacement of the gold standard and tissue deformation during radiotherapy, which leads to model prediction bias and affects the accuracy of radiotherapy.

Method used

By collecting the three-dimensional spatial position of the gold standard and the movement trajectory of the LED markers on the body surface in real time, a respiratory motion model is constructed. By combining the shrinkage ratio of the gold standard volume and the displacement of the center of gravity, the shrinkage and rotation compensation components are obtained, and the model parameters are updated in real time to predict the position of the gold standard.

Benefits of technology

It significantly improves the accuracy of target localization in tumor radiotherapy, ensuring that radiation accurately irradiates the tumor target area, improving the local tumor control rate and enhancing the patient's quality of life.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121921341B_ABST
    Figure CN121921341B_ABST
Patent Text Reader

Abstract

The application discloses a kind of real-time image guided method and system for organ motion compensation in tumor radiotherapy, comprising: by synchronous acquisition gold standard three-dimensional position and body surface LED marker point motion trajectory, construct respiratory motion model, combine gold standard volume atrophy ratio and barycenter displacement in radiotherapy process to obtain atrophy compensation component, then rely on historical radiotherapy data, gold standard vector and position relationship are weighted fusion to obtain accurate rotation compensation component, gold standard current position can also be input model real-time update parameter and combine two kinds of compensation components to predict next time gold standard position, while through actual position comparison, realize model cyclic optimization.The application greatly improves the accuracy of target area positioning in tumor radiotherapy, ensures that radioactive rays can accurately irradiate tumor target area, and then improves tumor local control rate and improves patient quality of life.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of image processing technology and relates to a real-time image-guided method and system for organ motion compensation in tumor radiotherapy. Background Technology

[0002] The efficacy of tumor radiotherapy is highly dependent on the precise localization of the tumor target area. However, physiological activities such as breathing and heartbeat can cause real-time movement of internal organs, resulting in continuous changes in the position of the tumor target area. Therefore, organ motion compensation technology is needed to monitor and correct the deviation of the tumor position in real time, so as to ensure that the radiation accurately irradiates the target area while protecting the surrounding normal tissues to the greatest extent, reducing the risk of radiation damage and complications, improving the local tumor control rate, and improving the patient's quality of life.

[0003] In current clinical practice, organ motion compensation in tumor radiotherapy often employs a real-time image-guided robotic tracking system. This technology uses an X-ray-detectable metal marker implanted near the tumor as an internal motion reference. Before treatment, the initial tumor position is determined by taking X-rays and planning CT scans. Then, the patient's respiratory data is collected, and the position of the marker and the surface fluctuations of the chest and abdomen are monitored simultaneously to establish a mathematical model relating surface motion and internal marker motion. During treatment, the system predicts the tumor position based on real-time surface motion data and drives the robotic arm to adjust the treatment head to achieve ray tracking. Simultaneously, X-rays are taken intermittently to verify the consistency between the actual marker position and the model's predicted value. If the error exceeds the standard, the model is updated.

[0004] However, traditional techniques suffer from significant model prediction bias. The core reason lies in their reliance on a linear relationship between surface and internal movement to construct respiratory prediction models, failing to account for the displacement of implanted gold markers during radiotherapy due to physiological and pathological changes in tissues. This leads to model baseline failure, manifesting in three main ways: First, radiotherapy causes tumors to shrink gradually, moving surrounding gold markers inward and altering the spatial distance between the surface and the gold markers recorded during model establishment, resulting in systematic prediction bias. Second, radiotherapy-induced local tissue edema and subsequent fibrosis can push or pull on the gold markers, disrupting the original spatial correspondence between the gold markers and surface markers. The model cannot detect these tissue deformations in real time and continues to predict tumor location based on the original parameters, causing prediction errors. Third, in soft tissues such as the lungs and liver, gold markers are prone to migration of several millimeters due to respiratory movements or tissue relaxation, rendering the gold marker baseline coordinates on which the model relies invalid. Even if the radiation precisely hits the predicted site, it may still deviate from the tumor target area, ultimately affecting the accuracy and efficacy of radiotherapy. Therefore, a method capable of guiding the gold marker position in real time is urgently needed. Summary of the Invention

[0005] The purpose of this invention is to solve the problem in the prior art of organ motion compensation in tumor radiotherapy, where traditional respiratory prediction models do not take into account the displacement of implanted gold markers during radiotherapy due to tumor shrinkage, tissue edema / fibrosis and soft tissue relaxation, leading to model prediction bias. This invention provides a real-time image-guided method and system for organ motion compensation in tumor radiotherapy.

[0006] To achieve the above objectives, the present invention employs the following technical solution: a real-time image-guided method for organ motion compensation during tumor radiotherapy, comprising: Step 1, acquiring X-ray images of the patient, obtaining the three-dimensional spatial position of each gold marker implanted within the tumor area in real time, and simultaneously tracking the motion trajectory of light-emitting diode markers pasted on the patient's chest and abdomen surface, thereby constructing a respiratory motion model; Step 2, during radiotherapy, based on the volume shrinkage ratio of each gold marker obtained from the X-ray image at the current time compared to the previous time, and combined with the displacement of the center of gravity of each gold marker at the current time compared to the previous time, obtaining the shrinkage compensation component of each gold marker; Step 3, based on historical radiotherapy data, the gold markers at any given time point to all gold markers at the corresponding time. Step 4: Based on the atrophy compensation component and the positional relationship between each gold standard, obtain the initial rotation compensation component of the gold standard at any given time relative to the gold standard at the previous time. Step 5: Based on the atrophy compensation component and the initial rotation compensation component of each gold standard, obtain the rotation compensation component of each gold standard obtained from the X-ray at the current time compared to the previous time. Step 6: Input the current position into the respiratory motion model for real-time parameter updates, and obtain the predicted position result of the gold standard at the next time based on the atrophy compensation component and the rotation compensation component. Step 7: Take another X-ray to obtain the actual position of the gold standard. Compare the predicted position result of the gold standard with the actual position of the gold standard to determine whether the error between the two exceeds a preset threshold. If it exceeds the threshold, repeat steps 1 to 5 until the treatment needs are met.

[0007] A further improvement of the present invention is as follows: Specifically, the acquisition of X-ray images of the patient, real-time acquisition of the three-dimensional spatial position of each gold marker within the implanted tumor area, and simultaneous tracking of the motion trajectory of LED markers pasted on the patient's chest and abdomen to construct a respiratory motion model are achieved by: acquiring the position of the gold markers implanted near the tumor in three-dimensional space using an X-ray imaging device at a frequency of 0.5-1 seconds per acquisition; acquiring the LED markers pasted on the patient's chest and abdomen skin using an optical tracking device and recording their up-and-down trajectory; aligning the timestamp of each frame of X-ray image with the corresponding position of the markers on the body surface; forming a time-synchronized multi-source motion dataset from several sets of data points including time, body surface position, and gold marker position; and using a multinomial regression algorithm to fit the quadratic function relationship between the body surface displacement and the gold marker displacement to construct a respiratory motion model.

[0008] Furthermore, during radiotherapy, based on the volume shrinkage ratio of each gold standard obtained from the X-ray at the current time compared to the previous time, and combined with the displacement of the center of gravity of each gold standard at the current time compared to the previous time, the shrinkage compensation component of each gold standard is obtained. Specifically, this involves: obtaining the volume of all gold standards at different times before the patient undergoes radiotherapy, and fitting the overall change pattern of the volume of all gold standards with respiratory motion to obtain a fitting function; during the patient undergoes radiotherapy, without considering the influence of respiratory motion, obtaining the overall shrinkage ratio of the gold standard volume obtained from the X-ray at the current time compared to the previous time; based on the overall shrinkage ratio, and combined with the displacement of the center of gravity of each gold standard at the current time compared to the previous time, obtaining the shrinkage compensation component of each gold standard.

[0009] Furthermore, the process of obtaining the volume formed by all gold markers at different times before radiotherapy, and fitting the overall change pattern of the volume formed by all gold markers with respiratory movements to obtain a fitting function, specifically involves: obtaining the convex hull of the gold marker point set based on the X-ray image at each time point, i.e., the smallest convex geometric figure that encloses all gold markers, and then calculating the volume of the convex hull based on the shoelace formula; constructing a fitting function with a sinusoidal fitting function. This represents the curve of the volume formed by the gold standard changing over time, and the least squares method is used to evaluate the function. Perform fitting to obtain the optimal parameters , , and .

[0010] Furthermore, the elimination of breathing interference involves: taking two adjacent time points... and Substitute the sine fitting function The theoretical volume values ​​caused solely by respiratory movements at the two time points were calculated. and The difference between the two is the theoretical volume change caused solely by respiratory movements during that time interval, and the formula is: The specific steps for obtaining the overall shrinkage ratio of the gold standard volume obtained from the X-ray at the current moment compared to the previous moment are as follows: in, Indicates the current Compared to the previous moment The overall shrinkage rate of the gold standard volume obtained from the X-ray image. Indicates the current time The volume of the gold standard obtained from the X-ray film. Indicates the current The previous moment The volume of the gold standard obtained from the X-ray film. This indicates the overall change in the volume of the gold standard obtained from the X-ray image at the current moment compared to the previous moment. This represents the current time obtained by fitting a sine function. The gold standard volume of the X-ray film. Indicates the current The previous moment The gold standard volume of the X-ray film. This indicates the change in the volume of the gold sample obtained from the X-ray at the current moment compared to the previous moment, as a result of respiration.

[0011] Furthermore, based on the overall shrinkage ratio and combined with the displacement of the centroid of each gold standard at the current moment compared to the previous moment, the shrinkage compensation component of each gold standard is obtained separately, specifically as follows: in, Indicates the current time Compared to the previous moment The X-ray film was obtained The shrinkage compensation component of each gold standard Indicates the current time The X-ray film was obtained The vector from each gold standard to the geometric center of all gold standards. Indicates the current time The previous moment The X-ray film was obtained The vector from one gold standard to the geometric center of all gold standards.

[0012] Furthermore, based on historical radiotherapy data, and based on the vectors pointing from the gold markers at the current and previous times to the geometric centers of all gold markers at the corresponding times, as well as the positional relationships between the gold markers, the preliminary rotation compensation component of the gold marker at any time relative to the gold marker at the previous time is obtained. Specifically, based on historical radiotherapy data, one gold marker is randomly selected as the reference gold marker, and the positional relationship between the reference gold marker and other gold markers is obtained. Based on the first Always refer to the gold standard Pointing to the geometric center of all gold standards at that moment vector The moment before Reference Gold Standard Pointing to the geometric center of all gold standards at that moment vector Obtain the rotation matrix Based on the positional relationship between the reference gold standard and other gold standards With rotation matrix The process involves obtaining the initial rotational compensation component of the gold marker at any given time relative to the gold marker at the previous time; the historical radiotherapy data refers to the three-dimensional spatial position of the gold marker group during the radiotherapy process; specifically, obtaining the initial rotational compensation component of the gold marker at any given time relative to the gold marker at the previous time involves: in, Indicates the first At the [time]th moment Each gold medal compared to the previous moment The initial rotational compensation component, Indicated by An exponential function with base 0. Indicates the first At the [time]th moment The gold medal to the first The distance to the gold standard Indicates the first At the [time]th moment The vector from the gold standard to the center of the gold standard Compared to the previous moment Vector to the geometric center of all gold standards The rotation matrix.

[0013] Furthermore, the step of obtaining the rotation compensation component of each gold standard based on the shrinkage compensation component and the preliminary rotation compensation component of each gold standard, compared to the X-ray image at the current time, specifically involves: using both time proximity and the magnitude of the shrinkage compensation component as dual weights, weighted fusing historical preliminary rotation compensation component data to obtain the rotation compensation component of each gold standard obtained at the current time compared to the X-ray image at the previous time; for all historical times... Perform maximum and minimum value normalization to obtain the shrinkage deformation characteristic function. Historical moments during radiotherapy Quantization is performed to obtain the time-dimensional normalized weight function. The shrinkage deformation characteristic function is fused with the time-dimensional normalized weight function, and based on the preliminary rotation compensation component data, the rotation compensation components of each gold standard obtained from the X-ray at the current time compared to the previous time are obtained. The formula is as follows: in, Indicates the current time Compared to the previous moment The first X-ray obtained The rotational compensation component of each gold standard This represents the maximum and minimum value normalization function. Indicating the first stage of radiotherapy A historic moment; This is the time-dimension normalized weight function.

[0014] Furthermore, the current position is input into the respiratory motion model for real-time parameter updates, and the predicted position of the gold standard at the next moment is obtained based on the shrinkage compensation component and the rotation compensation component. Specifically, the current position is the current moment during radiotherapy. The measured three-dimensional coordinates of the X-ray image are used to input the current position into the respiratory motion model. The respiratory motion model updates its parameters in real time and predicts the next moment. The gold standard base position considering only respiratory movements Based on the shrinkage compensation component and the rotation compensation component, the predicted position of the gold standard at the next moment is obtained, specifically: in, This indicates that the output should be the next time step after the current time step has been optimized. The The spatial location prediction results of the gold standard Indicates the next moment after the current moment. The The spatial location prediction results of the gold standard This indicates the current time compared to the previous time. The first X-ray obtained The rotational compensation component of each gold standard Indicates the current Compared to the previous moment The X-ray film was obtained The shrinkage compensation component of each gold standard.

[0015] A real-time image-guided system for organ motion compensation during tumor radiotherapy includes: a construction module, which acquires X-ray images of the patient, obtains the three-dimensional spatial position of each gold marker implanted in the tumor area in real time, and simultaneously tracks the motion trajectory of light-emitting diode markers pasted on the patient's chest and abdomen to construct a respiratory motion model; a first acquisition module, which, during radiotherapy, acquires the shrinkage compensation component of each gold marker based on the volume shrinkage ratio of each gold marker obtained from the X-ray image at the current moment compared to the previous moment, and combines this with the displacement of the center of gravity of each gold marker at the current moment compared to the previous moment; and a second acquisition module, which, based on historical radiotherapy data, determines the gold marker pointing to the corresponding moment at any given time compared to the previous moment. The system uses vectors representing the geometric centers of all gold markers and their positional relationships to obtain the initial rotational compensation component of each gold marker relative to the previous gold marker at any given time. A third acquisition module, based on the shrinkage compensation component and the initial rotational compensation component of each gold marker, obtains the rotational compensation component of each gold marker at the current time compared to the previous X-ray image. An update module inputs the current position into the respiratory motion model for real-time parameter updates and, based on the shrinkage compensation component and the rotational compensation component, obtains the predicted position of the gold marker at the next time. A comparison module re-examines the X-ray to obtain the actual position of the gold marker and compares it with the predicted position until the treatment requirements are met.

[0016] Compared with existing technologies, this invention has the following beneficial effects: This invention constructs a respiratory motion model by simultaneously acquiring the three-dimensional position of the gold marker and the motion trajectory of LED markers on the body surface. It obtains a shrinkage compensation component by combining the gold marker volume shrinkage ratio and center of gravity displacement during radiotherapy. Then, based on historical radiotherapy data, gold marker vectors, and positional relationships, it obtains a precise rotation compensation component through weighted fusion. Furthermore, it can input the current position of the gold marker into the model to update parameters in real time and predict the gold marker position at the next moment by combining the two types of compensation components. Simultaneously, it achieves iterative optimization of the model through comparison with actual positions. This invention significantly improves the accuracy of target area localization in tumor radiotherapy, ensuring that radiation accurately irradiates the tumor target area, thereby improving the local tumor control rate and enhancing the patient's quality of life. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating the real-time image-guided method for organ motion compensation in tumor radiotherapy according to the present invention. Figure 2 This is a schematic diagram of the real-time image-guided system for organ motion compensation in tumor radiotherapy according to the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0020] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0021] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0022] In the description of the embodiments of the present invention, it should be noted that if terms such as "upper," "lower," "horizontal," or "inner" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. Furthermore, terms such as "first" and "second" are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0023] Furthermore, the use of the term "horizontal" does not imply that the component must be absolutely horizontal, but rather that it can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal than "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0024] In the description of the embodiments of the present invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention according to the specific circumstances.

[0025] The present invention will now be described in further detail with reference to the accompanying drawings: See also Figure 1 This invention discloses a real-time image-guided method for organ motion compensation during tumor radiotherapy, comprising: S101, acquiring X-ray images of the patient, obtaining the three-dimensional spatial position of each gold marker implanted in the tumor area in real time, and simultaneously tracking the motion trajectory of light-emitting diode markers pasted on the patient's chest and abdomen to construct a respiratory motion model; before tumor radiotherapy, the patient maintains the treatment position and performs stable and normal breathing. The system simultaneously activates the X-ray imaging equipment and the optical tracking equipment to acquire baseline data for establishing the respiratory motion model.

[0026] The position of the gold marker implanted near the tumor in three-dimensional space was acquired using X-ray imaging equipment at a frequency of 0.5-1 seconds per acquisition. An optical tracking device was used to acquire LED markers attached to the patient's chest and abdomen skin and record their up-and-down trajectory. The timestamp of each frame of X-ray image was aligned with the position of the marker on the body surface at the corresponding time. Several sets of data points containing time, body surface position, and gold marker position were collected to form a time-synchronized multi-source motion dataset. A multinomial regression algorithm was used to fit the quadratic function relationship between body surface displacement and gold marker displacement to construct a respiratory motion model.

[0027] S102, During radiotherapy, based on the volume shrinkage ratio of each gold marker obtained from the X-ray image at the current time compared to the previous time, and combined with the displacement of the center of gravity of each gold marker at the current time compared to the previous time, the shrinkage compensation component of each gold marker is obtained. Before radiotherapy, the patient is in a stable treatment phase, and the tumor volume has not yet shrunk or proliferated. At this time, the fluctuation of the gold marker group volume is mainly dominated by respiratory physiological motion, showing obvious periodicity. In order to accurately separate the true deformation caused by treatment from the overall volume change of the gold markers, a pure respiratory motion model is established by fitting a sine function before radiotherapy. S102.1, Obtain the volume of all gold markers at different times before the patient undergoes radiotherapy, and fit the size of the volume of all gold markers with the overall change law of respiratory motion to obtain the fitting function; Based on the gold marker point set in the X-ray image at each time, obtain the convex hull of all gold marker point sets at each time, that is, the smallest convex geometry that wraps all gold markers, and then calculate the volume of the convex hull based on the shoelace formula; Construct a model with a sine fitting function. This represents the curve of the volume formed by the gold standard changing over time, and the least squares method is used to evaluate the function. Perform fitting to obtain the optimal parameters , , and .

[0028] S102.2 During radiotherapy, without considering the influence of respiratory motion, the overall shrinkage ratio of the gold-labeled volume obtained from the X-ray at the current moment compared to the previous moment is obtained. During radiotherapy, the change in gold-labeled volume observed on X-ray is the result of multiple factors, mainly including periodic normal changes caused by respiratory motion and non-periodic changes caused by clinical deformations such as tumor shrinkage / tissue edema. If the influence of respiration is not eliminated, normal fluctuations caused by respiration may be misjudged as shrinkage or swelling, leading to systematic errors in model prediction. Therefore, it is necessary to use the respiratory baseline model established before radiotherapy to remove respiratory interference from the current measured volume in order to truly reflect the net volume changes of the tumor and its markers.

[0029] The elimination of breathing interference is: taking two adjacent time intervals... and Substitute the sine fitting function The theoretical volume values ​​caused solely by respiratory movements at the two time points were calculated. and The difference between the two is the theoretical volume change caused solely by respiratory movements during that time interval, and the formula is: The specific steps for obtaining the overall shrinkage ratio of the gold standard volume obtained from the X-ray at the current moment compared to the previous moment are as follows: in, Indicates the current Compared to the previous moment The overall shrinkage rate of the gold standard volume obtained from the X-ray image. Indicates the current time The volume of the gold standard obtained from the X-ray film. Indicates the current The previous moment The volume of the gold standard obtained from the X-ray film. This indicates the overall change in the volume of the gold standard obtained from the X-ray image at the current moment compared to the previous moment. This represents the current time obtained by fitting a sine function. The gold standard volume of the X-ray film. Indicates the current The previous moment The gold standard volume of the X-ray film. This indicates the change in the volume of the gold sample obtained from the X-ray at the current moment compared to the previous moment, as a result of respiration.

[0030] S102.3 Based on the overall shrinkage ratio, and combined with the displacement of the center of gravity of each gold standard at the current time compared with the previous time, the shrinkage compensation component of each gold standard is obtained.

[0031] The obtained overall shrinkage ratio describes the macroscopic change in the volume of the gold standard cluster, but the direction and magnitude of the individual displacement of each gold standard may differ. The positional changes of individual gold standards are also mixed with respiratory motion and individual deformation. For each gold standard's displacement, the overall shrinkage ratio should be used as a global factor, combined with the gold standard's relative position to the center of gravity, to decompose the macroscopic shrinkage into the microscopic compensation amount of each gold standard. This aims to ensure the physical consistency of the compensation while eliminating residual respiratory noise. Therefore, multiplying the relative displacement vector of an individual gold standard by the overall shrinkage ratio allows estimation of the component attributable to overall deformation in that gold standard's displacement, i.e., the individual shrinkage compensation component.

[0032] Therefore, the shrinkage compensation component for each gold standard is obtained as follows: in, Indicates the current time Compared to the previous moment The X-ray film was obtained The shrinkage compensation component of each gold standard Indicates the current time The X-ray film was obtained The vector from each gold standard to the geometric center of all gold standards. Indicates the current time The previous moment The X-ray film was obtained The vector from one gold standard to the geometric center of all gold standards.

[0033] S103, based on historical radiotherapy data, obtains the vectors pointing from the gold markers to the geometric centers of all gold markers at any given time, and the positional relationships between the gold markers, to obtain the preliminary rotation compensation component of the gold markers at any given time relative to the gold markers at the previous time. The rotation change of the vector of a single gold marker relative to the geometric centers of all gold markers from one time to the next may contain noise or local mutations. However, in real tumor tissue, due to the continuity and rigidity of the tissue, the motion of adjacent gold markers should be correlated and smooth, without drastic mutations. Therefore, the rotation matrix calculated directly using a single gold marker may be unreliable. Therefore, the rotation information of surrounding gold markers is used to smooth or correct it. The true rotation component of a gold marker should be similar to the rotation components of its neighboring gold markers. The closer the gold markers are, the more similar their motion patterns are, and therefore, they should be given greater weight during correction.

[0034] The specific steps are as follows: Based on historical radiotherapy data, select one gold standard as the reference gold standard, and obtain the positional relationship between the reference gold standard and other gold standards. Based on the first Always refer to the gold standard Pointing to the geometric center of all gold standards at that moment vector The moment before Reference Gold Standard Pointing to the geometric center of all gold standards at that moment vector Obtain the rotation matrix Based on the positional relationship between the reference gold standard and other gold standards With rotation matrix The process involves obtaining the initial rotational compensation component of the gold marker at any given time relative to the gold marker at the previous time; the historical radiotherapy data refers to the three-dimensional spatial position of the gold marker group during the radiotherapy process; specifically, obtaining the initial rotational compensation component of the gold marker at any given time relative to the gold marker at the previous time involves: in, Indicates the first At the [time]th moment Each gold medal compared to the previous moment The initial rotational compensation component, Indicated by An exponential function with base 0. Indicates the first At the [time]th moment The gold medal to the first The distance to the gold standard Indicates the first At the [time]th moment The vector from the gold standard to the center of the gold standard Compared to the previous moment Vector to the geometric center of all gold standards The rotation matrix.

[0035] S104. Based on the shrinkage compensation component and preliminary rotation compensation component of each gold standard, obtain the rotation compensation component of each gold standard obtained from the X-ray at the current time compared to the previous time. Historical rotation data contains information from different times. The closer the historical time is to the current time, the more similar its motion state is to the current time, and the higher its predictive value. Therefore, its weight should be greater. Furthermore, if the shrinkage compensation component of a gold standard is large at a certain historical time, it indicates that significant deformation has occurred at that time, indicating drastic changes in the tumor structure. Significant deformation is often accompanied by more obvious changes in tissue structure and spatial relationships. Therefore, the rotation information at that time may be more valuable for predicting the current rotation caused by deformation, and its weight should also be greater. Therefore, based on the recent time, if the shrinkage compensation component is large, the weight of each gold standard group calculated in history relative to the preliminary rotation compensation component of the previous time is higher.

[0036] Therefore, using both temporal proximity and the magnitude of the shrinkage compensation component as dual weights, historical preliminary rotation compensation component data are weighted and fused to obtain the rotation compensation components of each gold standard obtained from the X-ray at the current moment compared to the previous moment; for all historical moments... Perform maximum and minimum value normalization to obtain the shrinkage deformation characteristic function. Historical moments during radiotherapy Quantization is performed to obtain the time-dimensional normalized weight function. The shrinkage deformation characteristic function is fused with the time-dimensional normalized weight function, and based on the preliminary rotation compensation component data, the rotation compensation components of each gold standard obtained from the X-ray at the current time compared to the previous time are obtained. The formula is as follows: in, Indicates the current time Compared to the previous moment The first X-ray obtained The rotational compensation component of each gold standard This represents the maximum and minimum value normalization function. Indicating the first stage of radiotherapy A historic moment; This is the time-dimension normalized weight function.

[0037] S105: The current position is input into the respiratory motion model for real-time parameter updates. Based on the atrophy compensation component and rotation compensation component, the predicted position of the gold standard at the next moment is obtained. Generally, the standard respiratory motion model predicts the gold standard position based on surface signals under the assumption of no clinical deformation. It only models respiratory motion and does not include the two key deformation factors of atrophy or proliferation and rotation caused by radiotherapy, leading to predictions that deviate from the true position. Since tumor shrinkage is a gradual process of cell apoptosis, necrosis, and absorption under radiation, similarly, tissue edema is a continuous manifestation of inflammatory response, and fibrosis is a long-term process of tissue repair and remodeling, we can now assume that this deformation trend will continue from the current moment to the next moment. Therefore, by superimposing the deformation compensation amount estimated at the current moment onto the base position predicted by the respiratory motion model, we can construct a representation of the next moment after the current moment has been optimized. No. Formula for predicting the spatial location of each gold standard.

[0038] The current location refers to the current moment during radiotherapy. The measured three-dimensional coordinates of the X-ray image are used to input the current position into the respiratory motion model. The respiratory motion model updates its parameters in real time and predicts the next moment. The gold standard base position considering only respiratory movements Based on the shrinkage compensation component and the rotation compensation component, the predicted position of the gold standard at the next moment is obtained, specifically: in, This indicates that the output should be the next time step after the current time step has been optimized. The The spatial location prediction results of the gold standard Indicates the next moment after the current moment. The The spatial location prediction results of the gold standard This indicates the current time compared to the previous time. The first X-ray obtained The rotational compensation component of each gold standard Indicates the current Compared to the previous moment The X-ray film was obtained The shrinkage compensation component of each gold standard.

[0039] S106, re-take X-ray to obtain the actual position of the gold standard; and compare the predicted position of the gold standard with the actual position of the gold standard to determine whether the error between the two exceeds the preset threshold; if it does, repeat S101 to S105 until the treatment needs are met.

[0040] During tumor treatment, X-ray images are re-taken to obtain the actual position of the gold standard. This measured value is compared with the model's predicted value; if the error is within the clinically acceptable range, treatment continues; if the error exceeds the threshold, the system automatically pauses irradiation and prompts medical staff to re-collect data and update the respiratory motion model to ensure the geometric accuracy of subsequent treatments. The robotic arm adjusts the attitude and beam direction of the treatment head in six degrees of freedom in real time based on the prediction results, dynamically correcting the irradiation target point so that the high-energy beam continuously aligns with and covers the tumor target area that moves due to respiration.

[0041] See Figure 2 This invention discloses a real-time image-guided system for organ motion compensation during tumor radiotherapy, comprising: a construction module, which acquires X-ray images of the patient, obtains the three-dimensional spatial position of each gold marker implanted in the tumor area in real time, and simultaneously tracks the motion trajectory of light-emitting diode markers pasted on the patient's chest and abdomen to construct a respiratory motion model; a first acquisition module, which, during radiotherapy, acquires the shrinkage compensation component of each gold marker based on the volume shrinkage ratio of each gold marker obtained from the X-ray image at the current moment compared to the previous moment, and combines this with the displacement of the center of gravity of each gold marker at the current moment compared to the previous moment; and a second acquisition module, which, based on historical radiotherapy data, compares the gold marker pointing at any given moment with the previous moment. The system employs a first module to obtain the vectors of the geometric centers of all gold markers at any given time, as well as the positional relationships between the gold markers, and to acquire the preliminary rotational compensation components of the gold markers at any given time relative to the gold markers at the previous time. A second module, based on the shrinkage compensation components and preliminary rotational compensation components of each gold marker, acquires the rotational compensation components of each gold marker obtained from the X-ray at the current time compared to the previous time. An third module, based on the shrinkage compensation components and the preliminary rotational compensation components, inputs the current position into the respiratory motion model for real-time parameter updates and, based on the shrinkage compensation components and rotational compensation components, acquires the predicted position of the gold markers at the next time. Finally, a comparison module re-examines the X-rays to obtain the actual positions of the gold markers and compares these predicted positions with the actual positions until the treatment requirements are met.

[0042] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A real-time image-guided method for organ motion compensation in tumor radiotherapy, characterized in that, include: Step 1: Acquire X-ray images of the patient, obtain the three-dimensional spatial position of each gold marker within the implanted tumor area in real time, and simultaneously track the movement trajectory of the LED markers pasted on the patient's chest and abdomen to construct a respiratory motion model; Step 2: During radiotherapy, based on the volume shrinkage ratio of each gold marker obtained from the X-ray at the current moment compared to the previous moment, and combined with the displacement of the center of gravity of each gold marker at the current moment compared to the previous moment, obtain the shrinkage compensation component of each gold marker; Step 3: Based on historical radiotherapy data, the vector pointing from the gold marker to the geometric center of all gold markers at any given moment compared to the previous moment, and the positional relationship between each gold marker, obtain the preliminary rotation compensation component of the gold marker at any given moment relative to the gold marker at the previous moment; Step 4: Based on the shrinkage compensation component and the preliminary rotation compensation component of each gold marker, obtain the rotation compensation component of each gold marker obtained from the X-ray at the current moment compared to the previous moment; Step 5: Input the current position into the respiratory motion model for real-time parameter updates, and obtain the predicted position of the gold marker at the next moment based on the shrinkage compensation component and the rotation compensation component; Step 6: Take another X-ray to obtain the actual position of the gold marker; The predicted location of the gold standard is compared with the actual location of the gold standard to determine whether the error between the two exceeds a preset threshold. If the volume exceeds the limit, repeat steps 1 to 5 until the treatment needs are met. During radiotherapy, based on the volume shrinkage ratio of each gold standard obtained from the X-ray image at the current time compared to the previous time, and combined with the displacement of the center of gravity of each gold standard at the current time compared to the previous time, obtain the shrinkage compensation component for each gold standard. Specifically: obtain the volume composed of all gold standards at different times before the patient undergoes radiotherapy, and fit the overall change pattern of the volume composed of all gold standards and respiratory motion to obtain a fitting function; during the patient's radiotherapy, eliminate the influence of respiratory motion, and obtain the volume of the X-ray image at the current time compared to the previous time. The overall shrinkage ratio of the gold marker volume obtained from the X-ray film; based on the overall shrinkage ratio, combined with the displacement of the center of gravity of each gold marker at the current moment compared to the previous moment, the shrinkage compensation component of each gold marker is obtained; the volume formed by all gold markers at different times before the patient undergoes radiotherapy is obtained, and the overall change law of the size of the volume formed by all gold markers and respiratory motion is fitted to obtain the fitting function, specifically: based on the gold marker point set in the X-ray image at each moment, the convex hull of the point set is obtained, that is, the smallest convex geometric figure that encloses all gold markers, and then the volume of the convex hull is calculated based on the shoelace formula; a sinusoidal fitting function is constructed. This represents the curve of the volume formed by the gold standard changing over time, and the least squares method is used to evaluate the function. Perform fitting to obtain the optimal parameters , , and To eliminate the influence of respiratory movements, the time intervals between two adjacent moments were... and Substitute the sine fitting function The theoretical volume values ​​caused solely by respiratory movements at the two time points were calculated. and The difference between the two is the theoretical volume change caused solely by respiratory movements during that time interval, and the formula is: To obtain the overall shrinkage ratio of the gold standard volume obtained from the X-ray at the current moment compared to the previous moment, specifically: in, Indicates the current Compared to the previous moment The overall shrinkage rate of the gold standard volume obtained from the X-ray image. Indicates the current time The volume of the gold standard obtained from the X-ray film. Indicates the current The previous moment The volume of the gold standard obtained from the X-ray film. This indicates the overall change in the volume of the gold standard obtained from the X-ray image at the current moment compared to the previous moment. This represents the current time obtained by fitting a sine function. The gold standard volume of the X-ray film. Indicates the current The previous moment The gold standard volume of the X-ray film. This indicates the change in the volume of the gold sample obtained from the X-ray at the current moment compared to the previous moment, as a result of respiration.

2. The real-time image-guided method for organ motion compensation in tumor radiotherapy according to claim 1, characterized in that, The process involves acquiring X-ray images of the patient, obtaining the three-dimensional spatial position of each gold marker implanted within the tumor area in real time, and simultaneously tracking the movement trajectory of LED markers pasted on the patient's chest and abdomen to construct a respiratory motion model. Specifically, the X-ray imaging device acquires the position of the gold markers implanted near the tumor in three-dimensional space at a frequency of 0.5-1 seconds per frame; an optical tracking device acquires the LED markers pasted on the patient's chest and abdomen skin and records their up-and-down trajectory; and the timestamp of each frame of X-ray image is aligned with the position of the markers on the body surface at the corresponding time. Several sets of data points, including time, body surface location, and gold standard location, were collected to form a time-synchronized multi-source motion dataset. A multinomial regression algorithm was used to fit the quadratic function relationship between body surface displacement and gold standard displacement to construct a respiratory motion model.

3. The real-time image-guided method for organ motion compensation in tumor radiotherapy according to claim 1, characterized in that, Based on the overall shrinkage ratio, and combined with the displacement of the center of gravity of each gold standard at the current moment compared to the previous moment, the shrinkage compensation component of each gold standard is obtained separately, specifically as follows: in, Indicates the current time Compared to the previous moment The X-ray film was obtained The shrinkage compensation component of each gold standard Indicates the current time The X-ray film was obtained The vector from each gold standard to the geometric center of all gold standards. Indicates the current time The previous moment The X-ray film was obtained The vector from one gold standard to the geometric center of all gold standards.

4. The real-time image-guided method for organ motion compensation in tumor radiotherapy according to claim 3, characterized in that, Based on historical radiotherapy data, and using the vectors pointing from the gold markers at the current and previous times to the geometric centers of all gold markers at the corresponding times, as well as the positional relationships between the gold markers, the preliminary rotational compensation component of the gold marker at any given time relative to the gold marker at the previous time is obtained. Specifically, based on historical radiotherapy data, a gold marker is randomly selected as a reference gold marker, and the positional relationship between the reference gold marker and other gold markers is obtained. Based on the first Always refer to the gold standard Pointing to the geometric center of all gold standards at that moment vector The moment before Reference Gold Standard Pointing to the geometric center of all gold standards at that moment vector Obtain the rotation matrix Based on the positional relationship between the reference gold standard and other gold standards With rotation matrix The process involves obtaining the initial rotational compensation component of the gold marker at any given time relative to the gold marker at the previous time; the historical radiotherapy data refers to the three-dimensional spatial position of the gold marker group during the radiotherapy process; specifically, obtaining the initial rotational compensation component of the gold marker at any given time relative to the gold marker at the previous time involves: in, Indicates the first At the [time]th moment Each gold medal compared to the previous moment The initial rotational compensation component, Indicated by An exponential function with base 0. Indicates the first At the [time]th moment The gold medal to the first The distance to the gold standard Indicates the first At the [time]th moment The vector from the gold standard to the center of the gold standard Compared to the previous moment Vector to the geometric center of all gold standards The rotation matrix.

5. The real-time image-guided method for organ motion compensation in tumor radiotherapy according to claim 4, characterized in that, The method involves obtaining the rotation compensation components of each gold standard based on its shrinkage compensation component and preliminary rotation compensation component, compared to the previous X-ray image. Specifically, this is achieved by weighting and fusing historical preliminary rotation compensation component data with both time proximity and the magnitude of the shrinkage compensation component as dual weights, to obtain the rotation compensation components of each gold standard compared to the previous X-ray image. This process is repeated for all historical moments. Perform maximum and minimum value normalization to obtain the shrinkage deformation characteristic function. ; Historical moments during radiotherapy Quantization is performed to obtain the time-dimensional normalized weight function. The shrinkage deformation characteristic function is fused with the time-dimensional normalized weight function, and based on the preliminary rotation compensation component data, the rotation compensation components of each gold standard obtained from the X-ray at the current time compared to the previous time are obtained. The formula is as follows: in, Indicates the current time Compared to the previous moment The first X-ray obtained The rotational compensation component of each gold standard This represents the maximum and minimum value normalization function. Indicating the first stage of radiotherapy A historic moment; This is the time-dimension normalized weight function.

6. The real-time image-guided method for organ motion compensation in tumor radiotherapy according to claim 5, characterized in that, The process involves inputting the current position into the respiratory motion model for real-time parameter updates, and obtaining the predicted position of the gold standard at the next moment based on the shrinkage compensation component and the rotation compensation component. Specifically, the current position refers to the current moment during radiotherapy. The measured three-dimensional coordinates of the X-ray image are used to input the current position into the respiratory motion model. The respiratory motion model updates its parameters in real time and predicts the next moment. The gold standard base position considering only respiratory movements ; Based on the shrinkage compensation component and the rotation compensation component, the predicted position of the gold standard at the next moment is obtained, specifically: in, This indicates that the output should be the next time step after the current time step has been optimized. The Spatial location prediction results of gold standard Indicates the next moment after the current moment. The Spatial location prediction results of gold standard This indicates the current time compared to the previous time. The first X-ray obtained The rotational compensation component of each gold standard Indicates the current Compared to the previous moment The X-ray film was obtained The shrinkage compensation component of each gold standard.

7. A real-time image-guided system for organ motion compensation in tumor radiotherapy, characterized in that, include: The system includes a construction module that acquires X-ray images of the patient, obtains the three-dimensional spatial position of each gold marker implanted within the tumor area in real time, and simultaneously tracks the movement trajectory of LED markers pasted on the patient's chest and abdomen to construct a respiratory motion model. A first acquisition module, during radiotherapy, acquires the shrinkage compensation component of each gold marker based on the volume shrinkage ratio of each gold marker obtained from the X-ray image at the current moment compared to the previous moment, combined with the displacement of the center of gravity of each gold marker at the current moment compared to the previous moment. Specifically, acquiring the overall volume shrinkage ratio of the gold markers obtained from the X-ray image at the current moment compared to the previous moment involves: in, Indicates the current Compared to the previous moment The overall shrinkage rate of the gold standard volume obtained from the X-ray image. Indicates the current time The volume of the gold standard obtained from the X-ray film. Indicates the current The previous moment The volume of the gold standard obtained from the X-ray film. This indicates the overall change in the volume of the gold standard obtained from the X-ray image at the current moment compared to the previous moment. This represents the current time obtained by fitting a sine function. The gold standard volume of the X-ray film. Indicates the current The previous moment The gold standard volume of the X-ray film. This represents the change in the gold-labeled volume obtained from the X-ray at the current moment compared to the previous moment, as a result of respiration; the formula for calculating the shrinkage compensation component is shown below: in, Indicates the current time Compared to the previous moment The X-ray film was obtained The shrinkage compensation component of each gold standard Indicates the current time The X-ray film was obtained The vector from each gold standard to the geometric center of all gold standards. Indicates the current time The previous moment The X-ray film was obtained The system consists of: a first acquisition module, a second acquisition module, and a third acquisition module. The second acquisition module uses historical radiotherapy data to obtain the vectors pointing from the gold markers at any given time to the geometric centers of all gold markers at the corresponding time, as well as the positional relationships between the gold markers. The system obtains the initial rotational compensation component of each gold marker relative to the previous time. The third acquisition module uses the shrinkage compensation component and the initial rotational compensation component of each gold marker to obtain the rotational compensation component of each gold marker at the current time compared to the previous time's X-ray. The system also includes an update module that inputs the current position into the respiratory motion model for real-time parameter updates and obtains the predicted position of the gold markers at the next time based on the shrinkage compensation component and the rotational compensation component. Finally, a comparison module re-examines the X-rays to obtain the actual position of the gold markers and compares the predicted position with the actual position until the treatment requirements are met.