Radiation dose and bone quantity fusion-based risk assessment method and system for dental implant

Abstract

The application discloses a kind of risk assessment method and system of dental implant based on radiotherapy dose and bone quantity fusion, comprising: obtaining the radiotherapy planning data and oral CT image data of target object;Radiotherapy planning data and oral CT image data are registered in three-dimensional space, and three-dimensional digital fusion model is constructed;In response to the selection of any to-be-implanted site on the model, the second irradiation dose and the second bone mass parameter statistical quantity of the target voxel set corresponding to the selected implantation site are automatically extracted;The safety of the second irradiation dose and the second bone mass parameter statistical quantity is evaluated based on the preset radiotherapy dose and bone mass joint determination rule.It is converted into local implantation site dose by constructing voxel-level double-quantity fusion model, and double-insurance evaluation is realized by combining bone mass, which solves the problem of single evaluation dimension, low precision and lack of quantitative tools in the prior art, and provides objective and accurate decision support for implantation and repair of head and neck tumor radiotherapy patients.

Description

Technical Field

[0001] This invention relates to the fields of medical digital image processing and dental implant planning technology, and in particular to a risk assessment method for dental implantation based on radiotherapy dose and bone fusion. Background Technology

[0002] Radiation therapy is an important treatment for head and neck tumors (such as nasopharyngeal carcinoma and oral cancer). However, high doses of radiation can cause irreversible vascular damage and osteonecrosis of the jawbone, significantly increasing the risk of radiation-induced osteonecrosis (ORN) and implant failure in subsequent dental implant restorations. Therefore, accurate risk assessment is crucial for patients who have previously received radiation therapy before undergoing implant surgery.

[0003] Currently, clinical assessment of implant placement in patients after radiotherapy faces the following main problems: 1) Single and crude risk assessment dimensions: Existing protocols often rely on the patient's "total radiotherapy dose" or "target area prescription dose" to determine implant risk. However, the dose distribution generated by modern precision radiotherapy techniques (such as IMRT and VMAT) is highly uneven, and the actual dose received by the periphery of the target area or even the contralateral jawbone may be far lower than the total dose. This "one-size-fits-all" assessment method leads to a large amount of bone actually located in the low-dose safe zone being incorrectly classified as high-risk, causing many patients to lose the opportunity for implant restoration. 2) Lack of precise quantitative tools: Doctors usually rely solely on experience to review independent radiotherapy planning reports and oral CT images, making it impossible to accurately match the abstract dose distribution with specific bone anatomical structures in the same three-dimensional space. The decision-making process is highly subjective and lacks objective, quantitative site-level data support. 3) Blindly designed implant plans: Without precise knowledge of the dose risk at specific implant sites, the design of implant location, number, and size lacks targeted avoidance strategies, making it difficult to achieve optimal function and aesthetics while ensuring safety.

[0004] In summary, existing technologies lack a technical solution that can deeply integrate radiotherapy dose information with bone mass information and conduct accurate and quantitative risk assessment based on this. Summary of the Invention

[0005] The purpose of this invention is to provide a risk assessment method for dental implants based on the fusion of radiotherapy dose and bone volume, so as to solve the problems of single risk assessment dimension, low accuracy and lack of quantitative tools in the prior art.

[0006] The first aspect of this invention provides a method for risk assessment of dental implants based on radiotherapy dose and bone fusion, comprising: Acquire radiotherapy planning data and oral CT image data of the target object, wherein the radiotherapy planning data includes spatially distributed first radiation dose information, and the oral CT image data includes at least first bone mass parameter information of the jawbone; The radiotherapy plan data and the oral CT image data are registered in three-dimensional space, and the first irradiation dose information is mapped to the voxel space of the oral CT image data to construct a three-dimensional digital fusion model containing the bone density attribute and irradiation dose attribute of each voxel. In response to the selection of any implantation site on the three-dimensional digital fusion model, the second irradiation dose and the second bone mass parameter statistics of the target voxel set corresponding to the selected implantation site are automatically extracted based on the fusion model. The safety of the second irradiation dose and the second bone mass parameter statistics is assessed based on the preset joint determination rules of radiotherapy dose and bone mass, and the risk assessment information of the implantation site is output according to the assessment results.

[0007] In one or more embodiments, prior to performing three-dimensional spatial registration of the radiotherapy planning data with the oral CT image data, the following steps are included: The spatial coordinate system of the oral CT image data is used as the reference coordinate system, and the spatial coordinate system of the radiotherapy plan data is used as the floating coordinate system; Anisotropic diffusion filtering was performed on the positioning CT images in the oral CT image data and the radiotherapy planning data respectively to suppress image noise while preserving the high-frequency feature signals of the jawbone cortical bone edge; Perform gray-level normalization based on bone density distribution histogram to map the gray-level values ​​of two sets of image data collected at different time points and from different imaging devices to a unified standardized range, thereby eliminating the gray-level offset field introduced by differences in imaging parameters. In the reference coordinate system and the floating coordinate system, anatomically invariant target bony landmarks are used as registration anchors based on curvature feature analysis and automatic extraction of grayscale thresholds, or in response to user interaction. The target bony landmarks include at least the skull base, mandibular angle, or mental foramen. A spatial position constraint vector is constructed based on the registration anchors to limit the search space boundary of the subsequent registration optimization process and avoid local optimal solutions caused by missing or deformed anatomical structures in the lesion area.

[0008] In one or more embodiments, the three-dimensional spatial registration employs a hierarchical rigid registration algorithm based on anatomical constraints: Construct a cost function with the goal of maximizing mutual information. , where MI(I CT ,I PlanLet be the mutual information measure between the oral CT imaging data and the localization CT images in the radiotherapy planning data, and let T be the rigid body transformation matrix to be solved. and , respectively, are the three-dimensional coordinate vectors of the corresponding registration anchor points in the oral CT image data and the radiotherapy planning data, and λ is the anchor point constraint penalty factor; The optimal rigid body transformation matrix T is obtained by solving the problem using a multi-resolution pyramid optimization strategy. opt ; Through the optimal rigid body transformation matrix T opt The first irradiation dose information is transformed to the coordinate system of the oral CT image data to generate a registered third irradiation dose, which has the same voxel grid as the first bone volume data.

[0009] In one or more embodiments, the three-dimensional spatial registration further includes: For nonlinear deformation regions with postoperative tissue deformation, non-rigid correction is performed after rigid registration: Construct a nonlinear deformation field model based on the B-spline free deformation model, using the optimal rigid body transformation matrix T. opt The transformed space is the initial reference space; Construct a comprehensive objective function that includes an image similarity measure and a deformation smoothing regularization term, the expression of which is: E total =E similarity +α(x)*E smooth , of which E similarity For similarity terms, E smooth For smoothing, α(x) is the spatial adaptive regularization weight coefficient. The weight map is dynamically generated using a Gaussian decay function based on the Euclidean distance between the current voxel and the registration anchor point. Within a preset neighborhood centered on the registration anchor point, a high regularization weight is applied to maintain rigid constraints; in the target region where there is local nonlinear deformation between the radiotherapy planning data and the oral CT image data, a low regularization weight is applied to release deformation degrees of freedom. Solving the comprehensive objective function yields the spatial displacement field describing the local nonlinear displacement; Based on the spatial displacement field, the registered dose data is elastically mapped to generate nonlinear registered dose data that conforms to the anatomical structure of the oral CT image data.

[0010] In one or more embodiments, constructing a three-dimensional digital fusion model comprising bone mineral density and radiation dose attributes for each voxel includes: Based on the registered third irradiation dose, a multidimensional tensor data structure consistent with the spatial grid of the oral CT image data is constructed. Based on a preset CT value and bone density conversion function, the voxel gray values ​​of oral CT image data are mapped to bone density attribute values, which serve as the first attribute channel of the tensor data structure. The voxel dose value in the registered third irradiation dose is determined as the irradiation dose attribute value and used as the second attribute channel of the tensor data structure. The multidimensional tensor data structure is encapsulated so that each voxel node in the fusion model corresponds to a multidimensional feature vector containing spatial coordinates, bone density attribute values, and irradiation dose attribute values, thereby establishing a correlation mapping between spatial location, bone tissue physical characteristics, and cumulative irradiation dose, forming a data index for voxel-level multimodal parameter fusion.

[0011] In one or more embodiments, the automatic extraction of the second irradiation dose and second bone mass parameter statistics of the target voxel set corresponding to the selected implantation site based on the fusion model includes: In response to the interactive selection operation of the three-dimensional digital fusion model, the spatial coordinates of the implantation site and the specification parameters of the implant to be implanted are analyzed. Centered on the spatial coordinates, a three-dimensional spatial query mask simulating the shape of the implant is constructed based on the specifications of the implant to be implanted, and voxels falling within the mask are selected to generate a target voxel set; Based on the tensor data structure of the three-dimensional digital fusion model, the bone density attribute value and radiation dose attribute value of each voxel in the target voxel set are indexed and extracted in parallel to form a multidimensional feature dataset. Statistical analysis is performed on the multidimensional feature dataset to calculate a second bone mass parameter statistic that characterizes the bone mass status of the implantation site, and a second radiation dose statistic that characterizes the cumulative radiation risk of the implantation site is calculated simultaneously.

[0012] In one or more embodiments, the safety assessment of the second irradiation dose and the second bone mass parameter statistics based on a preset rule for determining both radiotherapy dose and bone mass includes: The preset radiotherapy dose risk threshold range and bone mass safety threshold range are retrieved. The radiotherapy dose risk threshold range includes at least the critical value of high-risk and contraindicated doses, and the bone mass safety threshold range includes at least the lower limit value of bone mineral density and the lower limit value of bone mass volume percentage. Construct a two-parameter risk grading matrix and map the second radiation dose statistic and the second bone mass parameter statistic to the corresponding coordinate positions in the matrix; The joint judgment logic is as follows: if the second radiation dose statistic exceeds the contraindication dose threshold, it is directly judged as an absolute contraindication level; if the second radiation dose statistic is in the high-risk range and the second bone mass parameter statistic is lower than the lower limit of bone mineral density, it is judged as a relative contraindication level; if the second radiation dose statistic is lower than the preset safe dose and the second bone mass parameter statistic meets the implantation requirements, it is judged as a safe level. Based on the results of the joint determination, a corresponding risk quantification score is generated, which is used to characterize the probability of radiation-induced osteonecrosis and the risk of implant osseointegration failure.

[0013] A second aspect of the present invention also provides a risk assessment system for dental implants based on radiotherapy dose and bone fusion, realizing the risk assessment method for dental implants based on radiotherapy dose and bone fusion as described in this embodiment, including: The data acquisition module is used to acquire radiotherapy planning data and oral CT image data of the target object. The radiotherapy planning data includes spatially distributed first radiation dose information, and the oral CT image data includes at least first bone mass parameter information of the jawbone. The fusion model construction module is used to perform three-dimensional spatial registration between the radiotherapy planning data and the oral CT image data, map the first irradiation dose information to the voxel space of the oral CT image data, and construct a three-dimensional digital fusion model containing the bone density attribute and irradiation dose attribute of each voxel. The parameter extraction module is used to automatically extract the second irradiation dose and second bone mass parameter statistics of the target voxel set corresponding to the selected implantation site based on the fusion model in response to the selection of any implantation site on the three-dimensional digital fusion model. The risk assessment module is used to assess the safety of the second irradiation dose and the second bone mass parameter statistics based on the preset radiotherapy dose and bone mass joint determination rules, and output the risk assessment information of the implantation site according to the assessment results.

[0014] A third aspect of the present invention also provides an electronic device, comprising: The memory is used to store the processing program; A processor, when executing the processing program, implements the risk assessment method for dental implantation based on radiotherapy dose and bone fusion as described in the embodiments of the present invention.

[0015] A fourth aspect of the present invention also provides a readable storage medium, characterized in that the readable storage medium stores a processing program, which, when executed by a processor, implements the risk assessment method for dental implantation based on radiotherapy dose and bone fusion as described in the embodiments of the present invention.

[0016] Compared with the prior art, the present invention has the following beneficial effects: This invention proposes a risk assessment method for dental implants based on the fusion of radiotherapy dose and bone volume. Through deep fusion of multimodal data, adaptive flexible registration, voxel-level quantization extraction, and dual-parameter joint judgment steps, it systematically solves the technical problems of existing technologies, such as the separation of radiotherapy dose and bone volume data, single assessment dimension, and strong subjectivity, bringing the following significant beneficial effects: 1. This invention constructs a three-dimensional digital fusion model and a multi-dimensional tensor data structure, mapping radiotherapy dose data (from TPS) and bone anatomical data (from CBCT) with distinctly different physical properties to a unified spatial coordinate system. This allows each voxel to simultaneously carry the dual attributes of "bone density" and "irradiation dose," solving the problem that doctors cannot simultaneously obtain radiotherapy dose and bone mass information of anatomical locations from the same perspective. It achieves a leap from single bone mass assessment to panoramic assessment of "bone mass + dose" dual parameters, providing a complete data foundation for implant repair in radiotherapy patients.

[0017] 2. This invention employs a hierarchical strategy of preprocessing-rigid registration-non-rigid correction, introducing spatially adaptive regularization weight coefficients and using a Gaussian decay function to dynamically adjust the constraint strength based on the distance between the voxel and the anchor point: maintaining rigid constraints in stable bone structure areas and releasing deformation degrees of freedom in postoperative defect or transplant reconstruction areas. This effectively overcomes the nonlinear deformation problems often associated with jawbone defects, soft tissue collapse, and transplant bone reconstruction in post-radiotherapy patients, avoids the large errors generated by traditional rigid registration in deformation areas, ensures voxel-level accuracy of dose mapping, and significantly improves the accuracy of risk assessment.

[0018] 3. This invention abandons the traditional single-point measurement method, constructs a three-dimensional spatial query mask simulating the implant morphology, directly filters out all voxel sets within the real three-dimensional space occupied by the implant, and extracts statistics in parallel using tensor structures. This solves the problem that traditional evaluation methods cannot reflect the overall three-dimensional environment of the implant. By calculating D... max and B min Extreme value indicators can capture local high-risk hotspots or weak points in bone that are masked by average values, making the assessment results more consistent with the actual clinical stress and stress environment, and improving the reliability of surgical planning.

[0019] 4. This invention constructs a dual-parameter risk grading matrix of radiotherapy dose risk threshold and bone mass safety threshold, and executes joint judgment logic (such as logical judgment of dose and bone mass). This avoids the risk of osteonecrosis caused by focusing only on bone mass and ignoring radiotherapy risks, and also avoids the over-medicalization behavior of rejecting patients based solely on radiotherapy history. Through quantitative scoring and grading output, subjective judgments relying on doctors' experience are transformed into standardized objective data support, providing a scientific "double insurance" for selecting suitable implantation cases. Attached Figure Description

[0020] Figure 1 This is a schematic diagram illustrating the steps of risk assessment for dental implantation based on radiotherapy dose and bone fusion in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the design of implant placement positions on a three-dimensional digital fusion model in an embodiment of the present invention. Detailed Implementation

[0021] 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, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Example 1 The term "comprising" and its variations as used herein are open-ended inclusion, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0023] It should be noted that the concepts of "first" and "second" mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0024] It should be noted that the terms "a" and "a plurality of" used in this application disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0025] Example 1 like Figure 1 As shown, this embodiment of the invention provides a risk assessment method for dental implantation based on radiotherapy dose and bone fusion, specifically for dental implantation in radiotherapy patients, including the following steps: Step S1: Acquire radiotherapy planning data and oral CT imaging data of the target subject. The radiotherapy planning data includes spatially distributed first radiation dose information, and the oral CT imaging data includes at least first bone mass parameters of the jawbone, such as jawbone height, width, and bone mineral density. In existing clinical workflows, radiotherapy data and oral imaging data are stored in different systems (TPS and PACS), with heterogeneous data formats, forming "information silos." This makes it difficult for doctors to simultaneously obtain dose and bone mass information for a specific anatomical location from the same perspective. To solve this technical problem, this step acquires two types of heterogeneous data through a standardized interface, providing a standardized data source for subsequent multimodal data fusion and breaking down data barriers between departments.

[0026] Specifically, the radiotherapy planning data originates from the hospital's radiotherapy planning system. This data is in DICOM RT format and contains three-dimensional spatial dose distribution information. It typically consists of a localization CT image sequence and a strictly registered three-dimensional dose distribution matrix. Each voxel has a corresponding radiation dose value, i.e., the first irradiation dose information, in Gy, with a spatial resolution of at least 2.5 mm. The oral CT image data originates from a cone-beam computed tomography (CBCT) system, which is in DICOM format and contains three-dimensional morphology and bone density information of the jawbone. CBCT images have higher spatial resolution, making them suitable for observing trabecular bone structure and clearly displaying the subtle anatomical features of the cortical and cancellous bone of the jawbone. Each voxel in this data has a corresponding CT value. This CT value is converted into an actual bone density value using a calibration formula based on a CT density phantom, thereby obtaining the first bone mass parameter information. The specific form of the calibration formula is: BMD = a * HU + b, where BMD is the target output bone density (unit: mg / cm³). 3 ); HU is the input CT value; a and b are device-specific calibration coefficients. a represents the linear proportional relationship between the CT value and bone mineral density, reflecting the sensitivity of the CT equipment to bone tissue; b represents the linear offset between the CT value and bone mineral density, reflecting the zero-point offset of the CT equipment. a and b are obtained by scanning a CT density phantom of known density—the phantom contains cylindrical inserts of various known physical densities (such as air, water, cortical bone equivalent materials, and cancellous bone equivalent materials). By measuring the average CT value of each insert and performing linear regression analysis, the device-specific calibration coefficients can be obtained. This transformation converts the abstract CT value into a physically meaningful bone mineral density value, providing a quantitative basis for subsequent bone mass assessment.

[0027] Step S2: Perform three-dimensional spatial registration between the radiotherapy plan data and the oral CT image data, map the first irradiation dose information to the voxel space of the oral CT image data, and construct a three-dimensional digital fusion model containing the bone density attribute and irradiation dose attribute of each voxel. This abandons the traditional, crude concept of "target dose" in clinical practice, and accurately maps the dose information to each voxel of the oral CT image data through inverse spatial transformation and dose resampling algorithm.

[0028] Because CBCT images are typically accompanied by significant scattering noise, and different imaging parameters (such as KV values) vary between devices, resulting in substantial differences in image grayscale distribution, direct registration can lead to misregistration or decreased accuracy. Furthermore, post-radiotherapy patients often experience jawbone defects or bone graft reconstruction, causing nonlinear deformation of local anatomical structures, making rigid registration alone insufficient for achieving the required accuracy. To address these issues, this step employs a layered registration strategy: "preprocessing - rigid registration - non-rigid correction." Step S201: Coordinate System Definition Using the spatial coordinate system of the oral CT image data as the reference coordinate system and the spatial coordinate system of the radiotherapy planning data as the floating coordinate system, a registration benchmark is established to ensure that the final fusion model is based on the oral clinical perspective.

[0029] Step S202: Image filtering preprocessing Anisotropic diffusion filtering was performed on the localization CT images in the oral CT image data and the radiotherapy planning data, respectively, to suppress image noise while preserving high-frequency feature signals of the jawbone cortical bone edge. First, anisotropic diffusion filtering was performed. Using the Perona-Malik model, diffusion was suppressed in edge regions with large image gradients (such as cortical bone edges) and promoted in flat regions with small gradients (such as the medullary cavity), thereby preserving high-frequency edge features while suppressing noise. Specifically, the diffusion coefficient... Defined as: .in, Let K be the gradient magnitude of the image, and K be the gradient threshold parameter. When the gradient magnitude is large (edge ​​regions), the diffusion coefficient approaches 0, and the edges are preserved; when the gradient magnitude is small (flat regions), the diffusion coefficient approaches 1, and the noise is smoothed. Through this processing, while suppressing CBCT scattering noise, the cortical bone edge features are accurately preserved, providing high-quality anatomical images for subsequent registration and significantly improving the signal-to-noise ratio of the image.

[0030] Step S203: Grayscale normalization processing A grayscale normalization operation based on the bone density distribution histogram is performed to map the grayscale values ​​of two sets of image data acquired at different time points and from different imaging devices to a unified standardized interval. This eliminates the grayscale offset field introduced by differences in imaging parameters and solves the problem of inaccurate mutual information measurement caused by inconsistent image grayscale values. This step achieves grayscale normalization through histogram matching: First, the grayscale histogram of the reference image (oral CT image data) is calculated to determine its grayscale distribution characteristics; then, the grayscale histogram of the image to be normalized (radiotherapy planning and localization CT) is calculated to establish the mapping relationship between the grayscale values ​​of the two images; finally, the grayscale value of each voxel in the image to be normalized is converted into the corresponding value in the grayscale space of the reference image through a mapping function. The specific mapping function is as follows: , of which F src Let be the cumulative distribution function of the source image. This is the inverse function of the cumulative distribution function of the reference image. This operation eliminates grayscale shifts caused by different times and devices, ensuring a consistent basis for mutual information calculation and guaranteeing the accuracy of subsequent registration algorithms.

[0031] Step S204: Registration Anchor Point Extraction and Constraint Construction In the reference coordinate system and the floating coordinate system, anatomically invariant target bony landmarks are used as registration anchors based on curvature feature analysis and automatic extraction of grayscale thresholds, or in response to user interaction. The target bony landmarks include at least the skull base, mandibular angle, or mental foramen. A spatial position constraint vector is constructed based on the registration anchors to limit the search space boundary of the subsequent registration optimization process and avoid local optimal solutions caused by missing or deformed anatomical structures in the lesion area.

[0032] This step guides the registration direction using prior anatomical knowledge. Specifically, based on curvature feature analysis (such as principal curvature and shape index) and grayscale thresholding, anatomically invariant bony landmarks are automatically identified in the image. These landmarks (skull base, angle of the mandible, mental foramen) are typically unaffected by tumor invasion after radiotherapy and have stable anatomical positions, serving as anchor points to avoid interference from lesion areas on registration. A spatial position constraint vector is then constructed based on these anchor points. ,in, For anchor point coordinates in radiotherapy planning data, Let represent the coordinates of the corresponding anchor point in the oral CT image data, and T be the transformation matrix to be solved. This constraint vector is added as a soft constraint to the cost function of subsequent registration optimization, guiding the registration towards the clinically desired direction and avoiding local optimal solutions caused by local lesion areas.

[0033] Step S205: Hierarchical rigid registration based on anatomical constraints In this embodiment, the 3D spatial registration adopts a hierarchical rigid registration algorithm based on anatomical constraints: for the macroscopic position and angle deviations between the two sets of data, a cost function is constructed with the goal of maximizing mutual information. , where MI(I CT ,I Plan ) represents the mutual information measure between oral CT image data and localization CT images in radiotherapy planning data, where T is the rigid body transformation matrix to be solved, including rotational and translational components; and denoted as the three-dimensional coordinate vectors of the corresponding registration anchor points in the oral CT image data and radiotherapy planning data, respectively. λ is the anchor point constraint penalty factor, used to balance the weights between the mutual information term and the anatomical landmark constraint term. The optimal rigid body transformation matrix T is obtained by solving a multi-resolution pyramid optimization strategy. opt ; through the optimal rigid body transformation matrix T opt The first radiation dose information is transformed to the coordinate system of the oral CT image data to generate a registered third radiation dose. The registered third radiation dose has the same voxel grid as the first bone volume data. The dose value of each voxel is the mapped dose value after resampling by the optimal rigid body transformation matrix.

[0034] In one embodiment, a multi-resolution pyramid optimization strategy is used for solving the problem: In the first stage, global coarse registration is performed in the low-resolution layer: the oral CT image data and the radiotherapy planning data are downsampled respectively, and the transformation parameter space is searched in the down-resolution image space with a large step size to quickly eliminate the overall spatial position deviation between the two sets of data and obtain the initial transformation matrix T0. In the second stage, local fine registration is performed at the original resolution layer: using the initial transformation matrix T0 as the optimization starting point and the anatomical region defined by the spatial position constraint vector as the search boundary, the cost function E is adjusted using L-BFGS or gradient descent. align Iterative optimization is performed; the optimization terminates when the change in the cost function value is less than a preset convergence threshold or the number of iterations reaches a preset maximum value, thus obtaining the optimal rigid body transformation matrix T. opt ; Through the optimal rigid body transformation matrix T opt The radiotherapy planning data is transformed to the coordinate system of the oral CT image data, achieving sub-voxel-level high-precision alignment of the two sets of data on the overall anatomical skeleton, realizing macroscopic alignment of images in three-dimensional space, and solving the problem of coordinate system inconsistency.

[0035] Step S206: Non-rigid correction based on spatial adaptive regularization In this embodiment, 3D spatial registration also includes non-rigid correction based on spatial adaptive regularization. Specifically, for nonlinear deformation regions with postoperative tissue deformation, non-rigid correction is performed after rigid registration. Here, postoperative tissue deformation includes local nonlinear deformation caused by postoperative tissue defects or bone graft reconstruction: a nonlinear deformation field model based on the B-spline free deformation model is constructed, with the optimal rigid body transformation matrix T... opt The transformed space serves as the initial reference space; a comprehensive objective function is constructed, comprising an image similarity measure and a deformation smoothing regularization term, with the expression: E total =E similarity +α(x)*E smooth , of which E similarity For similarity terms, E smooth For the smoothing term, a spatially adaptive regularization weight coefficient α(x) is introduced. Based on the Euclidean distance between the current voxel and the registration anchor point, a Gaussian decay function is used to dynamically generate the weight map. ,in, α is the Euclidean distance between the current voxel and the nearest registration anchor point. max and α min These are the high and low regularization weight thresholds (usually α). max Take a value of 1.0-2.0, α min (Take a value of 0.01-0.1), where σ is the Gaussian kernel width. Within a preset neighborhood centered on the registration anchor point, a high regularization weight is applied to maintain rigid constraints; in the target region where there is local nonlinear deformation between the radiotherapy planning data and the oral CT image data, a low regularization weight is applied to release the deformation degrees of freedom.

[0036] Within a preset neighborhood centered on the registration anchor point, a high regularization weight is applied to maintain rigid constraints; in the target region where there is local nonlinear deformation between the radiotherapy planning data and the oral CT image data, a low regularization weight is applied to release deformation degrees of freedom. Solve the comprehensive objective function to obtain the spatial displacement field describing the local nonlinear offset; perform elastic mapping on the registered dose data based on the spatial displacement field to generate nonlinear registered dose data that fits the anatomical structure of the oral CT image data.

[0037] Step S207: Dose resampling based on inverse spatial transformation and gradient preservation To prevent dose risk underestimation due to interpolation smoothing effects, this step employs a dose resampling algorithm with inverse spatial transformation and gradient preservation. For each target voxel in the oral CT image data, its corresponding floating-point coordinates in the radiotherapy planning data source space are calculated based on the optimal rigid body transformation matrix and the inverse transformation of the spatial displacement field. A trilinear interpolation algorithm is used to calculate the dose value, and a gradient preservation constraint mechanism is introduced: based on the local dose gradient magnitude, an anisotropic interpolation kernel is used in high-gradient regions (such as dose drop boundaries) to reduce the interpolation weight along the gradient direction, thereby preserving the high-risk dose boundary characteristics in the radiotherapy dose distribution and ensuring the accuracy and reliability of the generated third irradiation dose data.

[0038] Step S208: Construction and Verification of 3D Digital Fusion Model In this embodiment, the construction of a three-dimensional digital fusion model including the bone mineral density attribute and irradiation dose attribute of each voxel includes: First, based on the registered third irradiation dose, a multidimensional tensor data structure consistent with the spatial grid of the oral CT image data is constructed. Assuming the oral CT image data has dimensions X×Y×Z and voxel sizes Δx×Δy×Δz, the multidimensional tensor data structure F has the same dimensions, with each element being a voxel node.

[0039] Secondly, based on a preset CT value and bone density conversion function, the voxel gray values ​​of the oral CT image data are mapped to bone density attribute values, which serve as the first attribute channel of the tensor data structure: F BMD (x,y,z)=a*I CT (x,y,z)+b, where I CT (x,y,z) represents the CT value of the oral CT image data at position (x,y,z), and a and b are the device-specific calibration coefficients determined in step S1.

[0040] Next, the voxel dose value in the registered third irradiation dose is determined as the irradiation dose attribute value, serving as the second attribute channel of the tensor data structure: F Dose (x,y,z)=D registered (x,y,z), where D registered (x,y,z) represents the dose value mapped to position (x,y,z) after rigid registration and non-rigid correction.

[0041] Finally, the multidimensional tensor data structure is encapsulated so that each voxel node in the fusion model corresponds to a multidimensional feature vector V(x,y,z)=[x,y,z,F] containing spatial coordinates, bone density attribute values, and radiation dose attribute values. BMD (x,y,z),F Dose[x,y,z], thereby establishing a correlation mapping between spatial location, bone tissue physical characteristics, and cumulative radiation dose, forming a data index for voxel-level multimodal parameter fusion. Furthermore, it includes a step for verifying the spatiotemporal consistency of the fusion model: based on the mapping positions of the stable bone landmarks in the fusion model, calculating the registration residual metric. In response to the registration residual metric exceeding a preset clinical safety threshold, an adaptive correction mechanism is triggered to adjust the regularization weight coefficients in non-rigid registration and iteratively execute the mapping steps. A spatial confidence distribution map C(x) = 1 - local registration residual (x) / clinical safety threshold is generated based on the registration residual metric. The confidence distribution map is displayed in pseudo-color overlay in a three-dimensional visualization interface. For areas with large registration residuals, the data confidence label is synchronously associated when outputting risk assessment information. Users can view specific registration residual values ​​by clicking on the confidence area to assess the reliability of the data.

[0042] Step S3: In response to the selection of any implantation site on the three-dimensional digital fusion model, the second irradiation dose and the second bone mass parameter statistics of the target voxel set corresponding to the selected implantation site are automatically extracted based on the fusion model; Step S301: In response to the interactive selection operation of the three-dimensional digital fusion model, the spatial coordinates of the implantation site and the specification parameters of the implant to be implanted are parsed; the three-dimensional spatial coordinates P(x,y,z) are determined in response to the user's click, selection or coordinate input operation on the three-dimensional digital fusion model; the implant geometric parameters include at least the implant length L and diameter D, which are derived from a clinically preset implant database.

[0043] Step S302: Using the spatial coordinates as the center, construct a three-dimensional spatial query mask simulating the implant morphology based on the specifications of the implant to be implanted, and filter voxels falling within the mask to generate a target voxel set; specifically, based on the three-dimensional spatial coordinates and implant geometric parameters, perform spatial indexing in the three-dimensional digital fusion model to construct a three-dimensional region of interest surrounding the implantation site; the three-dimensional region of interest is a cylindrical or conical voxel space generated along a preset implantation axis with the three-dimensional spatial coordinates as the implant neck center point, and its spatial range precisely matches the bone tissue area occupied by the implant to be implanted. Traverse all voxels within the three-dimensional region of interest, and use a spatial position determination algorithm (such as ray casting or signed distance field SDF) to filter out voxels completely contained within the implant geometric contour, generating a target voxel set V = {v i |i = 1, 2, ...,N}, where N is the total number of target voxels, and the bone density attribute value B of each voxel in the target voxel set is read in parallel. i With the radiation dose attribute value Di .

[0044] Step S303: Based on the tensor data structure of the three-dimensional digital fusion model, the bone density attribute value and radiation dose attribute value of each voxel in the target voxel set are indexed and extracted in parallel to form a multidimensional feature dataset. Since the fusion model constructed in step S2 uses a tensor data structure, each voxel node has a definite index address. This step utilizes multi-threaded parallel computing technology to simultaneously read the voxel values ​​of each voxel in the target voxel set. i Bone mineral density attribute value B i With the radiation dose attribute value D i This parallel extraction mechanism greatly improves data reading efficiency, enabling millisecond-level response times for massive voxel data.

[0045] Step S304: Perform statistical analysis on the multidimensional feature dataset to calculate a second bone mass parameter statistic characterizing the bone mass status of the implantation site, and simultaneously calculate a second radiation dose statistic characterizing the cumulative radiation risk of the implantation site (such as mean and minimum bone mineral density). The second radiation dose statistic characterizing the radiation risk of the implantation site includes the maximum radiation dose D of all voxels within the target voxel set. max ==max(D i ): Used to detect dose hotspots within the planting area. D max The upper limit of risk directly related to radiation-induced osteonecrosis; if this value exceeds the contraindication threshold such as 50 Gy, even with a low average dose, it indicates an extremely high risk; mean radiation dose. Used to assess the overall radiation load level of the bone tissue surrounding the implant.

[0046] A second bone mass parameter statistic characterizing the bone mass status of the implantation site, the second bone mass parameter statistic including the minimum bone mineral density B of all voxels in the target voxel set. min =min(B i ): Used to identify weak points in the bone within the implantation area, B min The location where the implant's initial stability is weakest is a key indicator for determining whether bone augmentation surgery is needed; mean bone mineral density. Used to assess the overall bone quality of the implantation area and assist in selecting the appropriate implant model, the system automatically calculates and records the dose value within the specific delineated area, rather than directly referencing the target dose prescription in the radiotherapy plan, thus achieving a precise transition from "target dose" to "site dose".

[0047] Step S4: Based on the preset combined rule for determining radiotherapy dose and bone mass, a safety assessment is performed on the second irradiation dose and the statistical values ​​of the second bone mass parameters. Based on the assessment results, risk assessment information for the implantation site is output, specifically including: Retrieve preset radiotherapy dose risk threshold intervals and bone mass safety threshold intervals. The radiotherapy dose risk threshold intervals include at least high-risk and contraindicated dose thresholds, and the bone mass safety threshold intervals include at least lower limits for bone mineral density and bone mass volume percentage. Construct a two-parameter risk grading matrix and map the second radiation dose statistics and the second bone mass parameter statistics to the corresponding coordinate positions in the matrix. Execute the joint decision logic: If the second radiation dose statistic D max If the dose exceeds the contraindication threshold, such as >50 Gy, it is directly determined to be an absolute contraindication. This determination corresponds to the pathological state of irreversible damage to the microvascular endothelial cells of the jawbone, osteocyte inactivation, and complete loss of bone repair capacity, and the clinical decision is "contraindicated for implantation". If the second radiation dose statistic is in the high-risk range, such as 40-50 Gy, and the second bone mass parameter statistic B min A bone mineral density below the lower limit is considered a relative contraindication. This contraindication corresponds to clinical scenarios where bone tissue compensatory capacity is exhausted and initial implant stability is unreliable. The clinical decision is to require multidisciplinary consultation and recommend preoperative bone augmentation surgery. If the second radiation dose statistic D... max If the dose is below the preset safety level and the second bone mass parameter statistical value meets the implantation requirements, it is determined to be at a safe level. This determination corresponds to a state of good bone tissue microenvironment and possessing the biological conditions for osseointegration, and the clinical decision is that routine implantation is possible. Based on the results of the combined determination, a corresponding risk quantification score is generated. This risk quantification score is used to characterize the probability of radiation-induced osteonecrosis and the risk of implant osseointegration failure. Figure 2 As shown, the core of this embodiment lies in using a "dual-quantity" digital 3D model based on both radiotherapy dose and bone volume to provide double-insurance assessment of implant site safety. First, the radiotherapy dose for the implant site is precisely delineated, rather than a uniform overall target area dose. The green and pink circles in the figure represent the delineated dose circles for the implant site; the system automatically displays the radiation dose at that location. Simultaneously, the implant placement is designed on CBCT to ensure sufficient bone volume in the implant. Similarly... Figure 2As shown in the diagram for the same tooth position, implant placement can only proceed if both of the above conditions are met: a safe radiation dose (<50 Gy) and a safe bone mass (e.g., 200 HU). This ensures the feasibility and safety of implant placement for radiotherapy patients. The dose-bone mass joint determination rule can be understood as including a multi-threshold combination determination logic: setting at least one dose threshold and at least one bone mineral density threshold; if the extracted dose value exceeds the first dose threshold, or the bone mineral density value is lower than the first bone mineral density threshold, the corresponding area is determined to be a prohibited area; if the extracted dose value is between the first and second dose thresholds, and the bone mineral density value is not lower than the first bone mineral density threshold, the corresponding area is determined to be a risk area; if the extracted dose value is lower than the second dose threshold, and the bone mineral density value is not lower than the first bone mineral density threshold, the corresponding area is determined to be a safe area.

[0048] In one embodiment, the dose-bone mass joint determination rule includes a continuous risk scoring function: constructing a risk score R = f(D,B) = α·D norm + β·(1 - B norm ), where f is a preset weighting function, D is the dose value, B is the bone mineral density value, and D norm and B norm The normalized results are the dose value and bone mineral density value, respectively, with α and β being preset weighting coefficients; based on the calculated R value, the risk level is divided within the continuous interval of 0-1, generating a smooth risk distribution cloud map.

[0049] In one embodiment, the step of outputting the risk assessment information of the implantation site based on the assessment results specifically includes: generating an assessment report containing risk level identifiers, risk quantification scores, and clinical recommendations based on the risk grading matrix; color-coding and rendering the implantation site and its neighborhood on the three-dimensional digital fusion model, wherein different colors map to different risk levels, and highlighting areas with excessive dose or insufficient bone mass; and interactively displaying the radiation dose-volume histogram curve and bone density distribution histogram within the area in response to a user's query command for the implantation site, assisting clinicians in developing personalized implantation surgery plans or avoidance strategies.

[0050] In one or more embodiments of this application, the implantation location, angle, and size of the implant are automatically determined within a safe area based on the evaluation results. This includes: generating multiple candidate implantation locations within a safe or risk area using a heuristic search algorithm; simulating the implantation state of implants of different sizes for each candidate location, and verifying whether the implantation has entered a prohibited area or a severely deficient bone area using a collision detection algorithm; and comprehensively evaluating the verified candidate schemes based on at least one of the following indicators: dose distribution uniformity, bone distribution uniformity, and distance from important anatomical structures, to output the optimal implantation scheme. The method also includes local optimization of the implantation angle and size of the candidate implants: fine-tuning the implantation angle near the candidate location to reduce the overall dose received by the implant or increase the bone contact area; and adjusting the implant length and diameter within a preset specification range to maximize initial stability while meeting safety constraints.

[0051] In one or more embodiments of this application, the evaluation method further includes: importing the dual-quantity fusion three-dimensional digital model and the personalized implant repair plan into an augmented reality device, registering the patient's intraoperative position in real time through optical tracking technology, and overlaying and displaying the safe zone, risk zone, prohibited zone and predetermined implant position in the surgical field of view to achieve real-time intraoperative navigation.

[0052] In one or more embodiments of this application, the evaluation method further includes: uploading the radiotherapy plan data, oral imaging data and personalized implant restoration plan to a cloud platform, supporting multidisciplinary doctors to view, edit and discuss them simultaneously on different terminals, and recording collaborative operation history to generate consultation reports.

[0053] In one or more embodiments of this application, the evaluation method further includes: acquiring the patient's biological marker data, constructing a multimodal prognostic prediction model by combining the biological marker data with dose values ​​and bone mineral density values, and using the model to optimize the threshold of the dose-bone mass joint determination rule or the selection of implantation scheme.

[0054] The core principle of this embodiment lies in constructing a three-in-one digital mapping mechanism of "space-dose-bone volume". By using high-precision three-dimensional spatial registration technology, it breaks down the data barriers between the radiotherapy planning system and the oral imaging system, accurately anchoring the abstract physical quantity of radiation dose to the specific anatomical voxel space of the jawbone. This achieves a refined leap from macroscopic target area dose assessment to microscopic site dose assessment. On this basis, three-dimensional morphological masking technology is used to simulate the real implantation environment, simultaneously quantifying and extracting bone volume parameters and radiation dose statistics of the target area. The "dose-bone volume" dual-parameter joint judgment logic replaces the traditional single-dimensional assessment, thereby realizing the transformation from empirical qualitative judgment to precise quantitative analysis. This effectively avoids the risk of osteonecrosis caused by high-dose radiation and the risk of implant failure caused by insufficient bone volume, providing radiotherapy patients with a scientific, safe and personalized basis for implantation decision-making.

[0055] Example 2 Based on the same inventive concept, this invention provides a risk assessment system for dental implants based on radiotherapy dose and bone fusion, characterized in that it implements the risk assessment method for dental implants based on radiotherapy dose and bone fusion as described in Example 1, including: The data acquisition module is used to acquire radiotherapy planning data and oral CT image data of the target object. The radiotherapy planning data includes spatially distributed first radiation dose information, and the oral CT image data includes at least first bone mass parameter information of the jawbone. The fusion model construction module is used to perform three-dimensional spatial registration between the radiotherapy planning data and the oral CT image data, map the first irradiation dose information to the voxel space of the oral CT image data, and construct a three-dimensional digital fusion model containing the bone density attribute and irradiation dose attribute of each voxel. The parameter extraction module is used to automatically extract the second irradiation dose and second bone mass parameter statistics of the target voxel set corresponding to the selected implantation site based on the fusion model in response to the selection of any implantation site on the three-dimensional digital fusion model. The risk assessment module is used to assess the safety of the second irradiation dose and the second bone mass parameter statistics based on a preset rule for determining the combined radiotherapy dose and bone mass, and outputs the risk assessment information of the implantation site based on the assessment results. The implementation principle of the above module has been described in the foregoing embodiments, and therefore will not be repeated here.

[0056] It should be noted that the division of the various modules in this device / system embodiment is merely a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, these modules can be implemented entirely in software through processing element calls; they can also be implemented entirely in hardware; or some units can be implemented by processing element calls to software, while others can be implemented in hardware.

[0057] Example 3 Based on the same inventive concept, an electronic device is also provided in some embodiments of this application. This electronic device includes a memory and a processor, wherein the memory stores a processing program, and the processor executes the processing program according to instructions. When the processor executes the processing program, the risk assessment method for dental implantation based on radiotherapy dose and bone fusion as described in the foregoing embodiments is realized.

[0058] In some embodiments of this application, a readable storage medium is also provided, which can be a non-volatile readable storage medium or a volatile readable storage medium. The readable storage medium stores instructions that, when executed on a computer, cause an electronic device containing such a readable storage medium to perform the aforementioned risk assessment method for dental implantation based on radiotherapy dose and bone fusion.

[0059] It is understood that the aforementioned risk assessment methods for dental implants based on radiotherapy dose and bone fusion, if implemented as software functional modules and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0060] Computer-readable storage media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium may also be any readable medium other than a readable storage medium that can transmit, propagate, or transfer a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0061] The program code for executing the technical solutions disclosed in this application can be written in any combination of one or more programming languages. These programming languages ​​include object-oriented programming languages ​​such as Python and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A risk assessment method for dental implants based on radiotherapy dose and bone fusion, characterized in that, include: Acquire radiotherapy planning data and oral CT image data of the target object, wherein the radiotherapy planning data includes spatially distributed first radiation dose information, and the oral CT image data includes at least first bone mass parameter information of the jawbone; The radiotherapy plan data and the oral CT image data are registered in three-dimensional space, and the first irradiation dose information is mapped to the voxel space of the oral CT image data to construct a three-dimensional digital fusion model containing the bone density attribute and irradiation dose attribute of each voxel. In response to the selection of any implantation site on the three-dimensional digital fusion model, the second irradiation dose and the second bone mass parameter statistics of the target voxel set corresponding to the selected implantation site are automatically extracted based on the fusion model. The safety of the second irradiation dose and the second bone mass parameter statistics is assessed based on the preset joint determination rules of radiotherapy dose and bone mass, and the risk assessment information of the implantation site is output according to the assessment results.

2. The method for risk assessment of dental implants based on radiotherapy dose and bone fusion according to claim 1, characterized in that, Before performing three-dimensional spatial registration of the radiotherapy planning data with the oral CT image data, the following steps are included: The spatial coordinate system of the oral CT image data is used as the reference coordinate system, and the spatial coordinate system of the radiotherapy plan data is used as the floating coordinate system; Anisotropic diffusion filtering was performed on the positioning CT images in the oral CT image data and the radiotherapy planning data respectively to suppress image noise while preserving the high-frequency feature signals of the jawbone cortical bone edge; Perform gray-level normalization based on bone density distribution histogram to map the gray-level values ​​of two sets of image data collected at different time points and from different imaging devices to a unified standardized range, thereby eliminating the gray-level offset field introduced by differences in imaging parameters. In the reference coordinate system and the floating coordinate system, anatomically invariant target bony landmarks are used as registration anchors based on curvature feature analysis and automatic extraction of grayscale thresholds, or in response to user interaction. The target bony landmarks include at least the skull base, mandibular angle, or mental foramen. A spatial position constraint vector is constructed based on the registration anchors to limit the search space boundary of the subsequent registration optimization process and avoid local optimal solutions caused by missing or deformed anatomical structures in the lesion area.

3. The method for risk assessment of dental implants based on radiotherapy dose and bone fusion according to claim 1, characterized in that, The three-dimensional spatial registration adopts a hierarchical rigid registration algorithm based on anatomical constraints: Construct a cost function with the goal of maximizing mutual information. , where MI(I CT ,I Plan Let be the mutual information measure between the oral CT imaging data and the localization CT images in the radiotherapy planning data, and let T be the rigid body transformation matrix to be solved. and , respectively, are the three-dimensional coordinate vectors of the corresponding registration anchor points in the oral CT image data and the radiotherapy planning data, and λ is the anchor point constraint penalty factor; The optimal rigid body transformation matrix T is obtained by solving the problem using a multi-resolution pyramid optimization strategy. opt ; Through the optimal rigid body transformation matrix T opt The first irradiation dose information is transformed to the coordinate system of the oral CT image data to generate a registered third irradiation dose, which has the same voxel grid as the first bone volume data.

4. The method for risk assessment of dental implants based on radiotherapy dose and bone fusion according to claim 3, characterized in that, The three-dimensional spatial registration also includes: For nonlinear deformation regions with postoperative tissue deformation, non-rigid correction is performed after rigid registration: Construct a nonlinear deformation field model based on the B-spline free deformation model, using the optimal rigid body transformation matrix T. opt The transformed space is the initial reference space; Construct a comprehensive objective function that includes an image similarity measure and a deformation smoothing regularization term, the expression of which is: E total =E similarity +α(x)*E smooth E similarity For similarity terms, E smooth For smoothing, α(x) is the spatial adaptive regularization weight coefficient. The weight map is dynamically generated using a Gaussian decay function based on the Euclidean distance between the current voxel and the registration anchor point. Within a preset neighborhood centered on the registration anchor point, a high regularization weight is applied to maintain rigid constraints; in the target region where there is local nonlinear deformation between the radiotherapy planning data and the oral CT image data, a low regularization weight is applied to release deformation degrees of freedom. Solving the comprehensive objective function yields the spatial displacement field describing the local nonlinear displacement; Based on the spatial displacement field, the registered dose data is elastically mapped to generate nonlinear registered dose data that conforms to the anatomical structure of the oral CT image data.

5. The method for risk assessment of dental implants based on radiotherapy dose and bone fusion according to claim 1, characterized in that, The construction of the three-dimensional digital fusion model, which includes the bone mineral density attribute and radiation dose attribute of each voxel, includes: Based on the registered third irradiation dose, a multidimensional tensor data structure consistent with the spatial grid of the oral CT image data is constructed. Based on a preset CT value and bone density conversion function, the voxel gray values ​​of oral CT image data are mapped to bone density attribute values, which serve as the first attribute channel of the tensor data structure. The voxel dose value in the registered third irradiation dose is determined as the irradiation dose attribute value and used as the second attribute channel of the tensor data structure. The multidimensional tensor data structure is encapsulated so that each voxel node in the fusion model corresponds to a multidimensional feature vector containing spatial coordinates, bone density attribute values, and irradiation dose attribute values, thereby establishing a correlation mapping between spatial location, bone tissue physical characteristics, and cumulative irradiation dose, forming a data index for voxel-level multimodal parameter fusion.

6. The method for risk assessment of dental implants based on radiotherapy dose and bone fusion according to claim 1, characterized in that, The second irradiation dose and second bone mass parameter statistics automatically extracted from the target voxel set corresponding to the selected implantation site based on the fusion model include: In response to the interactive selection operation of the three-dimensional digital fusion model, the spatial coordinates of the implantation site and the specification parameters of the implant to be implanted are analyzed. Centered on the spatial coordinates, a three-dimensional spatial query mask simulating the shape of the implant is constructed based on the specifications of the implant to be implanted, and voxels falling within the mask are selected to generate a target voxel set; Based on the tensor data structure of the three-dimensional digital fusion model, the bone density attribute value and radiation dose attribute value of each voxel in the target voxel set are indexed and extracted in parallel to form a multidimensional feature dataset. Statistical analysis is performed on the multidimensional feature dataset to calculate a second bone mass parameter statistic that characterizes the bone mass status of the implantation site, and a second radiation dose statistic that characterizes the cumulative radiation risk of the implantation site is calculated simultaneously.

7. The method for risk assessment of dental implants based on radiotherapy dose and bone fusion according to claim 1, characterized in that, The safety assessment of the second irradiation dose and the second bone mass parameter statistics based on the preset combined determination rule of radiotherapy dose and bone mass includes: The preset radiotherapy dose risk threshold range and bone mass safety threshold range are retrieved. The radiotherapy dose risk threshold range includes at least the critical value of high-risk and contraindicated doses, and the bone mass safety threshold range includes at least the lower limit value of bone mineral density and the lower limit value of bone mass volume percentage. Construct a two-parameter risk grading matrix and map the second radiation dose statistic and the second bone mass parameter statistic to the corresponding coordinate positions in the matrix; The joint judgment logic is as follows: if the second radiation dose statistic exceeds the contraindication dose threshold, it is directly judged as an absolute contraindication level; if the second radiation dose statistic is in the high-risk range and the second bone mass parameter statistic is lower than the lower limit of bone mineral density, it is judged as a relative contraindication level; if the second radiation dose statistic is lower than the preset safe dose and the second bone mass parameter statistic meets the implantation requirements, it is judged as a safe level. Based on the results of the joint determination, a corresponding risk quantification score is generated, which is used to characterize the probability of radiation-induced osteonecrosis and the risk of implant osseointegration failure.

8. A risk assessment system for dental implants based on radiotherapy dose and bone fusion, characterized in that, Implementing the risk assessment method for dental implants based on radiotherapy dose and bone fusion as described in any one of claims 1 to 7, comprising: The data acquisition module is used to acquire radiotherapy planning data and oral CT image data of the target object. The radiotherapy planning data includes spatially distributed first radiation dose information, and the oral CT image data includes at least first bone mass parameter information of the jawbone. The fusion model construction module is used to perform three-dimensional spatial registration between the radiotherapy planning data and the oral CT image data, map the first irradiation dose information to the voxel space of the oral CT image data, and construct a three-dimensional digital fusion model containing the bone density attribute and irradiation dose attribute of each voxel. The parameter extraction module is used to automatically extract the second irradiation dose and second bone mass parameter statistics of the target voxel set corresponding to the selected implantation site based on the fusion model in response to the selection of any implantation site on the three-dimensional digital fusion model. The risk assessment module is used to assess the safety of the second irradiation dose and the second bone mass parameter statistics based on the preset radiotherapy dose and bone mass joint determination rules, and output the risk assessment information of the implantation site according to the assessment results.

9. An electronic device, characterized in that, include: The memory is used to store the processing program; A processor that, when executing the processing program, implements the risk assessment method for dental implantation based on radiotherapy dose and bone fusion as described in any one of claims 1 to 7.

10. A readable storage medium, characterized in that, The readable storage medium stores a processing program that, when executed by a processor, implements the risk assessment method for dental implantation based on radiotherapy dose and bone fusion as described in any one of claims 1 to 7.

Images